SECTOR REPORT

EQUITY RESEARCH
6 December 2010

Europe
116.30 66.30 dc-07 av r-09 sept-10

Pharmaceuticals

Outperformance

DJ St oxx Healt hcar e Rel. DJ St oxx

Biosimilars: Attacking the last bastion

The copying of biologics has moved up a gear. Cracks are appearing in the walls of the last bastion of resistance to generics with the new law which authorises the marketing of biosimilars in the United States. The arrival in the arena of marketing heavyweights such as Pfizer, which has just signed an agreement with Indian generic maker Biocon, marks a new stage in the process. We are now expecting a widespread arrival of biosimilars when patents expire for most biologics. This is leading to a commoditisation of the status of certain groups, and means, for example, that the recent derating of Roche has a structural component. But it would be a mistake to write off the biologics industry. The economic basis for biosimilar copies is radically different from that of generics. Substitutable biosimilars will be allowed in the United States, but they should remain the exception rather than the rule. Biosimilars will therefore have a different business model from low value added conventional generics: a relatively strong marketing content, low cyclicality and a relatively high profitability. This is a real opportunity for generic makers, but also for pharmaceutical groups, which explains why Pfizer and Merck have suddenly arrived on the scene, and are likely to be followed by others soon. Biosimilar penetration will be highly differentiated by market segment, depending on barriers to entry. Biosimilars will ultimately come to dominate in simple proteins (EPO hormones, etc.) but will remain weak even in the long-term for insulins, because of capital intensity which will not allow significant price cuts. Monoclonal antibodies, protected by being far more complex than conventional proteins, will also feel less pressure and at a later stage. We are therefore still cautious on Roche (Neutral), which is suffering from persistent doubts about its growth capacity, but we maintain our confident stance on insulin producers Novo Nordisk (Buy) and sanofi-aventis (Buy). Stada (Buy) remains one of the few stock market vehicles to play the generic sector in Europe.

Source: Natixis

Analyst(s) Philippe Lanone (33 1) 58 55 05 03 philippe.lanone@natixis.com Batrice Muzard (33 1) 58 55 05 13 beatrice.muzard@natixis.com

Equity Markets Bloomberg access

equity.natixis.com NXSE

Company Novo Nordisk Roche Sanofi-aventis Stada Median

Rating Buy Neutral Buy Buy

Price

Target

P/E (x) 2010 2011 20.2 9.4 7.0 10.4 9.9 24.5 10.5 7.0 11.3 10.9

EV/EBIT (x) 2010 17.9 8.7 5.6 10.4 9.5 2011 15.4 7.7 5.3 9.6 8.7

EV/Turnover (x) 2010 5.7 2.9 2.1 1.4 2.5 2011 5.0 2.6 2.0 1.3 2.3

Distribution of this report in the United States. See important disclosures at the end of this report.

DKK574.00 DKK600.00 CHF138.40 CHF150.00 47.92 24.15 63.00 29.00

EQUITY MARKETS
CORPORATE & INVESTMENT BANKING / INVESTMENT SOLUTIONS / SPECIALIZED FINANCIAL SERVICES

S EC TOR R EPOR T

Pharmaceuticals

I2

S EC TOR R EPOR T

Contents
Investment summary 1. Valuation: What impact from biosimilars?
Valuing pharmaceuticals: Raptures of the deep European groups: Solid growth capacity

4 6
6 8

2. Biologics: what makes them different?

Highly specific products Players in a highly profitable market

15
15 16

3. Biosimilars: a new breed of drugs!

Quite a difference between "similar" and "same"! Regulations: the US is lagging behind, but is more ambitious than Europe Commercial impact: what can be learned from Europe?

21
21 23 28

4. The biosimilars players

A branded product business model Generic drug producers: few biosimilar candidates An eldorado that is also attracting the pharmaceutical groups

31
31 33 40

5. Big pharma: growth strategies in the face of biosimilars

Roche: growth in spite of everything Insulins: biologics apart What impact for sanofi-aventis and Novo Nordisk?

42
42 46 49

6. Appendix 7. Company profile Novartis

Our top pick

53 58 59
59

Novo Nordisk
Above the fray

63
63

Roche
Return to grace unlikely before 2011

67
67

Stada
The recovery is on track

71
71

Pharmaceuticals

I3

S EC TOR R EPOR T

Investment summary
Pfizer took on the persona of Dr Faustus in October 2010, when it found its Mephistopheles in the form of Indian generic manufacturer Biocon, with which it signed an agreement for the development of biosimilars. What seems at first an unholy alliance bears witness to the great attractiveness of the biosimilars market, into which in addition to large generic makers such as Teva, Sandoz and Stada major pharmaceutical groups (e.g. Merck, Pfizer and Amgen) are rushing enthusiastically. The potential market opening up is enormous. Biologics currently represent a market of $94bn, and a majority of them will have lost their patent protection by 2015. The market for biosimilar products though currently still taking its first hesitant steps - could reach $10bn by 2017. Up to now, it has not been possible for biologics to face generic competition, because of their complexity. But the legislation has changed. In Europe, the marketing of biosimilars (copies of biologics) has been possible since 2006. In the United States, the Obama law will go even further. Not only will the launch of biosimilars be possible, but in some cases (unlike the situation in Europe) they may be substitutable for the original drug, producing a situation similar to that in conventional generics. There will even be the possibility of attacking patents. This should however be much less widespread than for conventional generics. Generally speaking, as in Europe, this will be a separate category of products, with a business model halfway between a generic and a branded product. In Europe, biologics are recognised for their specificity, and biosimilars are not generics. Clinical trials are required before biosimilars can be approved, and they are marketed under a specific name. This means development and marketing costs, and a much less substantial price discount than for a generic. Whereas a conventional generic is launched at a discount of 80% to the price of a branded product, for a biosimilar the discount is around 20% to 30%. There are two consequences. For the generic maker this means the certainty of better profitability and lower profit volatility for biosimilars than for conventional generics. But special production expertise is essential, and a certain level of financial clout. This is why only the heavyweights in the generics sector are present, such as Sandoz and Teva. And for the pharma groups, this means that presence in biologics remains an advantage, as biologics will only experience gradual erosion once their patent has expired, and not the sudden sharp fall of chemical products. Experience of the situation in Europe, where some fifteen biosimilars have been launched since 2006, confirms that biosimilars only very gradually capture market share, limited to a few percentage points, which results in a moderate fall in the original products (-10% to -15%/year). In our view however, this situation will change. Once biosimilars have been widely introduced in all categories, and once prescribers have grown used to them, market share growth rates will be more rapid, and we have based our models on the most pessimistic hypotheses. In our universe of stocks, biosimilars are primarily a major opportunity for Novartis, whose generics subsidiary, Sandoz, is already the leader in this segment with 47% of the market in 2009. It was moreover a pioneer with the marketing of the first Western biosimilar in 2006, Omnitrope, a human growth hormone. It recorded $118m of sales in this field in 2009, and we expect it to reach $600m in 2015. Furthermore, Sandoz has signed agreements with Momenta, which may be in the process of making history in the United States. Thanks to original technology, this company has succeeded in obtaining approval for a substitutable generic for Lovenox, by sanofi-aventis, a drug which is however recognised as a biologic in Europe. If successes of this
Pharmaceuticals I4

S EC TOR R EPOR T

type continue (it is now attacking Copaxone by Teva), this could change our analysis, with the widespread introduction of substitutable generics for biologics. But as things stand today, we consider that this is no more than a long-term risk. In any case, the generics division should be an important growth lever for Novartis, with annual profit growth of 20%/year. We are maintaining our Buy rating for Novartis, with a target price of CHF70. Amongst the big pharmas, the two groups most affected are Roche and Novo, followed to a lesser extent by sanofi-aventis. In our view the disappearance of Roches premium to the sector shows that the market has taken on board a fear about falling sales of major products in the second part of the decade. Roche however has lengthy patent protection, with the bulk of its patents not expiring until 2018 and 2019. This is therefore primarily a problem for the next decade. An actuarial calculation, comparing valuations for products with and without biosimilars, thus shows an impact in the order of 8% of valuation. This means that Roches derating has a structural component. But the group does have growth strategies, despite biosimilars. The first one concerns new subcutaneous forms, which should take the place of some current sales, while in the process extending patent protection until 2024. Roche also has successors, with in particular T-DM1 for Herceptin, which despite delays remains highly promising, and GA 101 for Mabthera. Generally speaking, the groups advanced pipeline remains one of the best in the industry, with a potential of CHF12.7bn, i.e. 33% of pharma sales only counting products in phase III. We are therefore confident, even assuming the arrival of biosimilars, that pharma sales should grow by 5%/year until 2015, and by around 4% a year after that date. However, the market will wait before taking these prospects on board for positive news flow, which means forgetting about recent disappointments (taspoglutide, Avastin). Until then, we are maintaining our Neutral rating. The insulins market will also feel the competition from biosimilars. But two specific features will limit their impact: 1/ high capital intensiveness, and 2/ the crucial importance ofhe injection device. The necessary investments must provide a return, and the price flexibility of biosimilars will be limited. Furthermore, the insulins market is growing rapidly. Even if we assume that biosimilar penetration will ultimately be strong, we consider that Novos insulin franchise could t grow up to 2022, and by 2030 still maintain its 2012 level. The impact will therefore remain circumscribed, and the markets attention will continue to be focused on the very strong short-term momentum, thanks in particular to Victoza. We think that the premium relative to the sector will be maintained, and we are sticking with our Buy rating. Similarly, the sales of Lantus from sanofi-aventis should have no difficulty growing up to 2015, before stabilising with the arrival of biosimilars and the ramp-up of Degludec. The development of the Lantus/GL1 fixed combination may extend the growth of the franchise up to the end of the decade. We are therefore maintaining our Buy rating and our target price of 63.

Pharmaceuticals

I5

S EC TOR R EPOR T

1.

Valuation: biosimilars?

What

impact

from

The performance of European pharmaceuticals was sluggish in 2010, only just following the market trend. This was despite the fact that EPS levels were extremely resilient during the crisis. The market is quite simply terrified by the patent cliff, and despite another demonstration of resilience in 2010, it seems to take apocalyptical prospects on board when it comes to valuation. Against this background, in which aversion to risk tends to become a pathological search for scare stories, the growing confirmation of the imminence of biosimilars is a new spectre on which its fears can be focused. As we demonstrate in this report, the markets fears will once again turn out to be overdone, but they need to be taken into account from a stock market viewpoint, so as to evaluate the prospects of groups such as Roche and Novo Nordisk.

Valuing pharmaceuticals: Raptures of the deep

Chart 1: Historic relative performance of pharma sector
150 140 130 120 110 100 90 80 2007 2008 DJSPHRM/DJSTOXX 2009 PHARMUS/DJINDUS 2010 100

Source: Datastream

Following a very good performance since mid-2008, the sectors performance has proved disappointing in Europe in 2010. Since the start of the year, it has been a market performer, as the sentiment of uncertainty about its short and medium term growth capacity has continued to put pressure on its valuation.
Significant growth differential in relation to the market

The significant upward revision of profit estimates has gone unnoticed. And yet, if the financial performances of pharma in Europe are compared to those of the CAC 40, it clearly emerges that pharma has turned out to be the only sector to be completely immune to the 2009 recession, with no decline in profits. In fact, over the period 2007/2009, the EPS of CAC 40 companies fell by 40%. It will not return to its initial level until 2012. Five years for nothing! Over the same period, pharma has achieved a 40% increase over 5 years, a growth differential which seems set to continue up to 2014.

Pharmaceuticals

I6

S EC TOR R EPOR T

Chart 2: Comparative trends of market and pharma sector EPS (base: 100 in 2007)
180 160 140 120 100 80 60 40 2007 2008 2009 2010 EPS Market 2011 EPS Pharma 2012 2013 2014

Source: Natixis

Valuation ratios falling to all-time lows

This has made the valuation ratios of the sector fall to all-time low levels. P/E has thus been plumbing new depths. It seems that the market is expecting a meltdown in profits on a scale which is impossible in our view.
Chart 3: Historic relative PE of pharmaceutical sector in Europe and US (in x)
2.00 1.80 1.60 1.40 1.20 1.00 0.80 0.60 82 84 86 88 90 92 94 96 98 00 02 04 06 1.37 08

PHARMUS(PE)/TOTMKUS(PE)

PHARMEE(PE)/TOTMKEE(PE)

Source: Datastream

The causes of the worries are well known. Temporarily reassured by a revival in R&D productivity, the market has found two other causes of concern:
Admittedly patent expiry dates are getting closer

The difficult environment for the sector (health care spending restrictions in Europe, fears about price trends in the United States, etc). And above all renewed fears about resilience in the face of attacks from the generic makers. As the patent expirations of 2011/12 come closer, doubts about short-term growth are more keenly felt. But in a longer-term view, anxiety about the arrival of the first biosimilars (generics of biologics) is beginning to be heard. Accordingly, the groups which have previously enjoyed the strongest immunisation against the genetic threat - because of the biologic component of their sales (Roche, Novo Nordisk, etc.) - are coming under the spotlight, and questions are being asked. The aim of this report is to evaluate the threat posed by biosimilars in a long-term perspective to the valuation of the sector, and the valuations of Novo-Nordisk and Roche in particular.
Pharmaceuticals I7

S EC TOR R EPOR T

European groups: Solid growth capacity

But the sector has growth relays which can offset the loss of profits

In the absence of substitutable generics for biologics (in the current state of technology see Momenta below), the sudden arrival of biosimilars represents a new form of competition, and a sharp brake on price flexibility for biologics, but not a complete calling into question of the sector. As shown clearly in chart 2 which compares EPS momentum, the sector still has a solid growth capacity over the next 4 to 5 years. The composition of the portfolio of the major European stocks helps us to understand that there is a bedrock of profit which is unaffected by generic themes of any kind. Thus, exposure to emerging countries is a growth asset for some players. In addition, the industry to innovation power is appearing more clearly in the decisions taken by the FDA. This will provide real support for the top line, and will be a genuine catalyst for valuations in the sector to move upwards. Finally, our estimates are made on a constant structural basis, but this does not allow for the mechanical accretive effect on EPS in the event of major acquisitions financed by debt. 67% of European laboratories portfolios are not genericable Biosimilars will lead to a redrawing of the analysis criteria for pharmaceutical companies. This is why we have defined three categories of activity. The genericable part. We think it is prudent to think ahead, and we consider that all chemical products, whatever the length of their patent protection, will be affected by faster erosion in a long-term perspective. We are including those biologics which are set to rapidly come up against widespread competition from biosimilars. Sales of strong barrier to entry products, for which any rapid erosion is ruled out, whatever happens (insulins, etc.). Non-genericable activities, including vaccines, OTC and emerging countries. We have calculated that actual expiration represents on average 6% of OP per year for the sector in the current patent cliff period. This will easily be absorbed by the growth in activities which are not sensitive to patent issues. The proportion of growing non-cyclical activities is rapidly increasing for all groups, which are broadly following the same strategy, except for those whose portfolio already meets these criteria, i.e. Novo and Roche.
Table 1: Structure of sales: Two-thirds are not genericable
As % Genericable part (including simple biologics: EPO, interferons) 70 Entry barrier (including respiratory, monoclonal antibodies, insulins) 14% (respiratory excl. emerging) Advair+ Tykerb, Arzerra 5465 19% 9% (biologics+Alcon) 70% 42% (strictly monoclonal antibodies Europe/US) 13% (insulins) 28 Definitely not genericable (emerging, vaccines, OTC) 16% (o/w emerging 13%) Vaccines + emerging excl. vaccines+OTC 13,106m 45% 49% (consumer+ generics+ vaccines+emerging) Emerging 23% 22% 48% (vaccines+ emerging excl. vaccines +OTC + generics excl. emerging) 39

AstraZeneca Bayer GSK Novartis Novo Nordisk Roche Sanofi-aventis Sector average Source: Natixis

36 42 7 36 39 33

Pharmaceuticals

I8

S EC TOR R EPOR T

The proportion of activities in which rapid growth is set to continue either because generic competition is impossible, or because it is improbable and has minor impact represents on average 67% of the sales of the groups. This proportion will continue to increase following recent acquisitions: Alcon for Novartis, and soon Genzyme for sanofi-aventis. Accordingly, it is hard to imagine how any of the groups could see their profits decline. Two companies stand out in this classification: Novo Nordisk, which has a very low genericable proportion, and at the other end of the spectrum AstraZeneca, which has to cope with the fact that most of its products are genericable. Note the case of GlaxoSmithKline, for which the Advair question could change the situation totally, if Sandoz managed to obtain authorisation for a substitutable generic.
And 67% of 2010 sales are not genericable, or only with difficulty

Even after allowing for this pessimistic vision of the fate of biologics, and assuming a rapid fall in the genericable proportion as a whole (-20%/year), leading to its virtual disappearance in the space of 7 years, and without taking into account the impact of external growth, this sales structure does not suggest that there would be a fall in profits for the sector. On the other hand, it is set to go through a period of two years of near-stability, which corresponds in fact to the patent cliff which has triggered so many fears.
Chart 4: Growth despite biosimilars (simulation of sales trend based on previous table)
160 140 120 100 80 60 40 20 0 2010 2011 2012 2013 2014 2015 Grow th 2016 Barrier 2017 2018 2019

Generic possible
Source: Natixis

Biosimilars: Legitimate concerns

The markets concern about biosimilars is legitimate, for many questions have still not really been answered. To what extent will the FDA authorise biosimilars and on the basis of which clinical trials? Will biosimilars be interchangeable? Which types of products? Will sudden attacks on patented products become widespread, as in the case of conventional chemical drugs (ABLAs)?
Law on biosimilars is increasing investor concerns

In this report we attempt to provide some answers to these questions. The passing of the Obama law (March 2010), and above all the authorisation of a substitutable generic for Lovenox (July 2010), mark a historical turning point for the industry. The substitutable generic authorisation amounts in fact to recognition of the technological ability, in this case of Momenta, to analyse the structure of a complex product so as to reproduce it in a way which is
Pharmaceuticals I9

S EC TOR R EPOR T

considered to be perfect in the eyes of the FDA. This seems to us to open the door to the possibility of substitutable copies for certain products, and this is probably what is terrifying the market. We do indeed think that the special nature of many biologics will be attenuated, and that the corresponding stock market premium will therefore quite legitimately decline. This only applies however to the products which are easiest to reproduce.
The law will apply particularly initially - to those biologics which are easy to reproduce

Two categories of products seem in our view destined to escape any rapid erosion by generics, not only short-term but also long-term: Monoclonal antibodies, whose fabrication is too complex for a substitutable generic to be feasible in the near future, and for which any future biosimilar must present more significant clinical trials, as confirmed by the recent European guidelines. Insulin analogues. The Pfizer-Biocon agreement means there is no doubt that biosimilars for insulins will ultimately become widespread. But impact will remain limited: in addition to the inevitable clinical trials, capital intensity will limit price discounts. Note that an insulin unit on the scale required for global markets requires an investment of $50m to $100m, and that the administration system which is essential is highly complex (it took sanofi-aventis 10 years to develop a competitive pen device). Impact on valuation: The example of Roche The Roche share price trend, in particular, shows in our view that the market is now bundling together high barrier products and conventional chemical products. Roche which makes two-thirds of its sales with biologics has thus seen its valuation premium relative to the sector fall away, which is probably excessive. This is due in large part to the confusion sown by the imprecision of the new US law, and the announcements of groups which are now preparing biosimilars for monoclonal antibodies.

It has led to a reduction in Roches stock market premium

We have attempted to evaluate the real impact of the arrival of biosimilars on the main company concerned, i.e. Roche. It is 5 times less serious in profit terms than in the case of a conventional generic. With a conventional generic, we estimate that the profit for the originator group is reduced by 90%. In the case of a biosimilar, on our estimates, profit is simply divided by 2. We have carried out simulations of the impact on valuation of the arrival of a biosimilar (see the case of Herceptin in the annex), which is based on the following hypotheses: An impact on Europe and the United States (excluding the emerging countries). A 25%/year reduction in sales of the product after the arrival of the biosimilars. A margin which is gradually reduced by 10 percentage points after the arrival of the biosimilar (based on a margin which in any case has been reduced by the decline in sales excluding biosimilars). The calculation does not take into account defence strategies, and particularly the launch of subcutaneous forms, which will enable the recovery of a large proportion of the value.

Pharmaceuticals

I 10

S EC TOR R EPOR T

Chart 5: Impact of biosimilars on valuation of Herceptin (CHFm)

The impact of the arrival of biosimilars for Roches major biologics

16 000 14 000 12 000 10 000 8 000 6 000 4 000 2 000 0 Without biosimilars
Source: Natixis

Biosimilars impact

With biosimilars

If major product values are added together, this indicates an overall impact on valuation.
Table 2: Impact of arrival of biosimilars on valuation of major Roche products (CHFm) Product Avastin Mabthera Herceptin Lucentis Others Total
Source: Natixis

Value without biosimilars 25 851 13 513 14 556 5 546 6 250

Value with biosimilars 22 462 11 767 11 611 4 783 4 300

Difference 3 389 1 746 2 945 813 1 950 10 843

could lead to a theoretical 8% reduction in the value of the groups capitalisation

We arrive at a valuation difference of CHF12.6 /share, which represents 8.4% of the target price. As this calculation makes no allowance for new formulations, the real impact is lower. It does however justify a decline in the premium relative to the sector. However the share price is at a 2012 P/E discount of 5% to the sector, which corresponds to a fall of 25% relative to the historic premium. Clearly the share is undervalued, even after allowance is made for biosimilars. But the short term momentum has been affected by R&D setbacks. Innovation: THE differentiating factor The recent decisions taken by the FDA and EMEA and their experts highlight the increase in the productivity of the pipelines of the major groups, which slumped from 2003 to 2009. But genuine efforts have been made to adapt to meet the regulators requirements, and have been based on new development methods combining 1/ the selection of patients chosen for clinical trials on the basis of the expression of genetic anomalies, 2/ research into outcome data (improvement in patient survival for an anti-cancer drug, reduction in the risk of vascular incidents for a molecule in cardiology) and no longer just into an effect on a bio-marker, and 3/ the setting up of secondary effect surveillance plans. Lastly, the prioritisation of certain programmes has been carried out alongside the discontinuation of earlier developments, thus incidentally enabling the R&D/sales ratio to be maintained at a stable level over the last three years. The product news flow is now clearly more encouraging. And 2010 marks a turning point. Approvals and favourable recommendations by experts have come in quick succession. Mention should be
Pharmaceuticals I 11

S EC TOR R EPOR T

made of Brilinta (arterial thrombosis - AZn), Gilenya (MS, Novartis), Pradaxa (thrombosis, Boehringer Ingelheim) and more recently Benlysta (lupus, GSK). Better defined, the trials have also turned out to be positive: Xarelto (thrombosis, Bayer), teriflunomide (MS, sanofi-aventis). This trend should continue in 2011 with the eagerly awaited phase III trials of iniparib (breast cancer, sanofiaventis), Degludec (long-acting insulin, Novo Nordisk) and dalcetrapib (dyslipidemia, Roche).
The real response to this new threat is the groups ability to innovate

Certainly not everything in the garden is rosy, as shown by the negative decisions by FDA experts about Avastin (Roche) in breast cancer (even though it had already been marketed in the United States) and the suspension of major developments (taspoglutide from Roche/Ipsen), but these are signals of pipeline regeneration which the market is beginning to value, with the first prescriptions / sales, and which explain the difference in valuation between Novartis and its European counterparts.
Table 3: Valuation of Pharma sector 3/12/2010)
x 2011e AstraZeneca Bayer GlaxoSmithKline Ipsen Merck KGaA Novartis Novo Nordisk Roche Sanofi-aventis UCB European average Abbott Bristol-Myers Squibb Eli Lilly and Company Johnson & Johnson Merck & Co Pfizer US average World average Source: Natixis estimates 7.6 12.1 9.8 13.4 8.7 9.7 20.2 9.4 7.0 14.2 10.2 10.2 11.5 7.7 12.5 9.3 7.3 10.0 10.1 PE 2012e 8.5 11.0 9.5 11.7 8.2 9.2 17.7 8.7 7.2 12.7 9.7 9.5 13.0 9.0 11.7 8.8 7.5 9.8 9.7 2013e 8.3 10.3 8.9 9.0 7.4 8.4 15.2 7.5 7.2 11.2 8.9 8.5 13.8 9.0 10.7 8.7 7.2 9.3 9.1 2011e 4.3 6.5 6.8 7.4 6.5 7.7 10.9 6.7 4.8 9.5 6.7 7.0 5.9 5.0 7.3 6.3 4.2 6.1 6.4 EV/EBITDA 2012e 4.5 5.8 6.4 6.1 5.9 7.1 9.5 6.0 4.6 8.8 6.2 5.9 7.1 5.4 6.5 5.4 4.0 5.6 5.9 2013e 4.2 5.1 6.0 5.2 5.2 6.7 8.0 5.4 nm 7.7 5.1 5.2 7.7 5.8 5.6 4.9 4.0 5.8 5.4 2011e 1.9 1.5 2.4 1.4 1.6 2.4 4.0 2.6 2.0 2.0 2.3 2.1 2.1 1.6 2.4 2.6 2.0 2.2 2.3 EV/Sales 2012e 1.9 1.4 2.2 1.2 1.5 2.2 3.5 2.4 1.8 1.9 2.1 1.8 2.4 1.7 2.1 2.3 2.0 2.1 2.1 2013e 1.8 1.3 2.1 1.1 1.4 2.1 3.1 2.1 nm 1.7 1.8 1.7 2.7 1.9 1.9 2.3 2.2 2.1 1.9 2011e 4.7 10.5 7.2 9.8 10.5 8.0 12.3 7.7 5.3 13.4 7.8 8.9 6.4 5.8 8.6 7.8 5.0 7.3 7.6 EV/EBIT 2012e 2013e 5.0 9.0 6.8 8.0 9.4 7.6 10.8 6.8 5.0 12.4 7.2 7.5 8.4 6.7 7.5 6.7 4.8 6.7 7.0 4.7 8.0 6.4 6.6 8.4 7.2 9.1 6.1 nm 10.8 6.0 6.4 10.9 7.5 6.4 6.2 5.1 6.5 6.2

Our favourite stocks The strong share price momentum of Novo-Nordisk (Rating: Buy) is set to continue. The recent Q3 10 publication reflects the already large contribution of Victoza (Q3 sales of DK10,700bn). This rapid growth is set to continue in the short term, not least with the launch in 2012 of its long-acting insulin Degludec, currently in phase III, which we see offering peak sales of over DKK15bn in the long term. We value Novo-Nordisk at DKK600 per share based on a DCF analysis.
It is this ability to innovate which means we are maintaining our Buy rating for Novo and Novartis

Thanks to its diversification business model, and its better-than-average R&D productivity and topline dynamism, Novartis (Buy, target price CHF70) still has a very attractive profile. Faced with a more difficult context, with upcoming patent expirations in 2011/12 (29% of pharma sales), Novartis has three responses: 1/ diversification, particularly into generics and OTC, enabling it to tap into new markets (biosimilars) and increase its bargaining power with payers, 2/ its ability to innovate, highlighted on many occasions in 2010, and 3/ from now on, an optimisation of its asset base, by better use of production capacities, via a reduction in the number of sites (86 sites in all), and a
Pharmaceuticals I 12

S EC TOR R EPOR T

lowering of the WCR. In results terms, this should lead to the cost of sales holding steady, or even falling, and an increase in ROCE (22% in 2010 excluding Alcon) significantly higher than the cost of capital (9%). More specifically, the generics division remains one of the cornerstones of the groups strategy, aimed at decreasing the risk inherent in traditional pharmaceuticals. The generics business involves little development risk, and by definition no patent risk. Furthermore, the increase in the size of Sandoz has reduced the volatility of its results. The valuation of the generics leaders has thus increased significantly, and is becoming higher than that of conventional pharma players. Based on a 2012e EV/EBIT ratio of 8.5x, comparable with that of Teva, Sandoz could represent an EV of $17.5bn (out of a total of $178bn for Novartis, i.e. 11%). Novartis remains our favourite European stock.
Table 4: Breakdown of restated value of Novartis $m Novartis Pharma Sandoz Consumer Health Alcon (100%) Vaccines (Chiron) Corporate Charges Total Enterprise Value + Financial assets - Roche - Others - Minorities (Alcon) - Net debt Restated value ($m) Restated value per share (CHF)
Source: Natixis

12 Sales 59,528 32,650 10,100 6,604 7,674 2,500 59,528

12 EBIT 16,405 9,870 2,050 1,375 3,200 635 -725 16,405

Multiples 12 EV/sales 3.0 2.7 2.0 2.5 6.5 4.0 3.0

Multiples 12 EV/EBIT 10.8 9.5 8.5 10.0 15.6 15.0 10.5 10.8

Enterprise value EV/sales 178,293 88,808 20,503 16,510 50,064 10,000 -7,592 178,293 11,381 7,381 4,000 11,055 7,649 170,971 76.2 EV/EBIT 177,414 94,160 17,507 13,750 50,064 9,525 -7,592 177,414 11,381 7,381 4,000 11,055 7,649 170,092 75.8

Sanofi-aventis remains substantially undervalued, as the market is anticipating a doomsday scenario which will not happen

Since July 2010, sanofi-aventis (Buy, target price 63) has been suffering from the uncertainty surrounding the Genzyme acquisition. The markets usual touchy reaction about the buyers share price has been made worse by the stubbornness of the Genzyme management which refuses to open the door for negotiations, which suggests to us that taking a controlling interest will be a long process. In view of this, we are more cautious about the stock. The associated value creation cannot be fully reflected in the share price until completion of the deal is in sight. The value creation should easily reach 4 to 5 per share, depending on the level to which the bid is raised. Looking beyond the value creation linked to this operation, the group remains substantially undervalued in its current structure. The doomsday scenario the market is expecting will not happen (at most, a 2% decline in EPS is expected in 2012). In fact, the bedrock of growing non-genericable businesses, and the contribution of new products (Multaq, Jevtana), offset the loss of the profits of Plavix, Eloxatin, Taxotere and Lovenox-US (we estimate the impact at 11% a year from 2010 to 2012). However, we are continuing to adopt a more cautious stance on Roche (Neutral, target price CHF150). The group has admittedly undergone a major derating, which has wiped off all of its historic premium. There are several explanations for this: 1/ worries about short-term growth (disappointing quarterly figures) and particularly medium-term growth (are biosimilars a danger for the future of Herceptin, Mabthera and Avastin?), 2/ questions about the fate of Avastin in breast cancer (FDA decision on 17 December), and the fate of taspoglutide. The announcement of the first restructuring plan of the decade (savings estimated at CHF1.8bn after tax, i.e. about 11% of the
Pharmaceuticals I 13

The market is likely to hold back on Roche until the Avastin decision is taken. Neutral rating maintained

S EC TOR R EPOR T

groups adjusted net profit in 2013) has not overcome this scepticism. Pharma stocks do not react to this theme. The main requirement for a return to grace would be clear positive news flow, together with good organic growth figures. It seems that this is not likely to happen for several months. We are maintaining our Neutral recommendation.
We are confident about Stadas ability to recover, and are maintaining our Buy rating

As for Stada, we consider that after experiencing serious problems (particularly the insolvency of its distributors in Serbia and the negative impact of health reforms in Germany) the group is in an upward phase. We are therefore maintaining our Buy rating, with a target price of 29. Stada published encouraging Q3 10 results, bolstered by the Branded Products division (+12%) and growth in Russia (+13%), and has confirmed its 2014 guidance figures. We therefore think the worst is behind us (provisions of 29m in 2010): the share should outperform the sector, and its discount of 30% (2012e P/E of 9.5x vs. 13.3x for all generic manufacturers combined) should decline.

Pharmaceuticals

I 14

S EC TOR R EPOR T

2.

Biologics: what makes them different?

In this report, we analyse the impact of the emergence of generic copies, or biosimilars, of drugs known as biologics. This section looks at the key role played by this category of products for the sector. A biological drug is one whose active matter is a complex molecule on a large scale, produced not through chemical synthesis, but by biotechnological means (genetic engineering). This makes them much more difficult to reproduce than conventional chemical molecules, and has seen the emergence of a specific regulatory framework in the US (Biological Licence Application, since 1996). The launch of revolutionary biological therapies in a range of therapeutic fields (cancer, arthritis, MS, anaemia, etc.) saw the spectacular rise of this market from the 1990s onwards, giving a major boost to groups such as Amgen, Novo Nordisk and later Roche. Biologics generated 2010 sales of around $110bn, i.e. some 15% of the worldwide pharmaceuticals market, and boast above-average margins for the industry.

Highly specific products

The definition of biological medicines is not that clearly defined, but it mainly includes therapeutic proteins and, in the broad sense, all products produced using biotechnological means, i.e. with living organisms (bacteria, or genetically modified cells). They are proteins that have been modified to a greater or lesser extent and which differ from small chemical molecules obtained through synthesis, above all in terms of their complexity. This complexity relates to the size of the compounds, which can be several hundred times that of small conventional molecules, and to how they are organised spatially. Biological drugs are huge structures compared to the handful of building blocks that make up conventional small molecules The following illustration gives a comparison of the actual size of molecules.
Chart 6: Biologics: a different order of complexity

*Da: this measure indicates the weight of the molecule Source: Hospira

Pharmaceuticals

I 15

S EC TOR R EPOR T

Biological products: highly complex

This complexity is the reason behind the production method. In many cases direct synthesis is not possible. These molecules are therefore obtained through genetic engineering. This involves taking bacteria whose DNA has been specially modified, cultivating them and having them express the product in fermenters. The first medicine to be produced using this process was human insulin, made by Eli Lilly, in 1984. The products can be hard to characterise with any precision. The spatial distribution of the atoms is as significant as the composition itself and the slightest modification to the molecules structure can impact on their efficacy. In particular, proteins tend to develop chains of sugar on their surface (glycosylation), which can vary depending on operating conditions during production. A major consequence of these characteristics is that a minor change in production procedures can lead to significant differences in efficacy or side effects.

A defining feature of biological products is their highly specific production processes

This sensitivity to manufacturing procedures and the types of cells used is at the root of the regulatory difference between biologics and small molecules. Indeed, it is not possible to ensure an exact copy of these molecules, short of applying the same procedures as the original maker, which only it knows, down to the last detail. Hence the idea that process and product are intrinsically linked with biologics. Because of the specific nature of these products, regulatory authorities have created specific approval frameworks for biologics. In 1996, the FDA introduced the biological licence application (BLA), which differs from that for conventional chemical entities (New Drug Application). Under its terms, no generic can be authorised for biologics registered in compliance with this procedure.

Biologics are governed by a distinct regulatory framework

This regulatory specificity, coupled with the fact that generic copies are impossible to make, produced an Eldorado for the industry, all but free of any generics threat. This is one of the reasons for the significant historical premium awarded to companies like Roche and Novo, which are big players in the biologics field. But this situation is changing.

Players in a highly profitable market

A market worth over $94bn in 2009. Biological products fall into very disparate categories. The main ones are recombinant proteins and monoclonal antibodies.

Pharmaceuticals

I 16

S EC TOR R EPOR T

Table 5: Key categories of biological products Category Insulins 2009 world sales ($m) Examples 14 512 Lantus (sanofi-aventis) Humalog (Lilly) Novolog (Novo Nordisk) Levemir (Novo Nordisk) 1 902 Pegasys PEG-Intron 9 532 Aranesp (Amgen) Epogen (Amgen) 3 137 Genotropin (Pfizer) Humatrope (Lilly) 35 613 Mabthera (Roche) Herceptin (Roche) 6 251 Avonex (Biogen IDEC) Rebif (Merck KgaA) 5 738 Lovenox Fraxiparine 5 163 Neulasta (Amgen) Neupogen (Amgen) 6 272 Kogenate (Bayer Schering) NovoSeven (Novo Nordisk)

Alpha interferons EPO

Sales of $94bn in 2009

Growth hormones Monoclonal antibodies Beta interferons Heparins G-CSF Coagulation factors
Sources: Evaluate, Natixis

The main categories are as follows:

derived mainly from recominant proteins

Recombinant proteins (world market of over $47bn in 2009): this segment includes various product categories, from EPO (anaemia), to G-CSF (neutropenia), to insulins, growth hormones, coagulation factors, alpha interferons (anti-viral therapies) and beta interferons (MS therapies). These products are now well characterised and offer a relatively simple molecular structure. Monoclonal antibodies ($35bn market in 2009): These products, mainly indicated to treat cancer and autoimmune diseases (RA, MS), are much more complex in their molecular structure, are much bigger than an EPO and are more difficult to reproduce. Because they are more recent, they still boast extensive patent protection. They also remain the preserve of just a few groups, Roche being the unrivalled leader. The emergence of biosimilars is unlikely to affect this category in the short term. Vaccines ($21bn market): These are biological products par excellence, manufactured from genetic material (eggs, cell lines) with limited stability and subject to contamination risk. They tend to be complex and expensive to produce, giving generics manufacturers extra technological and industrial hurdles to overcome. Enzymes ($2bn market in 2009). This is a very specific category but also includes proteins, which are authorised under the BLA procedure. It therefore comes under the definition of biologics. Drugs of animal extraction: Strictly speaking, these are not biological products, i.e. manufactured using genetic engineering or in-vitro culture, but there are similarities in terms of the complexity of their molecular composition which is not always well characterised. Among the products in this category are treatments such as Wyeths Premarin (oestrogen) and low molecular weight heparins (worldwide market of $5.7bn) such as sanofi-aventis Lovenox. The problem here is the precise characterisation of the product, which can be a barrier to generics.
Pharmaceuticals I 17

and monoclonal antibodies

S EC TOR R EPOR T

For instance, Premarin was the subject of a decision by the FDA which blocked generics, based on the fact that its structure was not fully understood. But the treatment of these drugs clearly illustrates that they are at the border line of the biologics market: for Lovenox, for example, the European authorities considered it as a biologic and requested clinical studies. By contrast, the US authorities authorised a substitutable generic without clinical trials. These products will very likely cease to be regarded as biologics in the future.
Some products (peptides, heparins) have similar characteristics to biologics

Peptides, like Tevas Copaxone (MS), or GLP1 (diabetes, AstraZeneca, Novo), are not traditional biologics either, as they are not manufactured using genetic engineering, but by means of chemical synthesis. These are chains of aminoacids, which are less complex than proteins and can be synthesised directly. As such, they do not come under the scope of the definition described above. That said, they are complex to manufacture and synthesising them requires specific expertise. In the same way as biologics, they represent significant technological barriers to entry. very profitable Biological products are more profitable than the sector average (25-30%). The difference is even more pronounced for monoclonal antibodies.

Much higher margins than with chemical products

Their operating margin can often exceed 40%, by our estimates, thanks to three factors: In most cases, the drugs are part of a recent generation of treatments and therefore benefit from innovation effects. The therapeutic benefit they confer is often more clearly established than with conventional products, which means they can command very high prices. Treatment costs for latest-generation cancer therapies can exceed $40,000/year. The fact they have been on the market for a mere ten years or so means there is an oligopoly in effect in the majority of cases. A key reason is that the products are by definition injectable, due to the difficulty of developing an oral formulation for these large molecules. Selling expenses are often low because most of the products are for hospital use. Hospital sales forces tend in most cases to be limited to a few hundred reps, as opposed to several thousand for primary care. But it is clear that these factors will not remain static. The oligopoly situation will ease as new products/categories come to market. Similarly, cuts in health-care spending are likely to hit these drugs particularly hard. As for price sensitivity, this is unlikely to be called seriously into question until biosimilars are a common feature in these categories. The slump in sales will therefore go hand in hand with a deterioration in profitability.
Table 6: Operating margins of biotechnology groups En %

Groups with a big Roche pharma biologics presence boast a margin of Amgen 38%, i.e. significantly higher than their Novo Nordisk traditional pharma counterparts

Biogen Idec

09 sales Top-selling products (09) 38 996 MCHF Mabthera/Rituxan (CHF6.087bn) Herceptin (CHF5.266bn) 14 642 M$ Neulasta ($3.355bn) Enbrel ($3.493bn) 51 078 MDKK NovoRapid (DKK9.749bn) NovoSeven (DKK7.072bn) 4 377 M$ Avonex ($2.323bn)

2007 35 36 21 27 30 28

2008 36 37 27 30 33 29

2009 36 38 29 30 33 31

2010e 39 41 32 38 38 31

Average of 4 biotech. Average of top 5 pharma groups

Sources: Evaluate, Natixis Securities, Datastream

Pharmaceuticals

I 18

S EC TOR R EPOR T

A presence in the biologics field therefore remains a key advantage. The main players are shown in the following chart.
European companies most active are Roche and Novo

Chart 7: Pharma groups have varying degree of involvement in biologics (biologics contribution as % of 2009 revenues)
100% 80% 60% 40% 20% 0% Amgen Nov o Nordisk Genzy me Roche sanofi-av entis GSK Nov artis

Source: Natixis

but most biological products are set to come off patent by the decades end By our estimates, around $24bn worth of sales will lose patent protection by the end of 2015.
Table 7: Key patent expriries (2010e sales)
Year Product (indication) Company Bayer Amgen Lilly Teva Genzyme Novo Nordisk Sanofi-aventis Amgen Merck Kg/BMS Roche Roche Roche Amgen Amgen World sales ($m) 1 590 3 500 2 075 3 127 725 2 057 3 466 3 500 2 608 1 783 6 550 5 620 6 430 2 441 2 481 US sales ($m) 587 3 273 1 287 2 176 364 823 2 586 2 874 2 608 666 3 244 1 674 3 044 1 060 1 473 Europe US patent sales 587 3 273 2012 1 287 2013 2 176 2014 364 2014 823 2014 (2017 formulation) 2 586 2015 2 874 2015 2 608 2015 666 2018 3 244 2019 1 674 2019 3 044 2019 1 060 2024 1 473 2026 2012 2014 2014 2015 2014 Expired 2010 2015 Europe patent

Expired Betaseron (Arthritis) 2012 2013 2014 Biologics: patent expiries getting closer Enbrel (Arthritis) Humalog (Insulin) Copaxone (MS) Cerezyme Novolog Neulasta 2015 2018 2019 2024 2026 Lantus Epogen Erbitux Avastin (CHFm) Herceptin (CHFm) Mabthera (CHFm) Aranesp Avonex

Sources: Natixis, company presentations, Evaluate. Expiry dates can vary depending on the source.

For patent expiry dates, more caution needs to be exercised for biologics than for traditional products. The drugs authorised under the BLA procedure do not appear in the Orange Book, and not all patents are necessarily public. This is important to bear in mind given the key nature of production or formulation method patents. Expiry dates also vary depending on the source. According to Roche, Mabthera is protected until 2019 in the US, whereas Hospira hopes to be able to launch a generic form from 2015.
Pharmaceuticals I 19

S EC TOR R EPOR T

Around $40bn in sales affected by expiries out to 2019

In all events, it is estimated that these expiries will affect around $40bn in sales between now and the end of the decade, with an initial wave in 2014/15. The issue of the emergence of biosimilar rivals and their impact on sales are crucial factors for companies like Roche or Novo Nordisk in Europe and for Amgen in the US. Depending on the assumptions used for the arrival of biosimilars, their long-term prospects are radically different. But this is not a problem in the short term. Looking out to 2015, the market for biological products is still set to average annual growth of +5.8%.

Pharmaceuticals

I 20

S EC TOR R EPOR T

3.

Biosimilars: a new breed of drugs!

The information on the previous page about the inevitable loss of patent protection for most of the biological products is an indication of the biosimilar market's potential. However, there is more to consider than this quantitative aspect. The world of biosimilars will involve a very different business model from that of conventional generic drugs. It will consist of value-added products that will be much more lucrative and less cyclical. Biosimilars also offer a major opportunity for existing generics groups, and even for the pharma groups, some of which are already preparing their offensive in this promising sector. It is therefore as much of a new growth component for some groups as a threat for others. Clearly, companies such as Roche and Amgen have to adapt their strategy to take account of this new situation. We will now take a look at the specific characteristics of biosimilars and their financial implications, then examine the groups strategies.

Quite a difference between "similar" and "same"!

Biosimilar does not mean identical

The characteristics of biological products described earlier in this report mean that even a very minor change in the production process can lead to significant differences in the products biological activity, that is to say its efficacy and, most importantly, side effects. One example of this is the EPO product, Eprex. This J&J product had been available on the European market for years without any problems, when very serious cases of side effects emerged. The group had to temporarily withdraw the product from the market, then began an investigation into the source of the problem. After months of research, it emerged that a minute change in the production process involving the coating on a stopper, had done very serious damage. This idea prompted the first public hearings on biosimilars on both sides of the Atlantic, and led to the current regulations. The idea of sufficient similarity lies at the core of the thinking that has been done about biosimilars, and has been the subject of many debates, which are still continuing even though they have led to legislation on both sides of the Atlantic. There are two factions: the generic manufacturers, which believe that the differences between the original product and its copy are small enough to consider them interchangeable, without any further precautions being taken. The pharma industry protested against this approach as soon as the FDA public hearings began in 2006, and in the end broadly won its case in Europe and the US. In the vast majority of cases, approval of biosimilars will not be possible until clinical trials are undertaken, which firmly establishes them as inexact replicas.

Pharmaceuticals

I 21

S EC TOR R EPOR T

Similar, but different!

Inexact replicas: the example of Rituxan in India

Biosimilars are different from the original drugs. Analysis of the first biosimilars that hit the market, particularly in India, has supplied ammunition for those arguing for non-substitutability. Roche has made known its views on the Mabthera biosimilar launched by Dr Reddys in India using the brand name Reditux (in reference to the US name for the drug, Rituxan). Roche highlighted the many differences (much higher level of remaining host cell proteins, differences in glycosylation etc.). Chromatographic analysis shows the visible differences in composition.
Chart 8: Chromatographic differences between a monoclonal antibody and its biosimilar

Source: Roche

The stumbling block: immunogenicity

Clinical trials will be required

The main issue that this raises relates to immunogenicity. This is the products ability to provoke an immune response. For many peptides and proteins, the human body makes antibodies, i.e. it recognises them as foreign antigens. As biosimilar copies are not exact replicas, they will be recognised differently by the body, and it will produce specific antibodies. If the difference in composition is significant, this immune response can become a cause for concern. The main concern is the relative lack of knowledge about biosimilars ability to provoke an immune response which is a clinical problem. The different production of antibodies by the body can result in differences on the pharmacokinetic, pharmacological and safety front. Elimination of the product can also be affected. The consequences are unpredictable they could be benign or they could be fatal. The Eprex case (see above) is one example, but there are also concerns about possible anaphylactic shock when on an I.V., or cases of delayed hypersensitivity.

Pharmaceuticals

I 22

S EC TOR R EPOR T

and they will especially assess the immunogenicity of the copies of the biological products

An immune response can vary for many reasons: Firstly, the structure of the protein, whether this is due to a variable sequence of amino acids or a simple difference in structure, caused by a change in the physical, chemical or enzymatic treatment during the production or storage process. Differences in glycosylation in particular lead to differences in structure. It is therefore clear that a different production process can lead to a difference in immunogenicity. The formulation can have an effect, or even the packaging materials, or the conditions of use during a perfusion. The impurities themselves can influence the immune response. In some cases, these impurities that result from the original manufacturers precise production process can act as an adjuvant and affect the drugs efficacy. Their replacement by other impurities may not have the same effect. Particularly if the impurities contain remaining host cell proteins used in the manufacturing process. Problem: an immune response in humans can be difficult to predict using animal models. This is the basis of the argument for the necessity of comparative clinical trials in the development of biosimilars, and not just bioequivalence data. The regulatory authorities have made a lot of progress in a few years. The situation is now clarified in Europe, and the legislation is moving forward quickly in the US.

Regulations: the US is lagging behind, but is more ambitious than Europe

Europe has been a pioneer in the biosimilar field, with very innovative regulations introduced from 2006 and the market approval of around fifteen biosimilars since then. Biosimilars progress in Europe provides a textbook case from which we can extrapolate their future. But US regulations, which are still being worked out, are expected to be even bolder, authorising substitutable generics in certain cases, which is not possible in Europe! Europe: regulations are now firmly in place
The European example: precise guidelines for each class of drug

Biosimilars are already a reality in Europe, and between 2006 and mid-2010 14 biosimilar products were launched. The European Commission initiated discussions on the subject in 2005, and has drawn up legislation on the major issues. It has applied one key principle: to consider the biosimilars as new products. They are treated as new products and have to undergo comparative clinical trials to prove that there is no clinically significant difference between the copy and the original, in terms of efficacy and side effects. Established as non-substitutable, they are launched under different brand names from the original product. Guidelines have been laid down for all the major product categories, and biosimilar competition has begun to grow. Overarching guidelines were established in 2005: biosimilars efficacy and tolerance must be guaranteed by pharmacokinetic and pharmacodynamic data, clinical data and pharmacovigilance data (including a year of immunogenicity data) and checks on the biosimilars manufacturing production process.

including clinical trials carried out on several hundred patients

To obtain marketing approval, a biosimilar manufacturer has to provide studies demonstrating that its product is equivalent in terms of quality, safety and efficacy to the reference biological product
Pharmaceuticals I 23

S EC TOR R EPOR T

that is already on the market. The active substance must be similar in molecular terms (same molecular family) and in biological terms (same effects on the body). The presentation, dosage and route of administration must also be identical, otherwise additional data must be supplied in agreement with the EMEA. Obviously, for the sake of consistency, the same reference compound is used in all the studies. These overarching guidelines are broken down into product type, starting with the growth hormones. In the appendices, we have detailed the studies required for the major categories. The latest ones, published in November 2010, set the frame for monoclonal antibody biosimilars.
Biosimilars are therefore considered to be new products

The EMEA guidelines make a distinction between biosimilars that are a protein, an immunological product (vaccines or allergens) or a blood-derived product (coagulation factors, antithrombin, immunoglobulin, etc.). For the blood-derived products, Abbreviated New Drug Applications which only include similarity trials, will not be accepted. The biosimilars will have to have a new product application, with all the upstream studies that that encompasses, whether they are pre-clinical (pharmacokinetic, pharmacodynamic, efficacy, safety), clinical (pharmacology/toxicology) or involve the safety of the production process. This is especially important for Novo Nordisks NovoSeven, which loses patent protection in 2011, but this is much less of a concern than the biosimilar issues regarding insulins. All in all, the guidelines in Europe have resulted in a significant number of patients in clinical trials, as shown by the table below.
Table 8: Number of patients in clinical trials for marketed biosimilars
Biosimilar (category) Retacrit (EPO) Binocrit (EPO) Onmitrope (human growth hormone) Valtropin (human growth hormone) Nivestim (GCSF) Zarzio Ratiograstim Source: Hospira Company Hospira Sandoz Sandoz Biopartners Hospira Sandoz Ratiopharm Number of indications studied 3 2 1 1 1 1 2 Number of patients in the clinical trials 1 978 826 242 164 371 316 877

14 biosimilars are currently marketed in Europe

The picture varies greatly, from a quite similar number of patients to those for the original product in the case of Retacrit, to a few hundred. The cost of these trials, not counting the failures, will therefore range roughly from $80m to $150m. The US: Pandoras Box has been opened The situation in the US is currently very different from that in Europe. Remember that up until now, marketing copies of biological drugs was simply impossible. It has only since Obamas healthcare reform legislation was passed in March 2010 that it has been made possible via the Affordable Care Act. In return for a 12 year marketing exclusivity period granted to all new biological drugs, the law authorises the approval of biosimilars, in a section covering only 46 out of the 2000 or so pages of the US health reform legislation. But it is one of the most important points!

Pharmaceuticals

I 24

S EC TOR R EPOR T

In the US, the law on biosimilars is part of the Obama healthcare reforms

This new law is very vague about the practical aspects and leaves the field wide open to the FDA to define how the law will actually be applied, which has not yet been done. To this end, the FDA has recently begun a series of consultations with manufacturers to gather opinions before issuing regulations. The law stipulates the following: Approval of biosimilars will not be made by the same body that authorises conventional generics. The latter are covered by the Office of Generic Drugs. Approval of biosimilars will be given by the Office of New Drugs, as is the case for new molecules. The FDA has created a special unit, the Biosimilar Review Committee, to advise the other branches of the FDA on this subject, so that approvals are made on a consistent basis.

It specifies clinical trials

Approval will only be given when it has been demonstrated by analytical methods that: 1/ the copy is highly similar to the reference product, even if there are minor differences in terms of the non-clinically active components; 2/ the pre-clinical trials have been undertaken; and 3/ one or more clinical studies has demonstrated the products safety and efficacy for at least one indication. In addition, under the law, the biosimilar must have the same mechanism of action as the reference product if it is known, as well as the same method of administration, and the same dosage. The production unit also has to manufacture a sufficiently pure product. However, the law states that the FDA can decide to waive these conditions at its discretion! The FDA therefore has supreme authority when making its decisions.

Now that the legal framework has been set up, the FDA has to establish a regulatory framework

Its stance on many points has yet to be worked out. Some time ago, the FDA began to consider this issue, and initiated a series of consultations, including a public meeting in early November at which different opinions from the pharma and generics industries were expressed. The main stumbling blocks are as follows: The type of clinical studies required to demonstrate biosimilarity, i.e. what should we understand by highly similar? Also, the trials required to demonstrate interchangeability on top of biosimilarity. The FDA should take it one step further than the EMEA We see two differences in the rationale employed by the FDA and that of the EMEA. For starters, under the law, certain biosimilars can be considered interchangeable. This interchangeability signifies that a biosimilar copy of a drug can be used without needing approval from the practitioner who prescribed the original. Under the law, a biosimilar may be determined to be interchangeable if it can be expected to produce the same clinical effect as the standard. Yet again, it is up to the FDA to define what this implies. It is also up to the FDA to define a Risk Evaluation and Mitigation Strategy (REMS). However, the greatest difference with the situation in Europe has to do with the possibility of challenging patents. This mechanism is intended to replicate one that exists already through the Hatch-Waxman Act concerning chemical molecules. Under this Act, companies can file ANDAs (abbreviated new drug applications) to launch a copycat version of a drug that is still protected by a patent if they can have the patent invalidated by a court. This is only possible in the US and it has weighed heavily on the valuation of pharmaceuticals groups due to the high risk it poses to their products. The vast majority of major drugs available in the US have been targeted by ANDAs. The new law makes it possible to file ABLAs (abbreviated biological licence applications) under section 351 (k). There is one new feature: an arbitration procedure between the manufacturer of the
Pharmaceuticals I 25

S EC TOR R EPOR T

biosimilar drug and the originator of the molecule before an actual court battle. This is a very complex procedure that makes up 22 of the 46 pages in the Affordable Care Act (ACA).
Question marks remain over their interchangeability and the possibility of challenging patents upstream

This is partly due to the fact that, in contrast to the situation with conventional chemical molecules, the patents in this case are not officially identified. The FDA centralises patents covering conventional drugs in its Orange Book on its website. The patents involved are clearly stated and the generic drugs producers can tell which patent they will need to challenge in order to bring a copycat onto the market. They are required to tell the originator which patent they have decided to challenge. This is not the case with biological products. Given that such drugs are not eligible for the NDA procedure, the patents that cover them are not included in the Orange Book. This means that potential challengers cannot know with any certainty which patent the producer of the original molecule may invoke. The new law makes it compulsory for certain data to be exchanged between the generic producer and the originator of a molecule: the originator must provide a list of patents that may protect its molecule to any generic drugs producer who comes forward; the generic producer must provide the originator with data concerning its biosimilar drug, including a description of its production methods. The conditions and deadlines for this exchange of information are governed by very specific rules. Following the data exchange, the two companies are required to enter into an arbitration procedure before taking the matter to the courts. In view of the deadlines involved, this process is expected to last at least six months. It is important to note that, in the event of a challenge to their patents, biological drugs are not protected by the 30-month stay period that chemical molecules currently benefit from. However, the first biosimilar drug to win its case benefits from a one-year exclusivity period, compared to just six months at the moment in the case of chemical molecules. The purpose of this procedure is to put an end to the lengthy legal battles that are jamming up US courts. However, there is no guarantee that these measures will be used on a large scale. The procedure is complex and it is unlikely that the manufacturers of biosimilar drugs will be willing to hand over their data without batting an eyelid.

We believe that these possibilities will be used less than for traditional generics

Insofar as biosimilar drug makers carry out clinical trials, it could be in their interest instead to attempt to prove that they have developed a genuinely new product, and thus opt for a BLA procedure. This option presents a number of advantages by comparison with the 351 (k) procedure: if approved, the drug would be granted a 12-year exclusivity period; the application filed for the biosimilar drug would remain confidential; patent-related matters could be addressed without suffering the FDAs involvement. Yet again, it is difficult to predict how things will pan out. However, given the complexity of the procedure, we find it hard to imagine that ABLAs will be used as extensively as ANDAs. The more complex and controversial the technology (e.g. monoclonal antibodies), the less inclined generic drugs producers will be to opt for this approach.

Pharmaceuticals

I 26

S EC TOR R EPOR T

Momenta: fact is stranger than fiction The negotiations currently underway concerning legislation in the US will probably spawn a more stringent regulatory framework than in Europe. However, this may only be the beginning. Under a worst-case scenario for the biological drugs industry, the licence given to the FDA could ultimately lead to a situation very close to the one that many conventional generic drugs are in. Science is progressing and one company in particular believes that it can revolutionise the situation. This company is Momenta.
Momenta could revolutionise the regulatory framework

Momenta caused a big upset after being awarded in July FDA approval for a generic drug that can be substituted for Lovenox. This sanofi-aventis drug is recognised as a biological product in Europe. With its generic drug M-enoxaparin, Momenta has taken a different approach to the one generally used for biosimilars. Momenta considers that it has completely characterised Lovenox and its production process, using new analytical techniques, including specific enzymes, improvements on conventional analysis methods (NMR) and patented mathematical techniques. The company claims to have developed an exact copy of Lovenox. In order to make this claim, it performed an analysis of several batches, after which it maintained that its generic drug was within the acceptable variability limits observed with the sanofi-aventis drug. In other words, Momenta claimed that it was capable of producing an exact copy of Lovenox, using the production method employed by sanofi-aventis!

thanks to an analysis method that makes it possible to obtain exact copies

It took a long time for Momentas product to be given a green light (five years!). It is important to stress that the application was not only reviewed by the Office of Generic Drugs but also by the Office of New Drugs, in anticipation of the new legislation. Whats more, the FDA decided to split hairs when it came to the products immunogenicity. It wanted, in particular, to obtain assurance regarding the risk of heparin-induced thrombocytopenia (HIT), a potentially-serious side effect linked to modifications in the immune system. This could have seen the FDA ordering Momenta to carry out clinical trials. Considering the very low frequency of HIT, the company would have needed to enrol a large patient population (10,000?) in the hope of obtaining a statistically significant demonstration of the innocuousness of the generic drug. From an economic perspective, this would not have been an option. Momenta was able to convince the FDA that its analytical technique was powerful enough to avoid having to carry out anything more than limited preclinical trials. The FDA therefore approved the product as a substitutable generic drug. This is a groundbreaking decision as it signifies that the FDA accepts that Momentas characterisation technology can produce a copy that is close enough to the original to: 1/ warrant authorisation of automatic substitution; and 2/ avoid ordering clinical trials to prove this. In other words, a situation very similar to the one observed with conventional generic drugs, and hence potentially likely to wreak as much havoc in the pharmaceuticals industry. However, with the help of Sandoz, Momenta plans to apply its technology to practically every biological product on the market. Copaxone, a peptide developed by Teva, is first on the list. Copaxone is used in the treatment of multiple sclerosis. Sandoz filed an ADNA in 2007. This is expected to provide further evidence of the validity of Momentas technology. Momenta is still battling with Teva over patent issues and we would expect the case to go the courts in 2011, which could see the arrival of a generic form in 2012.

Following on from Lovenox, Copaxone is next in line for Momenta

Assuming Momenta succeeds with copaxone, it plans to move on to glycoproteins. This would not involve any radical change from Lovenox and Copaxone. By 2016, we think that Momenta could be in a position to launch generics that would be substitutable for simple proteins (growth
Pharmaceuticals I 27

S EC TOR R EPOR T

hormones, EPO, etc.). This would mark a small revolution, which we have decided not to factor into our models for the time being. The final stage would be to attack monoclonal antibodies. However, the fabrication process is far more complex and we would expect the burden of proof to be higher. On this basis, we would not expect any potential substitutable generics to arrive before 2020, if at all. If Momenta succeeds with the development of pure copies of biological drugs, this would move the goalposts for biological products. What had been a threat from biosimilars would become a threat from generic drugs, with even more disastrous consequences for the biological drugs industry.
possibly followed by even more complex molecules

That being said, based on Momentas success with Lovenox alone, which was a very specific situation, this does not imply that the companys technology will make it possible to copy all biological drugs. For one, in some cases, production costs would be prohibitive, e.g. insulin, which necessitates heavy investment, even though there would be no particular problem copying what is a well-known molecule. As for monoclonal antibodies, we would expect the FDA to keep a hard line on this issue, considering the multiple-stage production method used for such products, which is more complex than for simple proteins. We would not expect a substitutable generic form of a monoclonal antibody to be developed for many years, and certainly not in the coming decade. Hence, at this stage, we are basing our evaluations on biosimilar copies of monoclonal antibodies and insulins for instance, and not on substitutable generics. To sum up, we believe that the US situation as regards biosimilars should become closer to traditional generics than in Europe.

Commercial impact: what can be learned from Europe?

Fourteen biosimilar drugs have been launched in Europe since the middle of 2006. The European market thus gives us precious examples of the trends in market share for biosimilar drugs as well as in their prices.
Table 9: Main biosimilar drugs launched in Europe
Commercial name Omnitrope Valtropin Binocrit Epoetin Alfa Hexal Abseamed Retacrit Silapo Ratiograstim Tevagrastim Filgrastim Ratiopharm Biograstim Zarzio Filgrastim Nivestim Source: Hospira Generic name somatropine somatropine epoetine alfa epoetine alfa epoetine alfa epoetine zeta epoetine zeta filgrastime filgrastime filgrastime filgrastime filgrastime filgrastime filgrastime Company Sandoz Biopartners Sandoz Sandoz (Hexal) Medice Hospira Stada Ratiopharm Teva Ratiopharm CT Arzneimittel Sandoz Sandoz (Hexal) Hospira Standard Genotropin (Pfizer) Humatrope (Lilly) Eprex (J&J) Eprex (J&J) No Eprex (J&J) Eprex (J&J) Neupogen (Amgen) Neupogen (Amgen) Neupogen (Amgen) Neupogen (Amgen) Neupogen (Amgen) Neupogen (Amgen) Neupogen (Amgen) CHMP opinion Jan.-06 Feb.-06 Jun.-07 Jun.-07 Jun.-07 Oct.-07 Oct.-07 Feb.-08 Feb.-08 Feb.-08 Feb.-08 Nov.-08 Nov.-08 Mar.-10 Feb.-09 Feb.-09 Jun.-10 EMEA approval Apr.-06 Apr.-06 Aug.-07 Aug.-07 Aug.-07 Dec.-07 Dec.-07 Sep.-08 Sep.-08 Sep.-08

Pharmaceuticals

I 28

S EC TOR R EPOR T

Even with price discounts of 20%, the first biosimilars are finding it hard to make their mark

Amgen has communicated its market share in the EPO and G-CSF segments. In both cases, although its products carry discounts of 20% to the originals, growth in market share is weak and slow. In terms of value, after close to three years on the market, Aranesps biosimilar drugs account for just 6% of the treatment of anaemia in patients on dialysis. This market share has remained stable even with the arrival of new biosimilar EPOs. The problem in this market is that pharmaceuticals groups have reacted by lowering the price of their products by around 9% since November 2008. This, in turn, has led generic drugs producers to cut their own prices. This downward spiral has led to a contraction of 15% in the EPO market since 2007. This is evidence of how difficult it is to alter prescription habits: doctors are reticent to prescribe a product over which they have doubts with respect to its quality, and patients tend to prefer to stick to their branded treatment.
Figure 9 European EPO market market share trends in value

Source: Amgen

Market shares of 5%-6% in the two largest markets targeted

The situation is hardly better in the market for neutropenia treatments, where biosimilar G-CSF products have captured 5% of the market in the twenty months they have been available. However, they appear to be growing their market share a little more rapidly than biosimilar EPOs.
Figure 10: European G-CSF market market share trends in value

Source: Amgen

Pharmaceuticals

I 29

S EC TOR R EPOR T

In the longer term, we expect much more pronounced penetration for biosimilars

To sum up, biosimilar drugs in Europe, where substitutable generic forms (interchangeable) of biological products are not currently possible, have not triggered any major changes in their given markets. We believe that this situation is provisional. With the spread of biosimilar drugs across many drug classes, we would expect them to gain more credence with practitioners and take on more market share. On this basis, we are expecting erosion in the sales of the original products to speed up (-20% p.a. vs. -10%/-15% observed until now). In addition, this model will remain valid only for as long as substitutable generics remain prohibited. Over the very long term, the situation should come more into line with that of conventional generics. This could happen in less than ten years time if Momenta has its way.

Pharmaceuticals

I 30

S EC TOR R EPOR T

4.

The biosimilars players

There are few credible players on this very high potential market. The European experience and the future US guidelines teach us that the criteria of success lie not only in the capacity for development and production, but also in the ability to differentiate future products between themselves. This assumes analytical and development skills as well as marketing expertise.
Chart 11: Biosimilars market ($bn)
10 8 6 4 2 0 2 010 2011e 2012e 2013e US
Source: Merck

2014e EU G5

2015e

2016e

2017e

A branded product business model

The principal idea underlying regulations is that there is no certainty of obtaining a sufficiently exact replica of the original biological product that would avoid differences in effectiveness / side effects when different producers are involved. The only way of assuring sufficient similarity is to conduct clinical studies. There are two consequences if the product is not recognised as identical: The product will be sold under a different brand name, thereby implying marketing costs. The product will involve development costs that could prove substantial, as the healthcare authorities will often require phase III studies involving several hundred patients in order to assure the safety of the product.
Biosimilars will have a strong marketing content

Additionally, given industrial investments on a completely different scale than those required for a chemical synthesis, it is understandable that product costs are much higher than for traditional generics, thereby preventing price busting. The business model will therefore be completely different. The biosimilars will have an intermediary economic logic between pharmaceutical products and traditional generics.

Pharmaceuticals

I 31

S EC TOR R EPOR T

Table 10: Biosimilars a different breed of animal

Generics Development cost Development time Economic logic $0.5-5m 1-4 years Pure commodity: no differentiation between competitors for a given product, competition based exclusively on price Multiple competitors by segment None 15% Cost of production exclusively 15%-20% High substantial dependence on the US ANDA system, leading to highly volatile results as a function of processes / exclusivity periods Non-existent in a standard case possibility of slightly different formulations Chemical synthesis Biosimilars $50-150m 7-8 years Intermediate situation: price competition, but no substitution. A certain degree of differentiation between products is possible based on marketing Several competitors Weak 70% Sales force, formulation, product quality, price 20%-25% Moderate dependence in principle on the ABLA system consequently, launches after expirations and relatively low cyclicality for results High Genetic engineering / cell culture Original drug $500-1,000m 10-15 years Complete speciality: price determined as a function of the added therapeutic value. Very high price flexibility during the period of patent protection in innovative segments. Few competitors for a given product Very high during patent protection 100% Added therapeutic value, commercial presence 30%-35% Very long and very limited cycles (patent cliff in 2008/2012).

Structure of the competitive environment Pricing power Price level (% of original product) Competitiveness factor Normal operating margin (% of revenues) Cyclicality

Barriers to entry for new competitors Production technology Source: Natixis

Very high Chemical synthesis / genetic engineering

Technology also involves production methods

Biosimilars producers must effectively master the same technologies as producers of biotechnology products, i.e.: The development of cell lines and their culture. Industrial fermentation. Virology. Biologics-specific separation techniques. Production in sterile facilities or controlled microbiology. These are not chemical company expertises. As such, producers must evolve. This shift has been successfully made by the Swiss company Lonza, a former fine chemicals producer than is now entirely focused on subcontracted production of biological products (for example, for Roche Genentech). Candidates will not only have to make high investments for the development of each product (cf. table), but will also need skills that themselves require years to acquire. This contributes to the barriers to entry in the sector. There will consequently be fewer biosimilars companies than traditional generics producers. In summary, biosimilars (in contrast to traditional generics) are not commodity products, but indeed relatively differentiable valued added products that are less cyclical and more profitable than traditional generics due to much higher technological, financial and marketing barriers to entry.
Pharmaceuticals I 32

S EC TOR R EPOR T

The players here will therefore be different from traditional generics producers, above all due to the major role of marketing. These players will include: Major generic drug producers. Pharmaceutical groups seeking to position themselves in growing areas.

Generic drug producers: few biosimilar candidates

Numerous players from widely varying horizons are beginning to invest in this field, particularly in emerging markets, where Indian groups (Biocon, Dr Reddys, Wockhardt, etc.) are very active. But few groups look to have the expertise needed to satisfy the demands of the FDA. As we see it, the best placed groups are Sandoz, Teva and (ultimately) Pfizer.
Chart 12: Breakdown of the biosimilars market in 2009 (250m total)
Others Stada 3% 7% Hopsira Sandoz 47% Tev a/ Ratiopharm 27% 16%

Source: Sandoz

Table 11: Criteria of success

Sandoz Products on the market in Europe Europe: 3 products - Growth hormone - EPO - G-CSF* US: 1 product Growth hormone (Omnitrope) 8-10 drugs In-house Teva Europe: 2 products - Growth hormone - G-CSF* US: 1 product Growth hormone Europe US: Neugranin (long duration GCSF*) 2 monoclonal antibodies Agreement with Lonza Europe: partners US: 2 projects in phase I Partnership with Cellpharm In-house / Biocon Projects in the insulin area Hospira 2 products - EPO (partnership with Bioceuticals) - G-CSF*(Pliva) Stada 1 product - EPO Partnership with Bioceuticals Pfizer None Biocon Indian generic producer

Development capacity

Production Marketing Presence Solidity of financial structure Source: Natixis

4 sites Synergies with pharmaceuticals Global 1st biosimilar in Japan Excellent

Current difficulties Strong presence in hospitals and clinics and with payers Global Europe Good Weak

In-house / Biocon In-house Global Good Local

Global Good (however, numerous recent acquisitions)

Pharmaceuticals

I 33

S EC TOR R EPOR T

Teva: unsatisfactory no. 2 position for the world leader in generic drugs The objective of Teva, the world leader in generic drugs, is to double its revenues to $31bn looking out to 2015 while maintaining a net margin of 22%. This medium-term objective has the advantage of enabling the market to accept the generic risk weighing on Copaxone (which is being targeted by Momenta) and to be enthusiastic about future sources of growth. This targeted growth will involve acquisitions and the emergence of new generic segments such as respiratory products and biosimilars.
Acquisitions and a partnership with Lonza

Teva is developing in this segment through acquisitions and partnerships: First, the acquisition of Barr in 2009 enabled Teva to double its presence in biologics (by the number of its researchers, research and production sites and number of projects under development) and its production capacities. The acquisition in 2008 for $400m of CoGenesys (spin-off from Human Genome Sciences), a specialist in albumin fusion processes (albumin is one of the principal plasma proteins and is consequently a vector of choice for the diffusion of an active pharmaceutical principle). Finally, the signing in January 2009 of a historic agreement with the Swiss group Lonza, the leader in antibody production for major pharmaceutical groups. Through this agreement, the two groups set up a joint venture dedicated to the development of a portfolio of biosimilars.

Three R&D sites

Teva has expertise in both discovery methods and industrial processes. The groups three biotech R&D sites (located in the United States, Lithuania and Israel) provide it with mastery of recombinant protein expression, microbiotical fermentation, mammalian cell culture, purification processes and formulation analysis and development methods. These expertises are supplemented by the groups expertise in the production of sterile finished products, biological pharmaceutical ingredients (Lonza) and albumin fusion technology (CoGenesys) needed in the development of sustained release formulations. Sales of biosimilars have grown very gradually, rising from $50m in 2007 to $74m in 2009 through a G-CSF in Europe and a growth hormone in the United States and Europe. Among the most advanced R&D projects, we would note: Tevagrastim, a G-CSF for which a BLA application was filed in the United States in December 2009 (the product is already marketed in Europe). This product will be sold under the Neutroval brand name. Neugramin, a long duration G-CSF (biosimilar of Amgens Neulasta) that is in advanced phase development in Europe and the United States. Two monoclonal antibodies

One of the sources of growth needed in order to double sales by 2015

Like Sandoz, Teva has decided to reinforce its positions in activities that feature the same requirements as biosimilars, with an initial focus on respiratory diseases (26% of brand name product sales, corresponding to $2.4bn by 2015 and with average annual growth of over 20%). In Europe, where the development of copies of respiratory products is better regulated, Teva is testing four molecules for which filings are expected in 2011 and 2012, particularly a copy of Symbicort (budesonide + formoterol) with the Spiromax inhaler (filing in 2011) and a copy of Advair (fluticasone + salmeterol) with Spiromax or a variable dose inhaler (2012).

Pharmaceuticals

I 34

S EC TOR R EPOR T

Table 12: Forecast Teva sales

$m Generic drugs Innovative (copaxone & Azilect) Respiratory products API Womens health Biosimilars Group sales Source: Evaluate 2009 9,340 2,665 898 565 357 74 13,899 2010e 10,843 2,970 1,071 591 377 105 15,957 2011e 12,754 3,263 1,257 612 413 134 18,433 2012e 14,953 3,740 1,520 615 420 184 21,432 2013e 17,512 4,051 1,845 654 458 265 24,785 2014e 20,251 3,925 2,156 710 511 365 27,918 2015e 23,109 3,757 2,425 764 562 524 31,141 CAGR 2010/2015 (%) 16.3 4.8 17.8 5.3 8.3 37.9 14.3

Sandoz: the pioneer No. 2 in the generic drug industry, Sandoz (a subsidiary of Novartis) is the pioneer on the highly specific biosimilar market.
Sales of $190m expected in 2010 in the biosimilars area...

Sandoz has been present on this market since 2006, bolstered by its cutting-edge position in high valued added generics, its key presence in Europe (where the first biosimilars were authorised) and the financial support of its parent company. Novartis has effectively invested in new capacities (three sites are dedicated to recombinant protein production) and in R&D (four centres are dedicated to biosimilars) and has carried out a number of major deals, with the acquisition of Ebewe and Oriel. Sandozs biosimilar sales should therefore rise from a still modest $118m in 2009 to $190m in 2010. This segment posted a 76% increase in sales over the first nine months of 2010 compared to 10% for the entire division. Sandoz has consequently become a key asset for Novartis, both in terms of its growth dynamic and for the reinforced bargaining power with payers that it provides. A strategy that has long been focused on high value added generics
Table 13: Pillars of growth at Sandoz
Product line Biosimilars Competitive positioning in 2010 No. 1 Market share around 50% No. 2 Acquisition of Ebewe in 2009 Launch of Enoxaparin Top 5 Acquisition of Oriel Size of target market Sales of products with patents expiring by 2015: $64bn Sales of products with patents expiring by 2015: $8.3bn (anti-cancer treatments). Sales of products with patents expiring by 2015: $13bn ($17bn in 2016) Treatments for asthma and chronic obstructive pulmonary disease (COPD) World market: $80bn in 2009. 2009 sales ($m) 118 2015 sales ($m) 1,000e

Injectable generics

300e

600e

Respiratory products

1,000e

Total Sandoz sales

No. 2 behind Teva

7,493

12,400

Source: Novartis, Business Review 2010

With estimated sales of $8.3bn in 2010, Sandoz is the no. 2 generics producer worldwide behind Teva. After a period of rapid growth (+33% from 2004 to 2007, largely underpinned by the acquisitions of Eon Labs and Hexal in 2005), Sandoz entered into a normalised growth phase in 2008/09 before returned to strong sales growth in 2010. This renewed dynamism reflects 1/ a more aggressive launch strategy in the United States (tacrolimus, losartan, lansoprazole) and Europe
Pharmaceuticals I 35

S EC TOR R EPOR T

(generic version of Plavix) as well as 2/ the takeoff of sales of biosimilars in Europe and the United States with the launch at the end of July of M-enoxaparin
Chart 13: Growth in US and biosimilar sales in local currencies

76% 43% 24% -1% -5% 2% 5% 19%

-11% Q1 08
Source: Novartis

-17% Q3 08

-14% Q4 08

Q2 08

Q1 09

Q2 09

Q3 09

Q4 09

Q1 10

Q2 10

Q3 10

to which can be added $450m in sales for M-Enoxaparin

The development alliance with Momenta (cf. US section) constitutes one of the pillars of Sandozs strategy in the biosimilars area. We would highlight the fact that M-enoxaparin got off to a very strong start, with sales of $292m in Q3 10, its first quarter of commercialisation. This was higher than sanofi-aventis Lovenox sales in the United States on the primary care market targeted by Sandoz (40% of Lovenox prescriptions come from family physicians vs. 60% from hospitals). Note however that sales were inflated by an inventory effect that will tend to reduce sales over the coming quarter. The group will consolidate 100% of sales and 55% of profits until another generic producer enters the market. Momentas development projects include a generic version of Copaxone, the leading product of Teva, which in this situation plays the role of the besieged party and is using legal and regulatory tools (including citizen petitions among others) to attempt to block Momenta. In-house development also represents sources of future growth. Sandoz has launched three products in Europe and one product in the United States since 2006. The first products commercialised in Europe were the growth hormone and white and red blood cell stimulating factors (G-CSF and EPO). Combined sales are becoming more and more significant each quarter ($4m in monthly sales for the biosimilar EPO in 2008, $15.8m in 2010). The future will now be determined outside of Europe, in priority in the United States or even Japan, where Sandoz was the first generic producer to launch a biosimilar (EPO, Omnitrope in 2009). The challenge in Europe is to boost its market shares. This should be possible thanks to the commercial support of both Sandoz and the Pharma division. Additionally, eight or nine drugs are under development targeting biological products. Management has not provided details concerning these projects, but has clearly ruled out the insulin segment, which is viewed as insufficiently profitable. We believe that Sandozs programmes involve the alpha and beta interferons, with the group additionally emerging as one of the key players in the treatment of multiple sclerosis. Other high value added development programmes are in advanced phases, notably in the respiratory diseases area. The products being studied show similarities with biosimilars in terms of their difficulty of development rather than by their nature. Additionally, the asthma and COPD treatment markets will approach $17bn in 2015 (including Advair, Symbicort and Spiriva). From a regulatory point of view, Sandoz estimates that its products are completely capable of benefiting
Pharmaceuticals I 36

8 to 9 new drugs are under development

S EC TOR R EPOR T

from automatic substitution by payers (in contrast to Teva, which is not counting on its products being considered as direct substitutes by the US regulator). Nevertheless, Sandoz is entering this market very cautiously. After a relatively unfruitful partnership Vectura (abandon of the agreement concerning the generic version of Advair in the United States, continuation in Europe), Sandoz now appears to be betting on the US company Oriel Therapeutics, which it acquired on 1 June 2010 for $332m ($74m upfront and $258m linked to certain development milestones). Oriel came with three projects under development and innovative inhalation systems. We would emphasise the fact that inhalators constitute the key differentiating criteria on this market. Novartis could claim a complete line ranging from a subscription product (QAB 149, whose launch is underway in Europe and planned in the United States in 2011) to generic products.
Sandoz is more generally targeting differentiated generics that are difficult to develop

Finally, Sandoz is looking into injectable generics. Sandoz took over Ebewe in September 2009 for $1.2bn. The high valuation (2008 sales of 188m ($272m) reflected this companys exceptional profitability (2008 operating profit: 53m), unique technology concerning administration and formulation methods, strong position in the oncology area (which will clearly see strong growth over 2010/2012 due to patent expirations for major drugs with combined sales of $10bn) and growth outlook (+20%e after +26% in 2007 and +17% in 2006). Financial impact

With forecast operating profit growth of 9% per year, Sandoz represent a longterm source of growth for Novartis

We estimate that Sandoz will show average annual sales growth of 8.2% over the next five years. We therefore consider that Sandoz should be Novartis most dynamic division over 2010/2015. After a very good year in 2010 (sales +15% in local currencies, +10% organic growth over the first nine months), Sandozs top line momentum should be underpinned by 1/ the growth of the world market (estimated at 7-9% per year and more likely +15% in volume terms due to patent expirations for major blockbusters and the boom on the emerging markets and even Japan, where the penetration rate of generics remains very low at 17% in volume terms versus 70% in the United States and Germany) and 2/ the strong development of the biosimilar and respiratory product lines. This positioning on the biosimilars and more generally high value added generics (respiratory, injectables which will account for 23% in 2015 versus 5% in 2010) combined with constant industrial productivity efforts should enable an increase in the operating margin starting in 2015. In the meantime, we estimate that Sandozs has little leverage to boost its operating margin, which reached a record level of 21.2% over the first nine months of 2010. Productivity gains should serve to offset pressure on the German market (Sandozs no. 2 market) and the increase in R&D spending. As such, we are forecasting operating profit growth of 8.8% per year on average between 2010 and 2015 (+26% as of end September 2010).
Table 14: Forecast Sandoz sales and operating profit $m Sales % change - Biosimilars - Injectables - M-Enox - Rest of portfolio % change Adjusted operating profit Adjusted operating margin (%)
Source: Natixis

2009 7,493 118 300 7,075 1,395 18.6

2010e 8,350 11.4 190 350 450 7,360 4.0 1,669 20.0

2011e 9,250 10.8 240 400 650 7,960 8.2 1,880 20.3

2012e 10,100 9.2 500 450 400 8,750 9.9 2,050 20.3

2013e 10,900 7.9 650 500 350 9400 7.4 2,280 20.9

2014e 11,650 6.9 800 550 300 10,000 6.4 2,500 21.5

2015e 12,400 6.4 1,000 600 250 10,550 5.5 2,550 20.6

Pharmaceuticals

I 37

S EC TOR R EPOR T

Hospira Hospira is the world leader in injectable generics, commercialising lines of anti-cancer drugs and products used in intensive care wards. 2009 sales reached $3.9bn, boosted by the acquisition in 2007 of Mayne Pharma (an Australian company specialising in injectable products) for $2.1bn.
Hospira has become a significant player, with two biosimilars on the market

Following the launch of its biosimilar version of EPO (Retacrit) in 2007 and its G-CSF biosimilar (Nivestim) in March 2010, the group has solid ambitions in this area. In these two cases, Hospira has signed a commercial licence agreement with drug developers: Bioceuticals (in which Stada has a 15.9% stake) for the EPO biosimilar and Pliva for the G-CSF biosimilar. Hospira is one of the very few players to disclose its biosimilars pipeline. Two projects are currently in phase I studies in the United States and Canada: the pegylated version of filgrastim (white blood cell growth factor or G-CSF) and Retacrit. While these two drugs were developed by partners in Europe, Hospira is responsible for the clinical trials in the United States and Canada. Finally, three monoclonal antibodies are also in the screening test phase and could enter into preclinical trials soon.

Pharmaceuticals

I 38

S EC TOR R EPOR T

Chart 14: Hospira pipeline

and three monoclonal antibodies in early development
Cell line/ Retracit APPROVED Retracit nevestim Filgrastim IV oncology indication EU Screnal indication Filgrastim EU EU AUS Preclinical Phase I Phase II Phase III Submission Launch

Pegfilgastrim EPO PIPELINE mAb1 mA2 mAb3 Renal/oncology Oncology Oncology Oncology

Global US/CAN Global Global Global

Source: Hospira

Even if the company has a good commercial presence with payer organisations

However, the group has not yet recorded positive results in this segment. Despite the launch of a first biosimilar in Europe in 2008, growth remains weak here, reaching only +6.5% in Q3 vs. +9% overall for the company (excluding the impact of Eloxatin, whose sales fell to zero in Q3 10). Management has therefore lowered its guidance for sales growth in 2010 from 3-5% to 2-3%. Finally, concerning production, the group was inspected by the FDA and received a warning letter concerning the quality of certain processes. This led Hospira to record charges ($85m expected for the full year), partially extraordinary (penalties for failure to supply required information, consultant fees etc.) and partially recurring (operating expense for the recall of manufactured products).

its commercial performances are not yet convincing

Hospira has high ambitions based on its solid commercial positions (with hospitals, oncology clinics, dialysis centres and payer organisations). However, the group will have to prove its development abilities. Its credibility in this area will come from the entry into clinical testing of its first monoclonal antibodies and the approval of the EPO and G-CSF biosimilars in the United States (the group was not responsible for testing in Europe). Hospira will also have to demonstrate its abilities in terms of production quality control, a factor to which the regulatory authorities are even more sensitive in the case of protein production. As we do not consider commercial presence to be a differentiating factor, we do not believe that Hospira will be a leading player in this area over the medium term. Stada: an outsider in the biosimilars area Stada is developing its biosimilars through Bioceuticals, a specific non-consolidated project financing structure in which it holds a 15.86% stake. It has a call option here that will enable it to acquire a 100% stake starting in 2011 at a pre-determined price. Bioceuticals posted 2009 sales of 26.7m (vs. 16.1m in 2008) and a net profit of 0.9m (vs. -14.3m in 2008).

Stada could become a future player through its non-consolidated subsidiary Bioceuticals.

Among the treatments developed by Bioceuticals, the Epo-Zeta (Retacrit/Silapo) is the only one already being commercialised (since the beginning of 2008). Hospira has obtained exclusive distribution rights in Europe (with the exception of German and certain eastern European countries, where Hospira has only semi-exclusive rights alongside Stada). Through this partnership, Bioceuticals will receive milestone payments from Hospira over coming years that could reach up to an additional 14m. This amount will depend notably on the possible commercialisation of Epo-Zeta in the United States and Canada. Epo-Zeta could also have potential applications in nephrology and oncology. The other development projects are more uncertain. Research efforts concerning Filgrastim (phase I, immunostimulant, worldwide commercialisation licence granted to Stada
Pharmaceuticals I 39

S EC TOR R EPOR T

subsidiary Cell Pharm) were suspended in 2009 and efforts concerning the monoclonal antibodies will not proceed until Bioceuticals finds a partner (2009 R&D spending: 2.5m). Over the short term, the potential improvement in profitability (14.3% in 2014e vs. 13.5% in 2009) will come more from productivity efforts than the boom in biosimilars. The groups restructuring plan (Build the future) aims to optimise production costs looking out to 2013 through economies of scale and the transfer of production capacities to low cost countries in order to offset price reductions. This implies the sale or abandon of certain production facilities as well as the outsourcing of different functions, thereby leading to simplification in group structures and major workforce reductions in all divisions and regions with the exception of Germany (800 full time employees, 10% of the workforce). Productivity gains make credible the announced objectives for 2014, which target a doubling in net profit in five years based on average annual sales growth of 7% and margin improvements. Note that in Q3 10, the operational difficulties in Serbia were offset by the performances of the branded products division, with the group adjusted operating margin stable at 14%.
Table 15: breakdown of the Stada operating margin m Sales generics branded products Adjusted operating profit
Managements generics priority is to lift the branded products profitability of the generics division Adjusted operating margin (%)

2009 1,569 1,116 393 211 159 80 13.5 14.3 20.2

2010e 1,590 1,095 420 220 159 88 13.8 14.5 21.0

2011e 1,690 1,160 450 235 170 95 13.9 14.7 21.1

2012e 1,830 1,260 485 252 190 100 13.8 15.1 20.6

2013e 1,970 1,365 515 275 210 105 14.0 15.4 20.4

2014e 2,100 1,470 535 300 227 110 14.3 15.4 20.6

generics (%) branded products (%)

Source: Natixis

An eldorado that is also attracting the pharmaceutical groups

The biosimilars segments business model, which is characterised by high marketing content and very strong market growth, has attracted the interest of unexpected players in the form of the pharmaceutical groups, which usually find themselves on the defensive here. Certain groups such as Novo considered the idea but then abandoned it. However, several groups are preparing to launch biosimilars, including Merck, AstraZeneca, Amgen etc. Nevertheless, Pfizer is by far the group with the most advanced and most ambitious programme here.
Pfizer is targeting the insulin market

Pfizer: accelerated efforts Following Novartis, Merck, AZN and Amgen, the world leader Pfizer is the fifth group to publicly announce its interest in biosimilar. However, in contrast to its peers, Pfizer has decided to initially proceed through a partnership with the Indian group Biocon, which reported 12-month sales as of end June 2010 of $512m, up 44% (+24% in H1 11), with 50% coming from biological copies. This agreement (announced in October 2010) signals a generalisation of biosimilars in all categories of biologicals, with an initial focus on insulins.

Pharmaceuticals

I 40

S EC TOR R EPOR T

The agreement calls for Pfizer to pay Biocon $200m upfront as well as up to $150m in milestone payments in return for an exclusive worldwide commercialisation licence. Biocon remains responsible for clinical development, production, supplies and product approvals.
The Indian group Biocon is a credible player in endocrinology

This agreement initially concerns diabetes. Biocon has already commercialised Insugen in India, with this product representing 10% of the national market. This product is also in the launch phase in Latin America, Asia and North Africa. The filing for approval for this product (as well as for Basalog, a generic version of Lantus) is expected between 2012 and 2016 in Europe and the United States (where Biocons management is targeting IND status). Biocon has adopted a commercial strategy worthy of a sector major with a glucose test (Breeze 2) supplied at no charge to patients treated with Basalog. In contrast, the administration system (pre-filled syringe) is not competitive with the latest generation insulin pens. Biocon is also a leading player in nephrology with Erypro (an erythropoietin), Renodapt (the generic version of Roches Cellcept (mycophenolate mofetil) indicated for the prevention of rejection following kidney transplants) and Tacrograf. Above all, the group has moved into the monoclonal antibody segment, with: The launch of BioMAb EGFR (generic version of Roches Herceptin), approved in 22 emerging markets and launched at a 50% discount compared to the price of the branded product. The phase III development of T1h, an anti-CD6 monoclonal antibody used in the treatment of psoriasis (clinical results expected in July 2011).

1 step: copies of traditional insulins in 2012 nd 2 step: analogues in 2015

st

Upstream research on monoclonal antibodies conducted in connection with the agreement signed with Mylan. Pfizers interest here is to enter the traditional insulin biosimilars market in Europe in 2012 and the United States in 2015, taking market shares away from Lilly and Novo Nordisk, the only groups currently present in first generation insulins (human insulins) in Europe and the United States (26% of Novo Nordisks diabetes sales in 2010). Over the medium term, the objective is clearly to produce modern (analogue) insulin biosimilars and to compete with sanofi-aventis Lantus and Novos Levemir and NovoRapid. This will only be possible following patent expirations (cf. table 22), with the first product concerned being NovoRapid. Over the long term, this agreement will give Pfizer access to monoclonal antibody biosimilars, an area in which it has begun to make direct investments. Granted, the acquisition of Wyeth already enables the group to claim recognised development and production capabilities in the biologics area. The group has as such announced a $300m investment programme in the biosimilar production area.

Pharmaceuticals

I 41

S EC TOR R EPOR T

5.

Big pharma: growth strategies in the face of biosimilars

Roche and Novo Nordisk are already getting ready to respond to the biosimilar offensive. Roche is in the front line, since all of its blockbusters are concerned. It enjoys lengthy patent protection and the products concerned are generally monoclonal antibodies, but its range will therefore see erosion in Europe and the USA once patents start to expire. The group must therefore as a matter of necessity find fresh sources of growth. The strategies employed by pharmaceuticals groups to fight back against biosimilars may be summarised in three points: Develop new formulations. Develop new delivery systems. Develop new chemical entities as successors. The first two aim to establish a technological barrier and, more importantly, to obtain patent extensions.

Roche: growth in spite of everything

Roche: 67% of pharma sales is generated with biological products

Roche generates two-thirds of its sales with biological products and could potentially face competition from biosimilars on most of its major products. The chart below summarises sales of the groups biological products in Europe and the USA.
Chart 15: Sales in Europe / USA of Roches biological products (first 9 months of 2010; CHFm)
4 000 3 000 2 000 1 000 0
Herceptin MabThera/Rituxan NeoRecormon Pegasys Actemra Lucentis Avastin Xolair

Source: Natixis

45.6% of pharmaceutical sales will suffer biosimilar competition

All told, 45.6% of pharmaceutical sales will suffer competition from biosimilars resulting in an erosion of sales (excluding emerging countries, where it generally already exists and does not have the same impact). However, these products were generally launched recently and enjoy lengthy patent protection, which postpones this outcome accordingly.

Pharmaceuticals

I 42

S EC TOR R EPOR T

Chart 16: Patent expiries per year (Europe / USA sales of biologics, CHFm)
7 000 6 000 5 000 4 000 3 000 2 000 1 000 0 2 010
Sources: Roche, Natixis

2 011

2 012

2 013

2 014

2 015

2 016

2 017

2 018

2 019

2 020

In fact, apart from Mabthera and Herceptin, which will lose patent protection in Europe in 2014/2015, the other geographical zones and products are protected until the end of the decade. This particularly puts the impact of biosimilars into perspective since the progression of the three blockbusters, Avastin, Mabthera and Herceptin, is already weak in mature zones, and their growth already stems almost entirely from the Japan / emerging countries zones, which will not be called into question.
But mostly in the next decade

The problem therefore mainly concerns the decline in growth in sales in the period 2018/2030. Note that by this time, by construction we can but envisage a gradual fall, insofar as new generations of more competitive products are likely to have emerged. The impact of biosimilars, in the absence of automatic substitution, will therefore correspond to a more rapid fall in sales. In our models, we expect a decline of 5%, then 10% per year in the absence of biosimilars, and 20% and then 25% per year with biosimilars. This results in the valuation difference indicated in the valuation section (cf. page 10). However, this does not include the strategies to secure the long-term survival of these franchises. Roche, in particular, is developing both successors and new methods of administration. The different initiatives are summarised in the table below.
Table 16: Roches strategy for ensuring the long-term survival of its biological products Product Mabthera Herceptin Avastin Actemra Recormon
Sources: Roche, Natixis

Sales 2014e (CHFbn) Successor 8.6 GA 101- RG1678 (2013) 7.0 7.9 1.3 0.8 Mircera (2007) Pertuzumab (2012) T-DM1 (2013) RG7334anti-PIGF RG7347anti-NRP1

SC from Yes Phase III Yes Phase III Probable Yes Probable

Pharmaceuticals

I 43

S EC TOR R EPOR T

The most original: the new subcutaneous forms of monoclonal antibodies

The counterattack has already been launched. First with new subcutaneous forms of monoclonal antibodies

Monoclonal antibodies are administered by intravenous infusion (IV). Beside the inconvenience for patients, this method of injection requires premises and staff, generating high costs for healthcare systems. Thus, at present, patients treated following a breast cancer operation to prevent possible metastasis and a relapse (adjuvant treatment), have, sometimes for a period of years, to come back to the hospital every three weeks to be put on a drip for several hours.
Table 17: Treatment protocols Drug / indication Herceptin breast cancer / adjuvant treatment Protocol With Taxol: 90 minutes on a drip initially + 30 minutes per week for 12 to 18 weeks. Then 30 to 90 minutes every 3 weeks, up to 52 weeks. Monotherapy or with Taxol: 90 minutes IV drip then 30 minutes once a week until progression of the disease.

Herceptin breast cancer /metastatic

Sources: Prescribing information, Natixis

The aim of Roches developments is to reduce these costs and this inconvenience through use of subcutaneous injections (comparable to those used by diabetics for insulin). These injections can be carried out in a few seconds, with small needles that are not painful. They would be carried out by a nurse or by the patients themselves at home. The savings for healthcare systems are substantial, and will lead to these techniques being supported by the organisations that pay for healthcare, in particular public bodies in the US. However, this supposes a technological leap forward, for which the group will rely, among other things, on the technology of a small US biotechnology company, Halozyme, which uses a recombinant enzyme. The agreements between Halozyme and Roche were signed in December 2006. They concern: 5 developments for which Roche has already exercised its option and for which clinical trials have been decided. Up to 8 others for which trials have not yet been decided. The first subcutaneous development of a monoclonal antibody, Herceptin, entered phase III in mid-2009. The study provides for an adjuvant treatment for one year after the operation, with two years of monitoring, for 550 patients. The results will therefore not be known until the start of 2014. A launch could therefore be envisaged at the end of 2014/start of 2015. In March 2010, Roche announced the entry into phase III of Mabthera in the SC form. The study starts in December 2010 and will enrol 530 patients. We expect a launch in 2013 or 2014.
Which will extend the patent protection of these products

We believe that Roche intends to develop subcutaneous forms of all of its big monoclonal antibodies, including Avastin (but for a limited share of the latters sales). Actemra is being developed in a subcutaneous form, but not with the same technology. Last, it is more than probable that Roches new generations of products will be rapidly developed in subcutaneous form (ocrelizumab, T-DM1, etc.). All told, we believe that the products concerned represent 2014 sales of CHF30bn, of which CHF15bn/20bn could switch to a subcutaneous form. The aim is to increase bioavailability and improve comfort of use for the patient, but also to prolong the patent life of blockbusters, Halozymes technology being protected by patent until 2024. However, the subcutaneous forms will only save part of the sales, insofar as they will only be an asset if the product is not associated with a chemotherapy, which in any case will be administered
Pharmaceuticals I 44

S EC TOR R EPOR T

via an IV infusion, i.e. especially in the adjuvant or management indication. We estimate that three-quarters of Herceptin sales and more than half of Mabthera sales could be concerned. Innovation as a fresh source of growth The alternative is to develop successors, which has the advantage of capitalising on the commercial franchise while benefiting from higher prices. A typical example is the strategy applied with the development of T-DM1. This product consists of a Herceptin molecule (trastuzumab) chemically linked to a cytotoxic product. Herceptin, a targeted therapy, will therefore preferentially fix to the cancer cells, which will be destroyed. This bomber effect results in improved efficacy.
Roche also has successors to most of its blockbusters, such as T-DM1 for Herceptin

The indications are, in principle, the same as those of Herceptin, and a large share of the sales will cannibalise those of Herceptin. However, the group is aiming for a hefty pricing premium, justified not just by superior efficacy, but also by a better side effect profile (cf. table).
Table 18: Phase II results for T-DM1 T-DM1 (n=67) Patients with an objective response Safety-evaluable patients Patients suffering from side effects, n (%) Side effects (grade III) 3 most frequent side effects with Herceptin + Taxotere
-

32 (47.8%) (n=67) 63 (94.0) 25 (37.3) 1 (1.5) 5 (7.5) 7 (10.4)

Herceptin + Taxotere (n=70) 29 (41.4%) (n=68) 68 (100.0) 51 (75.0) 45 (66.2) 39 (57.4) 31 (45.6)

Alopecia (hair loss) Neutropenia (reduction in the number of white blood cells) Diarrhoea

Source: Roche

The development of T-DM1 has not been an easy ride, however. After announcing the filing in May 2010, Roche had to withdraw it in August, as the FDA had refused the filing based solely on the phase II results. It now has to conduct phase III trials. The first, the Emilia study (980 patients), was launched in 2009, and should be enough for a filing in 2012 as a second-line treatment. Another study, Marianne, will make it possible to envisage an indication as a first-line treatment in 2014. T-DM1 will therefore arrive on the market just in time to counter Herceptin biosimilars, but probably not soon enough to prevent some erosion of the European franchise. Conversely, between now and 2019, we believe that Roche will be able to replace most of its Herceptin sales by subcutaneous forms or successors.
Generally speaking, the advanced-stage pipeline is well filled, guaranteeing long-term growth

In the case of Mabthera, the successor GA 101 looks interesting. It shares the same mechanism of action as Mabthera, it is an anti-CD20, but it is an antibody modified for better efficacy. This was already confirmed in phase I studies in terms of response rate, and it is being developed in Mabtheras main indications: CLL and NHL. Two phase III studies are underway, which should enable a filing in 2013. The product could arrive on the market in 2014, enabling a rapid substitution, but once again, it will in principle arrive a little late to avoid some erosion of European sales. These are just designated successors. But Roche has a well-filled pipeline, which despite some recent delays unquestionably ensures the companys long-term future, with or without biosimilars.

Pharmaceuticals

I 45

S EC TOR R EPOR T

Table 19: State of Roches products in phase III or entering phase III Product T-DM1 RG1273 Pertuzumab RG 7159 GA 101 RG 7204- Braf inhibitor RG 3616 Hedgehog inhibitor RG 3638 MetMab Ocrelizumab RG1658 dalcetrapib CETP inhibitor RG 1439 Aleglitazar Total phase III and equivalent
Sources: Roche, Natixis

Indication Metastatic breast cancer Metastatic breast cancer over expressing HER2 CLL NHL Melanoma Basocellular carcinoma Solid tumours Multiple sclerosis Dyslipidaemia Diabetes - cardiovascular risk

Status Phase III underway Phase III underway Phases III underway Entering phase III Pivotal phase II Entering phase III Entering phase III Phase III Phase III

Launch expected 2013 2012 2014 (CLL), 2015 (NHL) 2012 2012 2015 2014 2014 2014 2012/2015

Peak sales potential (CHFm) 1,000 800 3,000 800 300 600 600 5,000 600 12,700

Products in phase III represent CHF12.7bn, or 33% of pharmaceutical sales in 2010 with, it is true, a large share for dalcetrapib, or 44% including Actemra, which was launched recently and whose peak sales potential has yet to be unlocked. Given the absence of generification of major products, apart from Xeloda (end of 2013) and maybe Tarceva (2018 patent) and Boniva (on which ANDAs are ongoing), the growth profile therefore remains well assured, as Roche is the only one of the big pharma groups whose sales will not suffer the loss of a major product in the USA. We therefore expect a progression of around 5% per year in its pharma top line between now and 2015, then 4% to 5% per year. When the patent protection ends, US/Europe sales of Herceptin, for example, will account for just 6% of the groups pharma sales. Even an erosion of this figure will be hardly noticeable.

Insulins: biologics apart

The Pfizer / Biocon agreement shows the reality of the threat from insulin biosimilars. We believe that a widespread arrival of insulin biosimilars is inevitable, as for the other categories. But the impact will not be major. Specific barriers
Insulins are a special case. These products involve specific entry barriers

The question of the generification of insulins, in particular modern insulins, does not seem to particularly worry investors. Novo Nordisks PE, which is almost double the sector average, reflects the total absence of concern on this theme, which is not even the subject of investor questions. This is a paradoxical attitude since the main reason for the weakness of pharmaceutical sector valuations is fears regarding the patent cliff in 2009/2012.
Table 20: Patent expiries on the main insulins Product Humalog Novolog Lantus Levemir Apidra
Source: Orange book

Laboratory Lilly Novo Nordisk Sanofi-aventis Novo Nordisk Sanofi-aventis

Launch date 1999 2000 2001 2006 2005

Patent expiry date in the USA June. 2014 Dec. 2014 (formulation 2017) February 2015 2019 2018/2022

Pharmaceuticals

I 46

S EC TOR R EPOR T

The patents on the main compounds are due to fall into the public domain soon (table above). Why are investors so relaxed about this? Because several traditional arguments mean that insulins are seen as not being vulnerable to generic competition. Of course, insulins are assured of a gentle decline, like any biological product, unlike the sudden collapse that follows the arrival of a substitutable generic in the USA. But the markets optimism is based on two other characteristics: the importance of the injector pen and capital intensity. In our view this will limit the cut in prices that may be granted to 15% vs. 30% for a traditional biosimilar.
Capital intensity is higher than for other product categories

Capital intensity is particularly high in this market segment. Generics manufacturers will have to make substantial capital investments. We estimate that a world-class insulin production unit requires investment of around $50/100m. Moreover, Pfizer plans to invest $300m. This is particularly the case as the current producers will capitalise on their market position to develop big units at a low unit price. Novo Nordisk thus spent between 2003 and 2009 $300m for a Brazilian unit and will spend $420m for its Chinese facility between 2008 and 2012. However, this does not prevent smaller producers from developing tools, probably at a much lower cost. We therefore estimate that the additional depreciation that will impact P&L accounts is equivalent to around 5% of sales, limiting accordingly the price reduction than may be granted. Pens: a cutting-edge technology Delivery devices have long been a major criterion of differentiation and a historical strength for Novo Nordisk. Indeed, patients are very sensitive to this aspect, as it helps reduce anxiety over injections.

Delivery systems are an additional entry barrier

The latest generation of delivery systems to arrive on the market is insulin injector pens. They now represent most of the market in Europe. Novo Nordisk long ago launched a rechargeable model, Flexpen, which gave it a lead over rivals. Indeed, this pen was rolled out for Novo Nordisks analogues Levemir and Novolog. Conversely, the take-off of sanofi-aventis Lantus was slowed by a less competitive pen, Opticlick, developed by a Swiss company, Ypsomed. Since then, sanofi-aventis has itself developed a new generation pen, Solostar, which accelerated sales. Today, Solostar represents 32% of Lantus prescriptions in the US (we describe these developments in detail in our May 2010 report Diabetes: the new winners). In any case, the fact that sanofi-aventis experienced difficulty in developing a competitive pen shows that the task will not be that easy for generic manufacturers, though it is a key aspect.

Pharmaceuticals

I 47

S EC TOR R EPOR T

Chart 17 : Insulin pens: a differentiating factor

Share of total insulin volume 100%

80%

60%

40%

Prefilled device
20%

Cartridge for durable Vial and syringe

0%
1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 2006 2007 2008 2009

Source: sanofi-aventis

But insulin biosimilars are already a reality

But the arrival of insulin biosimilars looks inevitable

Indian and Chinese insulin producers are already active. Four or five Indian biosimilar human insulin producers shared the market in 2009 (Wockhardt, Biocon, USV, Lupin), with a market share of around 15%. Novo Nordisk remained the undisputed leader with 59% of the market, followed by Lilly (20%). It seems that this market share was maintained thanks to price cuts to remain competitive. Certain generics manufacturers, such as Wockhardt and Biocon, look to have a lead. Wockhardt has launched a recombinant human insulin, Wosulin, which was developed in pen form. More importantly, Wockhardt carried out acquisitions of companies in Europe, France (Negma), the UK and Germany. It therefore has sizeable marketing capacity in Europe. Similarly, the company is expanding in the USA, and in the long run, is set become a major player in biosimilar insulins. However, Biocon has moved back in front thanks to the agreement with Pfizer (cf. above). Yet the real threat does not concern traditional insulins, whose cheap prices do not necessarily justify massive investments, accounting for the fact that biosimilars are as yet little developed. The situation is different for insulin analogues, which sell for high prices. We believe that biosimilars will increase in number, once the patents on the analogues have fallen. In fact, the race is already on. Wockhardt is the first company to have launched a biosimilar of Lantus, Glaritus, since followed by Biocon with Basalog. So far, the product is only available in India. So, developments are underway on both sides of the Atlantic, once the patents expire.

But with a probably smaller price discount than for traditional biosimilars

In other words, we believe that a widespread launch of biosimilars should be factored into the models as of 2015 for analogues. Conversely, the financial and technological entry ticket will rule out a hefty price discount. In western countries, they could be launched with a discount of around 15% vs. 30% for other biosimilars.

Pharmaceuticals

I 48

S EC TOR R EPOR T

What impact for sanofi-aventis and Novo Nordisk?

A fast-growing market
In a fast-growing market, the new biosimilar competitors will make a place for themselves without causing a fall in sales of existing analogues

A first point setting insulins apart from other markets is their strong growth, which is practically in double digits. The impact of the arrival of biosimilars should accordingly be seen in perspective. Indeed, the European experience (cf. above) shows that penetration of the market in the case of a traditional insulin is very gradual, no more than a few per cent per year. Even assuming that in 2015 this is stepped up, we would have to assume penetration above 10% per year for it to lead to a fall in the sales of the princeps product. In the case of insulins, the specific barriers make it even more difficult. Admittedly, one of the growth levers for the insulin segment, namely the switch from old generations to new generations at higher prices, will in the end fade at the end of the decade, when full substitution has taken place. But the rise in the number of diabetics and growing insulinisation will keep growth of the market at around 7% per year on a sustainable basis. The capture of market share by the biosimilars, initially gradual, should accelerate as of 2020.
Chart 18: Insulin market with and without biosimilars ($m)
50 000 40 000 30 000 20 000 10 000 0 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030

Princeps after biosimilars launch

Source: Natixis

Biosimilars

Negativ e price impact

Even assuming a pessimistic scenario, where the share of biosimilars reaches 40% of the market at constant prices in 2030 (i.e. the current share of traditional insulins, etc.), our baseline hypothesis is therefore a stabilisation in sales of products that lose patent protection. In other words, in 2015 for sanofi-aventiss Lantus and Novo Nordisks Novolog, and in 2019 for Levemir. This is based on the assumption that insulin biosimilars will have smaller price discounts than traditional biosimilars, because of the entry barriers described previously, which will limit the pricing flexibility of generic manufacturers. Novo Nordisks traditional insulins will face biosimilar competition in 2012 according to Pfizer, but sales of these insulins already began to erode gradually some years ago. Biosimilar competition will merely accentuate this trend. In addition, the lions share of sales is generated in emerging markets, where generic competition is already a fact of life in a number of countries. In short, we believe it is prudent to assume erosion of 5% per year as of 2013, with an acceleration in the decline to -10% per year as of 2015. With substitution by analogues adding to the pressure from biosimilars, traditional insulins are set to disappear in western countries by the end of the decade, and will then even see their sales decline in emerging markets. This scenario is backed up by a textbook case, namely sanofi-aventis Insuman. It is, to the best of our knowledge, one of the few traditional insulins whose sales by geographic zone are made public. Sales generated in western Europe are declining (-2.4% in the first nine months of 2010) and those in emerging countries are growing (+20% to 18m).

Pharmaceuticals

I 49

S EC TOR R EPOR T

Sanofi-aventis and especially Novo Nordisk have new generations of products that will ensure their growth in diabetes

This overall analysis does not take into account the effect of new products. By 2015, a new generation of products will have been launched on the market, even if we just stick to the field of insulins. Thus, pharmaceuticals groups have under development: Fixed short-acting GLP1 / insulin combinations. New long-acting super-insulins. New forms of administration for insulins. We described these different approaches in detail in our report Diabetes: the new winners published on 5 May 2010. In this note, we will just give the timetable for the launches of the different products and their sales potential. The advantage is that all these developments will enjoy lengthy patent protection, beyond 2025, and therefore secure the long-term survival of the groups insulin businesses. In the short term, the launch of Novo Nordisks Degludec is the focus of attention. From the point of view of biosimilars, however, this is not the most urgent issue, since Levemir, the groups long-acting insulin analogue, is protected until 2019. However, it will help fuel the groups growth. The first phase III results show a competitive profile relative to Lantus. We expect a launch on the market in 2012. Peak sales potential amounts to DKK5.6bn ($1bn). The following stage will comprise the promising GLP1-insulins fixed combinations. Sanofi-aventis remains the leader in this field, with a Lantus/ lixisenatide combination, which is expected to enter phase III at the start of 2012, and should therefore be on the market in 2014. The advantage of these combinations is that they offer both a basal effect (over the day) thanks to the insulin, and a prandial effect (short-term during at meal) with the GLP1. Lixisenatide looks to have a particularly well suited pharmacokinetic profile. Phase III studies on (short-acting) Lantus /Byetta associations and lixisenatide/ Lantus have shown that glycaemia is strongly improved, with no deterioration in the side effect profile. Novo Nordisk is also developing a fixed combination between Degludec and Victoza. We therefore believe that these fixed combinations will become established for several categories of patients. Novo Nordisk: the long-term future looks assured The Degludec / Victoza combination, which could arrive on the market in 2015, will create an additional market alongside Degludec and Victoza.

Novo Nordisks diabetes franchise will continue to grow until 2022 despite biosimilars

We built a long-term model (2010/2030) for Novo Nordisks diabetes franchise using conservative assumptions for the arrival of biosimilars, i.e.: An acceleration in the decline of sales of traditional insulins as of 2013 and especially as of 2015 (-10% per year). This leads to a virtual disappearance of this activity at the end of the period. Stability in the first few years following the loss of patent protection, then a decline of 5% per year as of 2022 for short-acting analogues. A gradual decline in activities excluding insulins. What is more, we only included in the model the developments currently in phase III, i.e. Degludec (phase III results already published). We assume peak sales for Victoza of DKK16bn ($3bn) in 2021, followed by a gradual decline. This leads to the following sales structure:

Pharmaceuticals

I 50

S EC TOR R EPOR T

Chart 19: Novo Nordisks diabetes franchise sales (DKKm)

100 000 80 000 60 000 40 000 20 000 0 2010 2012 2014 2016 2018 2020 2022 2024 2026 2028 2030 Others

Nov o Rapid
Source: Natixis

Nov oMix

Lev emir

Degludec

Human Insulins

Victoza

Novo Nordisks diabetes franchise should continue to grow until 2022, reaching sales of DKK80bn, almost doubling the 2010 level and showing average annual growth of 6%. Assuming there are no new developments in the period, the franchise would then experience a gradual decline, but in 2030 would still generates sales at the level seen in 2012, a feat that no other group can hope to match (except Roche), excluding the upstream pipeline. What is more, Novo Nordisk has long-term sources of growth that are not included in the model. A once-a-week formulation of Victoza, which could be launched in 2015/16. Oral GLP1s, which could be launched in 2017/18. A long-acting oral insulin, which could be launched in 2017/18. An ultra-short-acting insulin. These developments are still in the upstream phase, but are a reminder that the sales of pharmaceutical groups are, in principle, completely renewed every 20 years and that fears regarding long-term growth are groundless. Sanofi-aventis: a lasting franchise
Sanofi-aventiss diabetes franchise will continue to grow

We believe that, in the case of Lantus, the fixed combination could be substituted for 1.9bn in sales in 2022, and add incremental sales of 1bn. Sales generated in the fixed combination form would therefore amount to 2.9bn in 2022, or 59% of sales in the franchise, which would thus reach a higher peak (6.5bn in 2019 vs. 5.7bn) and maintain sales of nearly 5bn in 2022.

Pharmaceuticals

I 51

S EC TOR R EPOR T

Chart 20: Sales in the Lantus franchise (m)

7 000 6 000 5 000 4 000 3 000 2 000 1 000 0 2007 2008 2009 2010 2011 2012 2013 2014 2015 2016 2017 2018 2019 2020 2021 2022 Lantus - non-substituted part
Source: Natixis

Lantus standalone

Lantus + Combi

Pharmaceuticals

I 52

S EC TOR R EPOR T

6.

Appendix
Table 21: Main biological products by therapeutic class
$m Anticoagulants Lovenox/Clexane (m) AVE-5026 (m) Fraxiparine Fragmin Fragmin Hibor Clivarine Heparin Sodium Heparin Sodium Lowhepa Hirudoid Heparin Heparin Na Lock Enoxaparin sodium Novo-Heparin Fragmin Heparin Sodium Ivor Total anticoagulants Monoclonal antibodies Avastin (CHFm) Rituxan (CHFm) Herceptin (CHFm) Lucentis (CHFm) Xolair (CHFm) Humira Prolia Remicade Remicade Lucentis Actemra Soliris Erbitux Cimzia Simponi Simponi Stelara Remicade Erbitux Tysabri Tysabri Vectibix Xolair Synagis Humira Nimotuzumab Ilaris Actemra Arzerra Synagis Vectibix Campath/MabCampath ReoPro Simulect Removab Orthoclone OKT-3 Cimzia Group Sanofi-Aventis Sanofi-Aventis GlaxoSmithKline Eisai Pfizer Laboratorios Farmacuticos Abbott Laboratories Fresenius Hospira Ajinomoto STADA Arzneimittel Ajinomoto Mitsubishi Tanabe Pharma Hospira Mochida Pharmaceutical Kissei Pharmaceutical Laboratorios Farmacuticos Sigma-Tau 2009 3,043 358 157 222 57 188 27 28 23 26 22 18 13 5 4 5,738 6,222 6,087 5,266 1,198 620 5,488 3,088 431 1,232 44 387 971 105 3 32 33 509 683 509 544 233 338 782 71 3 90 6 260 42 232 110 2 46 2010e 2,643 348 190 211 54 122 40 32 22 27 23 20 13 5 3 4,630 6,550 6,430 5,620 1,475 670 6,568 70 3,214 2,660 1,551 209 531 1,101 270 83 234 279 671 682 607 626 298 376 600 160 22 176 58 260 35 88 218 110 45 46 1 2011e 1,810 351 219 205 58 28 96 46 36 23 27 24 20 13 5 3 3,565 6,600 6,870 6,000 1,630 730 7,398 587 3,076 2,793 1,768 459 690 1,206 451 250 464 471 800 726 700 645 363 436 513 238 14 54 274 118 260 83 105 207 110 58 46 2 2012e 1,755 352 240 199 59 80 74 53 38 25 27 24 11 20 11 5 3 3,558 7,020 7,300 6,400 1,650 770 8,057 1,294 2,990 2,853 1,932 676 836 1,308 651 453 686 580 871 803 745 638 416 489 531 300 21 84 335 165 260 114 106 196 110 65 46 3 2013e 1,701 20 350 255 193 66 84 76 59 40 26 26 24 24 20 11 5 3 3,554 7,470 7,750 6,800 1,650 800 8,575 1,926 2,927 2,867 2,042 960 983 1,362 861 677 865 677 935 857 747 626 462 514 551 357 44 139 393 214 260 141 145 187 110 73 46 4 2014e 1,623 60 346 266 187 89 88 78 65 42 27 26 24 24 20 10 5 3 3,540 7,920 8,050 6,950 1,650 815 8,973 2,527 2,868 2,770 2,107 1,285 1,147 1,406 1,046 836 947 821 987 908 750 623 506 539 562 402 134 230 438 267 260 166 184 179 110 81 46 5 2015e 1,502 80 333 278 181 110 92 80 71 44 28 25 24 24 20 9 5 3 3,434 8,200 8,050 7,000 1,650 835 9,314 3,017 2,755 2,547 2,127 1,551 1,313 1,408 1,198 1,006 1,052 983 1,040 950 760 617 549 560 561 448 300 346 438 346 260 191 225 159 110 89 46 6

Roche Roche Roche Roche Roche Abbott Laboratories Amgen Johnson & Johnson Merck & Co Novartis Roche Alexion Pharmaceuticals Merck KGaA UCB Merck & Co Johnson & Johnson Johnson & Johnson Mitsubishi Tanabe Pharma Bristol-Myers Squibb Elan Biogen Idec Amgen Novartis AstraZeneca Eisai YM BioSciences Novartis Chugai GlaxoSmithKline Abbott Laboratories Takeda Genzyme Eli Lilly Novartis Fresenius Johnson & Johnson Laboratorios Farmacuticos

Pharmaceuticals

I 53

S EC TOR R EPOR T

$m Reditux Total monoclonal antibodies Enzymes Cerezyme Fabrazyme Replagal Vpriv Aldurazyme Replagal Uplyso Total enzymes EPO Aranesp Epogen Mircera Procrit/Eprex Nesp NeoRecormon Epogin Mircera Epoetin alfa BS Retacrit Espo Espogen Total EPO Coagulation factors NovoSeven (DKKm) Advate Kogenate Helixate Humate P Hemofil M ReFacto Xyntha Haemoctin SDH Recombinate Koate-DVI NN7008 Total coagulation factors G-CSF Neulasta Neupogen Leukine Neutrogin/Granocyte Gran Leucomax Nivestim Sargramostim Stemgen G-CSF Total Growth hormones Norditropin SimpleXx (DKKm) Nutropin (CHFm) Genotropin Humatrope Saizen Serostim NutropinAq Tev-Tropin/Tjet ARX201 Rastan Eutropin NN8630 Growtropin SR-rhGH

Group Dr. Reddy's Laboratories

2009 4 35,613 793 430 194 3 155 27 1,601 2,652 2,569 165 2,245 278 965 475 21 149 14 9,532 7,072 1,700 1,238 490 545 291 47 124 358 47 6,272 3,355 1,288 21 349 149 2 5,163 4,401 400 887 449 267 87 56 45 6 29 12 -

2010e 4 42,626 725 187 307 104 170 52 1,545 2,441 2,608 250 2,088 431 937 444 23 33 136 17 9,406 7,950 1,864 1,275 540 592 263 279 108 127 292 55 6,852 3,466 1,265 57 326 175 10 2 1 5,303 4,800 430 908 435 268 94 63 45 23 31 15 -

2011e 4 47,230 1,029 336 357 172 188 56 4 2,143 2,347 2,414 392 1,971 529 898 435 36 40 43 46 18 9,169 8,500 2,006 1,318 575 618 287 276 131 137 274 64 7,241 3,417 1,222 85 296 177 10 3 1 5,210 5,200 450 912 426 274 101 72 50 31 31 17 4

2012e 4 51,791 1,028 433 395 237 204 68 15 2,379 2,265 2,307 576 1,872 517 791 394 81 62 52 39 20 8,976 8,850 2,154 1,376 609 643 304 285 142 147 253 71 7,606 3,452 1,210 125 258 170 10 4 1 1 5,231 5,620 470 917 422 275 99 79 50 41 31 18 8

2013e 4 56,035 1,051 477 436 272 217 80 30 2,563 2,209 2,131 698 1,601 510 697 340 154 72 60 33 22 8,528 9,200 2,304 1,409 644 668 324 285 162 158 236 78 1 7,955 3,490 1,182 155 236 169 10 6 1 1 5,249 6,065 480 912 414 275 99 85 55 8 49 31 20 12

2014e 5 59,535 1,078 518 467 296 231 80 47 2,717 2,152 2,031 819 1,306 504 623 291 190 80 68 32 24 8,119 9,550 2,458 1,416 679 694 344 287 179 170 219 85 9 8,288 3,540 1,135 185 207 163 10 7 1 1 5,249 6,441 490 909 406 278 99 89 60 23 55 30 12 22 16

2015e 5 62,049 1,099 554 491 316 243 80 62 2,845 2,089 1,928 936 1,162 498 545 273 224 88 76 32 25 7,876 9,800 2,613 1,403 714 719 365 290 194 183 201 92 31 8,598 3,503 1,066 215 182 156 10 9 2 1 5,143 6,815 500 902 380 279 99 90 65 45 60 30 26 23 20

Genzyme Genzyme Shire Shire Genzyme Dainippon Sumitomo Pharma Protalix BioTherapeutics

Amgen Amgen Roche Johnson & Johnson Kyowa Hakko Kirin Roche Chugai Chugai Kissei Pharmaceutical Hospira Kyowa Hakko Kirin LG Life Sciences

Novo Nordisk Baxter International Bayer CSL CSL Baxter International Pfizer Pfizer Biotest Baxter International Talecris Biotherapeutics Novo Nordisk

Amgen Amgen Genzyme Chugai Kyowa Hakko Kirin Merck & Co Hospira Hospira Swedish Orphan Biovitrum

Novo Nordisk Roche Pfizer Eli Lilly Merck KGaA Merck KGaA Ipsen Teva Pharmaceutical Merck KGaA Pharmstandard LG Life Sciences Novo Nordisk Dong-A Pharmaceutical Bioton

Pharmaceuticals

I 54

S EC TOR R EPOR T

$m Declage SciTropin Total growth hormones Sex hormones Gonal-F/Gonalef Puregon/Follistim Elonva Follitrope Total sex hormones Immunomodulators Enbrel Enbrel Enbrel Esbriet Pirespa Total immunomodulators Insulins Lantus (m) Apidra (m) Levemir (DKKm) NovoRapid/NovoLog (DKKm) Human insulin & devices (DKKm) NovoMix (DKKm) Humalog Humulin ReliOn Degludec Afrezza DegludecPlus Gensulin Biosulin SciLin Total insulins Alpha interferons Pegasys (CHFm) PEGIntron Intron A Sumiferon Multiferon Aimafix Infergen/Advaferon Total alpha interferons Beta interferons Avonex Rebif Betaferon/Betaseron Extavia Feron Total beta interferons Total biologics market

Group LG Life Sciences Bioton

2009 4 5 3,137 678 97 12 786 3,493 378 348 16 4,236 3,080 137 5,223 9,749 11,315 6,499 1,959 1,022 25 7 14,512 1,655 149 38 63 1 1 1,902 2,323 2,142 1,692 49 45 6,251 94,742

2010e 5 5 3,201 668 530 16 15 1,229 3,500 3,188 450 36 7,174 3,500 175 6,765 11,555 11,710 7,640 2,075 1,051 74 11 3 15,008 1,700 729 211 61 1 1 2,706 2,481 2,183 1,590 145 50 6,449 106,129

2011e 5 5 3,329 685 493 41 16 1,235 3,454 3,306 513 7 47 7,327 3,830 210 7,983 13,058 11,827 8,404 2,204 1,059 34 91 14 4 16,346 1,800 734 203 62 4 1 1 2,800 2,483 2,186 1,517 200 51 6,436 112,032

2012e 5 5 3,450 675 460 78 18 1,231 3,424 3,406 553 21 57 7,461 4,300 250 9,180 14,233 11,827 9,076 2,323 1,070 97 87 16 5 17,776 1,900 748 193 58 5 1 1 2,909 2,432 2,134 1,451 251 51 6,320 118,688

2013e 6 5 3,559 672 427 120 20 1,238 3,387 3,456 590 49 67 7,548 4,850 295 10,374 15,514 11,709 9,802 2,421 1,075 70 175 34 96 19 6 19,431 2,020 771 180 58 7 1 1 3,032 2,332 2,087 1,338 293 51 6,101 124,793

2014e 6 5 3,680 668 401 150 21 1,241 3,333 3,450 620 98 72 7,572 5,430 335 11,618 16,756 11,123 10,489 2,317 1,069 212 244 98 103 23 8 20,907 2,060 791 169 54 9 1 1 3,088 2,206 2,038 1,238 314 51 5,847 129,785

2015e 7 5 3,780 665 365 170 23 1,222 3,256 3,202 643 146 76 7,324 5,700 380 12,896 17,928 10,456 11,118 2,185 1,055 409 308 179 114 26 10 21,982 2,110 796 146 51 10 1 1 3,109 2,106 1,952 1,132 332 51 5,574 132,935

Merck KGaA Merck & Co Merck & Co LG Life Sciences

Amgen Pfizer Takeda InterMune Shionogi

Sanofi-Aventis Sanofi-Aventis Novo Nordisk Novo Nordisk Novo Nordisk Novo Nordisk Eli Lilly Eli Lilly Novo Nordisk MannKind Novo Nordisk Bioton Pharmstandard Bayer

Roche Merck & Co Merck & Co Dainippon Sumitomo Pharma Swedish Orphan Biovitrum Pharmstandard Astellas Pharma

Biogen Idec Merck KGaA Bayer Novartis Daiichi Sankyo

Pharmaceuticals

I 55

Table 22: Clinical and pharmacovigilance trials recommended in European biosimilar guidelines
Product Insulin (adopted 22/02/2006) Pharmacokinetic Crossover study, single dose, subcutaneous injection, preferably type I diabetics. Pharmacodynamic Double blind euglycemic hyperinsulinemic clamp study. Provide comparative data for speed of glucose IV and seric insulin concentration. Crossover study, single dose, subcutaneous and intravenous injections. Comparative trials on several doses in healthy volunteers: Absolute neutrophils count and CD34 cell count. Trial comparing biosimilar and reference product in prophylaxis and severe neutropeny in patients treated with cytotoxic chemotherapy. The patient group selected must be homogenous (same type of tumour, disease stage, planned and previous treatments). This trial has two arms where the duration and frequency of neutropeny caused by chemotherapy is known, otherwise a three-arms trial is required. Measurement of duration of severe neutropeny and secondly, the occurrence of febrile neutropeny, infections and total doses of G-CSF. EPO (adopted 22/03/2006) Crossover study, single dose, subcutaneous and intravenous injections in healthy volunteers. Reticulocytes count At least two parallel randomised comparative trials between the biosimilar and reference product in anaemic patients due to renal insufficiency (patients most sensitive to EPO). First study: a 6-month correction phase. Measure of response rate (% of patients reaching target) and change in haemoglobin. Second study: a 6-month maintenance phase. Measure of maintenance rate (% of patients remaining in target without transfusion) or change in haemoglobin. Comparative immunogenicity trials of at least 12 months to be presented before approval. Comparative safety data from efficacy trials are enough before commercial launch. Efficacy No need if clinical comparability was established by pharmacokinetic and pharmacodynamic trials. Safety Immunogenicity trial of at least 12 months, including 6 months of comparative phase to be submitted before marketing approval. The 12month results can be submitted after launch. Follow-up data for 6 months of range of patients in comparative clinical trials. Include enough patients to assess side effect profile (particularly bone pains and anomalies diagnosed in the laboratory) Pharmacovigilance Detailed post-marketing monitoring and risk management programme factoring in the risks identified during development and risks of immunogenicity.

G-CSF (adopted 22/02/2006)

Risk management programme, pharmacovigilance plan. Immunogenicity and rare severe side effects must be monitored (particularly in chronic treatment). The absence of efficacy must be observed (particularly in patients whose haematopoietic cell cultures are already mobilised).

Risk management programme, pharmacovigilance plan, monitoring of safety data (particularly rare side and serious effects such as pure red cell aplasia: PRCA).

Product Growth hormone (adopted 22/02/2006)

Pharmacokinetic Crossover study, single dose, subcutaneous injections in healthy volunteers after suppression of endogenous production of growth hormone with a somatostatin analogue.

Pharmacodynamic Measure of somatotropin activity markers: IGF-1 or others such as IGFBP-3.

Efficacy At least one parallel randomised comparative double blind trial in children that have a growth hormone deficiency and have never been treated (most sensitive population). Measure of speed of growth and standard deviation score. Growth rate before treatment must be measured over a period of 6-18 months, the comparative trial must last at least 6 months.

Safety Data obtained in efficacy trials are usually enough before launch. Comparison of immunogenicity over 12 months and blood dosages of IGF1, IGFBP-3, insulin and glucose.

Pharmacovigilance Risk management programme, pharmacovigilance plan factoring in risks identified during development and immunogenicity risks.

Monoclonal antibodies (issued 19/12/2010)

Comparative and multi-dose studies with a sufficient and homogenous sample of patients. Dose analyses responses requested. Criteria: concentration of the product (AUC, Cmax, etc.) elimination of the molecule from the organism and half life. For cancer treatments: a monotherapy test is allowed, to minimise the sources of variability between the biosimilar and its reference product. A specific study targeting the adjuvant treatment is advised.

Possibility of combining the pharmacodynamic studies with pharmacokinetics Dose-response study vital for all indications targeted. With these pharmacokinetics/dynamics studies, proof of biosimilarity may be established.

Comparative randomised, double blind trials. Criterion of non-inferiority to princeps. To demonstrate biosimilarity, the criteria used (study targets, duration, etc..) may differ from traditional guidelines. The aim is to demonstrate similarity with the princeps product, and not the action on patients. For cancer treatments, the response rate or time to progression will be accepted. Possible to extrapolate the demonstration of biosimilarity to other indications of the princeps product

Verification of safety comparable to the original product. Immunogeneticity studies requested.

Pharmacovigilance study requested, comparable to that used by the inventor of the molecule.

Sources: Natixis, EMEA

S EC TOR R EPOR T

7.

Company profile
Novartis
Our top pick

59
59

Novo Nordisk
Above the fray

63
63

Roche
Return to grace unlikely before 2011

67
67

Stada
The recovery is on track

71
71

Pharmacie

I 58

S EC TOR R EPOR T

6 December 2010
161. 7 141. 7 121. 7 101. 7 81. 7 61. 7 dc -07 j ui n-08 j anv -09 j ui l -09 f vr -10 s ept -10

Pharmaceuticals

Switzerland

Novartis
NOVZn.S / NOVN@VX
Novar t is Rel. St oxx 50

Buy

Our top pick

Management opted for a business diversification model at the start of the decade. This consisted of building up a franchise of vaccines, generic drugs (the worlds no.2 with Sandoz) and ophthalmologic products (after taking over Alcon), and has led to a reduction in the risk usually associated with any pharmaceutical stock. Sandoz is an undisputable growth lever (operating profit CAGR of +9% from 2010 to 2015e): its market stands to grow twice as fast as the pharma market, and Sandoz is investing in fields that will generate higher operating margins than conventional generic drugs. Value-added generics include biosimilars (a field in which the group is a pioneer and, for the time being, the market no.1), respiratory inhalers (developed with Oriel) and injectable drugs (Ebewe was taken over in 2009). Sandoz is thus a top-quality asset which, by our estimates, accounts for 11% of the groups restated value. The Pharma division, accounting for 70% of operating profit, is delivering excellent sales performances (7% organic growth for the first 9 months) and a continuously expanding operating margin (+90bp at end-September to 33.9%). This performance is being fuelled by an innovation effect (new products account for 21% of Pharma sales), which is not expected to run out of steam. For instance, Novartis is marketing a new oral MS drug, Gilenya, and we estimate its peak sales potential at $2.6bn. The ramp-up of new products will help the group weather the difficult 2011/12 period during which the group will lose patent protection on three of its major drugs (Diovan, Femara, Zometa), accounting for 29% of Pharma sales. The 2011/2015 period is set to show 1.7% average annual sales growth for this division. Last of all, the takeover of Alcon is a strategic operation because it bolsters Novartis position in a highly-profitable and relatively generic-proof segment, ophthalmology. The takeover of Nestles 52% stake will, moreover, have an accretive impact of 7% in 2012, which will prevent Novartis EPS from shrinking, even a little, in 2012/13. The last stage of the process (acquisition of the 23% stake held by the public) could also have a significantly accretive impact if management uses a different financing method (a cash acquisition rather than a rights issue). We reiterate our Buy rating on the stock. Novartis is our top pick in the European pharma sector based on its capacity to develop its activities, optimise its costs and manage its assets.
Close on 31/12 2009e 2010e 2011e 2012e Turnover ($m) 44,267 50,705 57,214 59,528 Reported net profit ($m) 8,400 10,003 9,848 10,628 Adjusted EPS (CHF) 4.90 5.48 5.78 6.04 Chg. EPS (%) 8.6 12.0 5.4 4.6 PE (x) 9.7 9.2 8.8 EV/EBIT (x) 10.4 8.5 8.0 P/CF (x) 8.3 6.6 6.5 Net yield (%) 4.2 4.6 4.7

Source: Natixis

Price Target Upside

12/03/2010

CHF53.45 CHF70.00 31.0%

12m -4.9% 8.1% 2.3% 1 Jan -5.4% 3.7% -1.0%

Performance Absolute Sector Stoxx 50 12-month high/low SMI Stoxx 50 Market capitalisation Free float Daily volume

1m -6.4% 1.0% 0.3%

CHF60.25 / CHF50.55 6440.9 2553.8 CHF121.1bn 100.0% CHF352m

Analyst(s) Batrice Muzard (33 1) 58 55 05 13 beatrice.muzard@natixis.com Philippe Lanone (33 1) 58 55 05 03 philippe.lanone@natixis.com Equity Markets Bloomberg access equity.natixis.com NXSE

Distribution of this report in the United States. See important disclosures at the end of this report.

Pharmacie

I 59

S EC TOR R EPOR T

Company profile
Novartis was born of the merger between Swiss groups Ciba and Sandoz in 1996. In 2009, the pharma division (64% of consolidated sales) generated sales of $28.5bn. Novartis is also involved in the vaccines business following the Chiron acquisition (April 2006), and has become a significant player in flu vaccines. The groups generics business (Sandoz), which has expanded significantly in recent years via the acquisition of local companies and organic growth, ranks no. 2 in the world behind Teva-Ivax. Novartis generics arm makes the group a unique case in our pharma universe. Consumer Health encompasses ophthalmology (CIBA Vision), strengthened by the acquisition of a 77% stake in Alcon, animal health and OTC activities (reinforced by the acquisition of BMS businesses).

Recent events
Q3 10 results were excellent, with an EPS up 16%. Novartis continues to benefit from powerful innovation: new products represented 20% of Q3 10 sales. The group consolidated 100% of Alcons earnings starting in September 2010. Novartis received authorisation for Gilenya in the USA for relapsing remitting multiple sclerosis.

Turnover 10
Alcon Consumer Health 5% 12%

Adjusted operating profit 10 ($m)

10 076

Vaccines & diagnostic 5% Pharmaceutic als 62%

Generic 16%

894

1 311

1 669

1 180

Alcon

-745 Consumer Corporate Generics health

Pharmaceuticals

Vaccines

Opportunities / Strengths
New drugs are ramping up (Exforge and Lucentis

Weaknesses / Threats
The

already have blockbuster status) and accounted for 21% of Pharma sales in Q3 10. Renowned development capacity in the fields of oncology (Tasigna, Afinitor), neurology (Gilenya in MS) and respiratory diseases (QAB 149, QVA 149). Major restructuring efforts ($1.830bn in total, paying off in 2010). Major exposure to emerging countries (24% of consolidated sales).

Pharma division's earnings are heavily dependent on Diovan whose patent expires in September 2012 in the USA (worldwide sales of $6.2bn in 2010; operating margin 50%; 25% of the group's adjusted operating income). Diovan's momentum could be curbed in the short term in the USA by the arrival in 2010 of generic versions of Cozaar, a rival anti-hypertension drug from Merck and whose patent expires in the USA in 2010. Positions need to be bolstered in vaccines (only two major product lines) and animal health (weak global position). Uncertainty about the acquisition of the remaining 23% stake in Alcon: the operation is being blocked by Alcon's management and the conditions for buying the 23% are far from optimal (rights issue, merger in compliance with Swiss law).

Share ownership
Free float Shares Votes 100.0% 100.0%

Diary dates
01/27/2011 04/19/2011 07/19/2011 10/25/2011 2010 earnings Q1 11 earnings Q2 11 earnings Q3 11 earnings

Pharmacie

I 60

S EC TOR R EPOR T

Financial Data 31/12

Breakdown by activity ($m) Turnover Pharmaceuticals Generic Consumer Health Vaccines & diagnostics Alcon Adjusted operating profit Pharmaceuticals Generic Consumer Health Vaccines & diagnostics Alcon Corporate Adjusted operating margin Pharmaceuticals Generic Consumer Health Vaccines & diagnostics Alcon Profit & loss statement ($m) Turnover Change Organic growth EBITDA Change EBIT Change Adjusted EBIT Change Operating margin Financial items Pre-tax profit on ordinary activities Exceptional items Corporate tax Goodwill amortisation/ impairment Equity associates Minority interests Net profit on divested activities Reported net profit Change Adjusted net profit Change Cash flow statement ($m) Cash flow from operations Nets Investments Decrease (Increase) in WCR Free cash flow Acquisitions Dividend Capital increase Divestments Miscellaneous Increase (Decrease) in cash Net debt Gearing 2008 41,459 26,331 7,557 5,812 1,759 0 10,309 8,214 1,450 1,121 309 0 -785 24.9% 31.2% 19.2% 19.3% 17.6% 2008 41,459 8.9% 11,704 20.0% 8,964 21.6% 10,309 11.0% 24.9% 94 9,058 0 -1,336 0 441 -38 70 8,195 21.3% 9,431 13.4% 2008 11,129 -2,106 -1,360 7,663 -8,261 -3,345 -473 0 -2,339 -6,755 -652 -1.3% 2009e 44,267 28,538 7,493 5,812 2,424 0 11,437 9,068 1,395 1,118 719 0 -863 25.8% 31.8% 18.6% 19.2% 29.7% 2009e 44,267 6.8% 12,283 4.9% 9,982 11.4% 11,437 10.9% 25.8% -353 9,629 0 -1,468 0 293 -54 0 8,400 2.5% 10,213 8.3% 2009e 12,238 -1,887 -47 10,304 -2,009 -3,941 224 226 -1,995 2,809 -3,461 -6.0% 2010e 50,705 30,900 8,350 6,275 2,780 2,400 14,385 10,076 1,669 1,311 1,180 894 -745 28.4% 32.6% 20.0% 20.9% 42.4% 37.3% 2010e 50,705 14.5% 14,845 20.9% 12,274 23.0% 13,895 21.5% 27.4% -675 11,599 0 -2,099 0 749 -246 0 10,003 19.1% 11,845 16.0% 2010e 13,951 -2,400 -1,400 10,151 -29,650 -4,486 0 0 0 -23,985 20,524 27.9% 2011e 57,214 32,200 9,250 6,350 2,200 7,214 16,282 10,450 1,880 1,292 560 2,800 -700 28.5% 32.5% 20.3% 20.3% 25.5% 38.8% 2011e 57,214 12.8% 16,951 14.2% 12,980 5.8% 16,282 17.2% 28.5% -975 12,005 0 -2,026 1 503 -635 0 9,848 -1.6% 12,486 5.4% 2011e 17,448 -2,870 -1,600 12,978 -800 -5,148 0 0 0 7,030 13,493 17.2% 2012e 59,528 32,650 10,100 6,604 2,500 7,674 16,405 9,870 2,050 1,375 635 3,200 -725 27.6% 30.2% 20.3% 20.8% 25.4% 41.7% 2012e 59,528 4.0% 17,480 3.1% 13,380 3.1% 16,405 0.8% 27.6% -525 12,855 0 -2,109 2 538 -658 0 10,628 7.9% 13,059 4.6% 2012e 17,700 -3,000 -1,750 12,950 -1,500 -5,606 0 0 0 5,844 7,649 9.1%

Novartis
CAGR 09/12 10.4% 4.6% 10.5% 4.4% 1.0% 12.8% 2.9% 13.7% 7.1% -4.1% 5.6%

CAGR 09/12 10.4%

12.5% 10.3% 12.8%

10.1%

8.2% 8.5%

CAGR 09/12 13.1% 16.7% 7.9% 12.5%

Pharmacie

I 61

S EC TOR R EPOR T

Balance sheet ($m) Net fixed assets o/w net goodwill o/w gross goodwill Financial fixed assets WCR Net assets on divested activities Total equity o/w shareholders' equity Provisions Net debt Per share data (CHF) Shares outstanding (millions) Diluted shares (millions) Reported EPS Adjusted EPS Goodwill Cash flow Net dividend Payout ratio Book value Financial ratios Personnel expenses (M) Personnel expenses (% of sales) Operating margin Adjusted operating margin Effective rate of tax Net margin ROE ROCE Capital employed (M) Interest cover (x) Debt/EBITDA (x) Gearing WCR (% of sales) Goodwill (% of book value) Net investments (% of sales) Investment ratios EV/Turnover (x) EV/EBITDA (x) EV/Adjusted EBIT (x) P/BV (x) P/CF (x) Adjusted P/E Reported P/E Net yield Free cash flow yield (%)

2008 33,919 1,175 2,100 23,499 1,547 50,437 50,307 9,180 -652 2008 2,265.7 2,265.5 3.92 4.51 0.00 5.32 2.00 51.1% 24.05 2008 8,200 19.8% 21.6% 24.9% -14.7% 18.8% 16.3% 22.1% 36,391 na -0.1 -1.3% 3.7% 2.3% -5.1%

2009e 36,445 1,175 2,100 25,369 2,085 57,462 57,332 9,898 -3,461 2009e 2,265.7 2,267.9 4.03 4.90 0.00 5.87 2.10 52.2% 27.48 2009e 8,200 18.5% 22.5% 25.8% -15.2% 18.4% 14.7% 22.6% 39,455 32.4 -0.3 -6.0% 4.7% 2.0% -4.3%

2010e 81,200 1,175 2,100 15,744 9,237 73,658 73,458 12,000 20,524 2010e 2,265.7 2,288.0 4.63 5.48 0.00 6.46 2.25 48.6% 34.00 2010e 8,400 16.6% 24.2% 27.4% -18.1% 18.7% 13.6% 11.9% 91,362 20.6 1.4 27.9% 18.2% 1.6% -4.7% 2010e 2.9 9.7 10.4 1.6 8.3 9.7 11.5 4.2% 8.2%

2011e 72,000 1,175 2,100 15,744 16,024 78,274 78,054 12,000 13,493 2011e 2,265.7 2,288.0 4.56 5.78 0.00 8.08 2.45 53.8% 36.13 2011e 8,400 14.7% 22.7% 28.5% -16.9% 17.4% 12.6% 14.3% 88,949 16.7 0.8 17.2% 28.0% 1.5% -5.0% 2011e 2.4 8.1 8.5 1.5 6.6 9.2 11.7 4.6% 10.5%

2012e 75,000 1,176 2,101 15,744 12,512 83,607 83,387 12,000 7,649 2012e 2,265.7 2,288.0 4.92 6.04 0.00 8.19 2.50 50.8% 38.60 2012e 8,401 14.1% 22.5% 27.6% -16.4% 18.1% 12.7% 14.5% 88,437 31.2 0.4 9.1% 21.0% 1.4% -5.0% 2012e 2.2 7.5 8.0 1.4 6.5 8.8 10.9 4.7% 10.4%

CAGR 09/12 27.2%

13.3%

CAGR 09/12 0.0% 0.3% 6.9% 7.3% 11.8% 6.0% 12.0% CAGR 09/12 0.8%

30.9%

Pharmacie

I 62

S EC TOR R EPOR T

6 December 2010

Pharmaceuticals
221. 4 171. 4

Denmark

Novo Nordisk
NOVOb.CO / NOVOB@DC
j ui n-08 j anv -09 aot -09 f v r -10 sept -10

Buy

121. 4

71. 4 dc -07

Novo Nordisk Rel. St oxx 50

Above the fray

DKK574.00 DKK600.00 4.5%
12m 70.7% 8.1% 2.3% 1 Jan 72.9% 3.7% -1.0%

Source: Natixis

Price Target Upside

12/03/2010

Performance Absolute Sector Stoxx 50 12-month high/low KFX Stoxx 50 Market capitalisation Free float Daily volume

1m 3.9% 1.0% 0.3%

The announcement of the Pfizer/Biocon agreement in insulin biosimilars came as a bombshell, causing Novo Nordisk to lose 10% in two trading sessions. Once it got over the shock, the market calmed down. It is true that, put brutally, this is a reminder that around 50% of the groups diabetes franchise will face biosimilar competition between now and 2015. And that we therefore cannot expect growth for these products beyond that date. The question is therefore whether the stocks 80% premium to the sector is justified. In reality, even factoring in a virtual disappearance in the long term of traditional human insulins (26% of 2010e diabetes sales), and stability, even a fall, in short-acting analogues and premixes (42%), the progression of Victoza will enable the diabetes franchise to continue to show growth until 2022. Sales should double between 2010 and 2022, before falling back to the 2012 level in 2030. Thereafter, a decline is inevitable for the current products, but Novo has a pipeline that holds out the hope of further growth thereafter. Thus the groups growth profile remains much better and more assured than that of any other pharmaceuticals group, and its PE 12e of 18x does not factor in any more than the others the upstream pipeline, as it can be justified by the current generation of products. In short, we are convinced that the current PE is not exaggerated, as it correctly reflects the bright long-term prospects. The anomaly lies instead in the valuations of the other pharmaceuticals groups. On top of this there is the strong short-term momentum, since conversely, the group will continue for two or three years to post strong growth, a unique quality in the sector. We therefore maintain our Buy rating to factor in this scarcity effect.

DKK582.00 / DKK331.75 437.4 2553.8 DKK352.6bn 100.0% DKK272m

Analyst(s) Philippe Lanone (33 1) 58 55 05 03 philippe.lanone@natixis.com Batrice Muzard (33 1) 58 55 05 13 beatrice.muzard@natixis.com Equity Markets Bloomberg access equity.natixis.com NXSE

Distribution of this report in the United States. See important disclosures at the end of this report.

Close on 31/12 2009 2010e 2011e 2012e

Turnover (DKKm) 51,078 59,970 66,320 72,478

Reported net profit (DKKm) 10,768 13,630 16,517 18,802

Adjusted EPS (DKK) 17.97 23.40 28.48 32.41

Chg. EPS (%) 14.7 30.2 21.7 13.8

PE (x) 24.5 20.2 17.7

EV/EBIT (x) 17.9 15.4 13.5

P/CF (x) 20.9 17.5 15.5

Net yield (%) 1.6 1.8 2.2

Pharmacie

I 63

S EC TOR R EPOR T

Company profile
Novo Nordisk generates 73% of sales and 57% of its operating profit in diabetes treatments. The company is a world leader in insulins, its traditional area of expertise and a segment where it is world leader. The group achieved biopharma status via NovoSeven (coagulation factor), its second growth driver, which accounted for 13% of sales in 2009. The group also sells growth hormones (9% of sales) and hormone replacement treatments (3%).

Recent events
Very impressive Q3 2010 results from Novo. Revenue was DKK15,584m, +14%, vs. DKK15,015m forecast. Victoza sales were almost double the consensus forecast, and the guidance for 2010 has been revised up. Once again, Novo has raised its full-year guidance, and is now looking for growth of 11-12% in the top line at constant exchange rates, vs. +9/10% previously, and growth of over 15% in operating profit vs. 12/15% in August 2010. July and October 2010: development failures for the two main rivals to Novo's Victoza, Roche's taspoglutide and Lilly's Bydureon.

Revenues 09
Biopharma ceuticals 27%

Operating profit 09

Diabetes 57% Diabetes 73%

Biopharma ceuticals 43%

Opportunities / Strengths
Benefits from substitution of insulin analogues, a

Weaknesses / Threats
Potential biosimilar competition from 2012 for

fast-growing segment. With sanofi-aventis, Novo is the only pharma company to offer a full range of analogues (shortacting NovoRapid, long-acting Levemir). Strong exposure to emerging markets (China fifthlargest market for the group). The delay to the entry on the market of Bydureon (Lilly) will ease penetration for Victoza. Novo is the only company whose diabetes pipeline covers the entire spectrum of future therapies (ultra-basal insulins, long acting GLP-1, combo basal insulin/GLP-1, oral GLP-1). Shares Votes 100.0% 100.0%

human insulins, and in 2015 for analogues, as well as the NovoSeven patent expiry means uncertainty regarding long-term prospects. The group is highly exposed to the dollar. Price environment becoming tougher in the US with the US healthcare reform.

Share ownership
Free float

Diary dates
12/15/2010 02/02/2011 04/28/2011 08/04/2011 10/27/2011 Investors day 2010 earnings Q1 11 earnings Q2 11 earnings Q3 11 earnings

Pharmacie

I 64

S EC TOR R EPOR T

Financial Data 31/12

Breakdown by activity (DKKm) Turnover Diabetes Bio-pharmaceuticals Reported operating profit Diabetes Bio-pharmaceuticals Reported operating margin Diabetes Bio-pharmaceuticals Profit & loss statement (DKKm) Turnover Change Organic growth EBITDA Change EBIT Change Adjusted EBIT Change Operating margin Financial items Pre-tax profit on ordinary activities Exceptional items Corporate tax Goodwill amortisation/ impairment Equity associates Minority interests Net profit on divested activities Reported net profit Change Adjusted net profit Change Cash flow statement (DKKm) Cash flow from operations Nets Investments Decrease (Increase) in WCR Free cash flow Acquisitions Dividend Capital increase Divestments Miscellaneous Increase (Decrease) in cash Net debt Gearing 2008 45,553 33,356 12,197 12,373 7,373 5,000 27.2% 22.1% 41.0% 2008 45,553 8.9% 14,815 24.0% 12,373 38.4% 12,698 27.7% 27.9% 322 12,695 0 -3,050 0 -124 0 9,645 13.2% 9,645 37.4% 2008 13,425 -1,772 -562 11,091 0 -2,795 -4,717 0 0 3,892 -7,844 -23.8% 2009 51,078 37,502 13,576 14,933 8,510 6,423 29.2% 22.7% 47.3% 2009 51,078 12.1% 17,484 18.0% 14,933 20.7% 14,933 17.6% 29.2% -945 13,988 0 -3,220 0 -55 0 10,768 11.6% 10,768 11.6% 2009 13,319 -3,000 -613 9,706 0 -3,724 -6,500 0 4,112 3,594 -11,438 -32.0% 2010e 59,970 45,000 14,970 19,100 11,330 7,770 31.8% 25.2% 51.9% 2010e 59,970 17.4% 21,730 24.3% 19,100 27.9% 19,100 27.9% 31.8% -1,400 17,700 0 -4,070 0 40 0 13,630 26.6% 13,630 26.6% 2010e 16,260 -3,500 -600 12,160 0 -4,532 -7,500 0 0 128 -11,566 -32.3% 2011e 66,320 50,380 15,940 21,591 12,200 9,391 32.6% 24.2% 58.9% 2011e 66,320 10.6% 24,500 12.7% 21,591 13.0% 21,591 13.0% 32.6% 0 21,591 0 -5,074 0 50 0 16,517 21.2% 16,517 21.2% 2011e 19,426 -3,850 -800 14,776 0 -5,337 0 0 0 9,439 -21,005 -58.7%

Novo Nordisk
2012e 72,477 55,737 16,740 23,724 14,120 9,604 32.7% 25.3% 57.4% 2012e 72,478 9.3% 26,903 9.8% 23,724 9.9% 23,724 9.9% 32.7% 850 24,574 1 -5,775 1 70 1 18,802 13.8% 18,802 13.8% 2012e 21,978 -4,100 -900 16,978 0 -6,227 0 0 0 10,751 -31,756 -88.6% CAGR 09/12 12.4% 14.1% 7.2% 16.7% 18.4% 14.4%

CAGR 09/12 12.4%

15.4% 16.7% 16.7%

20.7%

20.4% 20.4%

CAGR 09/12 18.2% 11.0% 20.5% 18.7%

Pharmacie

I 65

S EC TOR R EPOR T

Balance sheet (DKKm) Net fixed assets o/w net goodwill o/w gross goodwill Financial fixed assets WCR Net assets on divested activities Total equity o/w shareholders' equity Provisions Net debt Per share data (DKK) Shares outstanding (millions) Diluted shares (millions) Reported EPS Adjusted EPS Goodwill Cash flow Net dividend Payout ratio Book value Financial ratios Personnel expenses (M) Personnel expenses (% of sales) Operating margin Adjusted operating margin Effective rate of tax Net margin ROE ROCE Capital employed (M) Interest cover (x) Debt/EBITDA (x) Gearing WCR (% of sales) Goodwill (% of book value) Net investments (% of sales) Investment ratios EV/Turnover (x) EV/EBITDA (x) EV/Adjusted EBIT (x) P/BV (x) P/CF (x) Adjusted P/E Reported P/E Net yield Free cash flow yield (%)

2008 21,539 788 2,112 5,170 32,979 32,979 3,686 -7,844 2008 614.2 620.7 15.54 15.66 0.00 19.47 6.00 38.6% 53.13 2008 14,615 32.1% 27.2% 27.9% -24.0% 21.4% 29.2% 36.1% 26,709 na -0.5 -23.8% 11.3% 2.4% -3.9%

2009 22,076 1,037 1,813 5,030 35,734 35,734 4,623 -11,438 2009 614.2 604.3 17.82 17.97 0.00 22.04 7.50 42.1% 54.14 2009 16,113 31.5% 29.2% 29.2% -23.0% 21.2% 30.1% 41.9% 27,106 15.8 -0.7 -32.0% 9.8% 2.9% -5.9%

2010e 23,940 1,037 1,813 4,054 35,762 35,762 5,611 -11,566 2010e 614.2 593.0 22.98 23.40 0.00 27.42 9.00 39.2% 60.31 2010e 17,587 29.3% 31.8% 31.8% -23.0% 22.7% 38.1% 51.9% 27,994 13.6 -0.5 -32.3% 6.8% 2.9% -5.8% 2010e 5.7 15.7 17.9 9.5 20.9 24.5 25.0 1.6% 3.4%

2011e 26,285 1,037 1,813 -6,492 35,796 35,796 6,814 -21,005 2011e 614.2 593.0 27.85 28.48 0.00 32.76 10.50 37.7% 60.36 2011e 19,020 28.7% 32.6% 32.6% -23.5% 24.8% 46.1% 82.9% 19,792 na -0.9 -58.7% -9.8% 2.9% -5.8% 2011e 5.0 13.5 15.4 9.5 17.5 20.2 20.6 1.8% 4.2%

2012e 28,985 1,037 1,813 -18,757 35,838 35,838 7,959 -31,756 2012e 614.2 593.0 31.70 32.41 0.00 37.06 12.50 39.4% 60.43 2012e 20,570 28.4% 32.7% 32.7% -23.5% 25.8% 52.5% 176.3% 10,228 na -1.2 -88.6% -25.9% 2.9% -5.7% 2012e 4.4 11.9 13.5 9.5 15.5 17.7 18.1 2.2% 4.8%

CAGR 09/12 9.5%

0.1%

CAGR 09/12 0.0% -0.6% 21.2% 21.7% 18.9% 18.6% 3.7% CAGR 09/12 8.5%

-27.7%

Pharmacie

I 66

S EC TOR R EPOR T

6 December 2010
162. 3 142. 3 122. 3 102. 3 82. 3 62. 3 dc -07 j ui n-08 j anv -09 j ui l -09 f vr -10 s ept -10

Pharmaceuticals

Switzerland

Roche
ROCZg.S / ROG@VX
Roche Rel. St oxx 50

Neutral

Return to grace unlikely before 2011

Roche had a rough ride this summer, what with taspoglutides virtual failure, Avastins negative recommendation in breast cancer, delays for T-DM1 and lacklustre sales. This in turn sharpened market concerns surrounding the emergence of biosimilars of the groups blockbusters, a threat discernable in the new US law. We have drawn up earnings scenarios with and without biosimilars for monoclonal antibodies, albeit without the worst case scenario of substitutable generics. While the impact is significant in terms of the valuation, with around 8.4%, mainly for Herceptin, Avastin, Mabthera and Lucentis, it is mitigated by the fact that there are no patent expiries in the short term. Markets-wise, the pressure has gone far beyond this over the last 12 months, with the share even trading below the sector average. Roches medium/long-term growth prospects are among the best in the sector. Despite the recent setbacks, the group will benefit from recent launches (Actemra) and boasts a very healthy pipeline of phase III developments, in oncology (T-DM1, braf), but also in metabolism disorders with dalcetrapib (dyslipidemia), for which the first phase III results will be announced in Q3 11. Roche is due to provide an overview of its late-stage developments at a meeting on 9 December. After the current rough patch, we see it continuing to deliver stronger growth than its peers. 2010 was a tough year for Roche, as it undertook restructuring efforts on a scale not seen at the group in a decade. It expects these initiatives to generate savings of CHF2.4bn in 2012, with a reduction in staff of 4,800. The earnings guidance for 2011 is likely to be cautious, a factor that has put the share under pressure in past years. All told and despite the clear undervaluation, the market upswing is unlikely to show through until there is confirmation regarding Avastin and possibly not until the group reports its 2010 results. In the meantime, we are maintaining our Neutral rating.

Source: Natixis

Price Target Upside

12/03/2010

CHF138.40 CHF150.00 8.4%

12m 8.1% 2.3% 1 Jan 3.7% -1.0%

Performance Absolute Sector Stoxx 50 12-month high/low SMI Stoxx 50 Market capitalisation Free float Hoffmann families Daily volume

1m -4.6% 1.0% 0.3%

-17.5% -21.3%

CHF186.00 / CHF130.20 6440.9 2553.8 CHF119.4bn 84.5% 9.3% CHF189m

Analyst(s) Philippe Lanone (33 1) 58 55 05 03 philippe.lanone@natixis.com Batrice Muzard (33 1) 58 55 05 13 beatrice.muzard@natixis.com Equity Markets Bloomberg access equity.natixis.com NXSE

Distribution of this report in the United States. See important disclosures at the end of this report.

Close on 31/12 2009 2010e 2011e 2012e

Turnover (CHFm) 49,051 48,630 51,080 53,950

Reported net profit (CHFm) 7,784 10,640 11,955 12,980

Adjusted EPS (CHF) 12.18 13.13 14.66 15.85

Chg. EPS (%) 8.7 7.8 11.7 8.1

PE (x) 10.5 9.4 8.7

EV/EBIT (x) 8.7 7.7 6.9

P/CF (x) 8.0 7.4 6.9

Net yield (%) 4.6 5.1 5.4

Pharmacie

I 67

S EC TOR R EPOR T

Company profile
Roche is the worlds 5th-largest drug company. After the disposal in 2000 of its aromas and fragrances business, followed by its vitamins activity in 2003, Roche is now fully focused on human heath with 80% of its revenues in pharmaceuticals. Roche is the world leader in oncology, which accounted for 50% of its pharma sales in 2007. The group is also the world leader in diagnostics ahead of Abbott. Novartis holds 33% of Roche's shares with voting rights (6.1% of capital).

Recent events
October 2010: The progression in pharma sales accelerated slightly in Q3 10 to +4% excluding Tamiflu and currencies. October 2010: good phase II data for T-DM1 and MetMab, both from the Genentech pipeline September 2010 : another potential setback for Roche. After the threat of Avastin being withdrawn in the USA in breast cancer, it is now Europes turn to become worried. June 2010: FDA reqests more data on taspoglutide (diabetes) March 2010: Success of Avastin in ovarian cancer, but failure in prostate cancer March 2010: decision to suspend the development of ocrelizumab in RA.

2009 sales by geographic area

2009 sales by division

Europe 37%

North America 37%

Pharmaceuticals 80%

Diagnostics 20%

Others 26%

Opportunities / Strengths
A series of anti-cancer drugs (Avastin, Mabthera,

Weaknesses / Threats
The shareholding structure (locked up by the

Herceptin, Tarceva) should ensure sales and earnings growth until 2014. Fresh growth will come from a second wave of launches, with especially Actemra in rheumatoid arthritis. Several long-term strengths in the field of metabolic diseases (cholesterol) will strengthen long-term growth. More than half of pharma sales are generated with biologics, which will not face direct generic competition in the short term.

families) prevents share buyback policies.

Prospect of widespread biosimilar competition at

the end of the decade means some element of uncertainty on long-term growth, even if the launch of subcutaneous formulations is an adequate answer. The share remains very sensitive to any doubts that might arise from the short-term performances of the major products (Avastin).

Share ownership
Hoffmann families Novartis Free float Shares 9.3% 6.1% 84.6% Votes 50.0% 33.3% 16.7%

Diary dates
12/09/2010 12/17/2010 02/02/2011 Investors day PDUFA date Avastin 2010 earnings

Pharmacie

I 68

S EC TOR R EPOR T

Financial Data 31/12

Breakdown by activity (CHFm) Turnover Pharmaceuticals Diagnostics Reported operating profit Pharmaceuticals Diagnostics Others Reported operating margin Pharmaceuticals Diagnostics Profit & loss statement (CHFm) Turnover Change Organic growth EBITDA Change EBIT Change Adjusted EBIT Change Operating margin Financial items Pre-tax profit on ordinary activities Exceptional items Corporate tax Goodwill amortisation/ impairment Equity associates Minority interests Net profit on divested activities Reported net profit Change Adjusted net profit Change Cash flow statement (CHFm) Cash flow from operations Nets Investments Decrease (Increase) in WCR Free cash flow Acquisitions Dividend Capital increase Divestments Miscellaneous Increase (Decrease) in cash Net debt Gearing 2008 45,617 35,961 9,656 13,896 12,974 1,187 -265 30.5% 36.1% 12.3% 2008 45,617 -1.1% 6.0% 16,637 -2.5% 13,896 -4.0% 13,896 -4.0% 30.5% 236 14,132 28 -3,317 0 1 -1,875 8,969 -8.1% 9,666 -6.7% 2008 13,096 -3,557 -524 9,015 -5,223 -4,051 0 541 -936 -654 -16,682 -31.0% 2009 49,051 38,996 10,055 15,012 14,154 1,198 -340 30.6% 36.3% 11.9% 2009 49,051 7.5% 12.0% 18,028 8.4% 15,012 8.0% 15,012 8.0% 30.6% -2,045 12,967 -2,735 -1,722 0 0 -726 7,784 -13.2% 10,506 8.7% 2009 16,086 -3,219 324 13,191 -52,812 -4,395 0 300 3,167 -40,549 23,867 253.5% 2010e 48,630 38,150 10,480 16,060 14,750 1,665 -355 33.0% 38.7% 15.9% 2010e 48,630 -0.9% 7.0% 19,065 5.8% 16,060 7.0% 16,060 7.0% 33.0% -1,990 14,070 0 -3,230 0 0 -200 10,640 36.7% 11,327 7.8% 2010e 14,900 -3,250 -2,000 9,650 -500 -5,214 0 0 -800 3,136 20,731 147.3% 2011e 51,080 39,980 11,100 17,270 15,800 1,820 -350 33.8% 39.5% 16.4% 2011e 51,080 5.0% 7.0% 20,275 6.3% 17,270 7.5% 17,270 7.5% 33.8% -1,300 15,970 0 -3,795 0 0 -220 11,955 12.4% 12,650 11.7% 2011e 16,100 -3,500 -850 11,750 -500 -5,434 0 0 34 5,850 14,881 76.4% 2012e 53,950 42,050 11,900 18,620 16,970 2,000 -350 34.5% 40.4% 16.8% 2012e 53,950 5.6% 107.0% 21,625 6.7% 18,620 7.8% 18,620 7.8% 34.5% -900 17,720 0 -4,500 0 0 -240 12,980 8.6% 13,675 8.1% 2012e 17,200 -3,800 -900 12,500 -500 -6,038 0 0 -2 5,961 8,922 35.0%

Roche
CAGR 09/12 3.2% 2.5% 5.8% 7.4% 6.2% 18.6% -1.0%

CAGR 09/12 3.2%

6.3% 7.4% 7.4%

11.0%

18.6% 9.2%

CAGR 09/12 2.3% 5.7% -1.8% 11.2%

Pharmacie

I 69

S EC TOR R EPOR T

Balance sheet (CHFm) Net fixed assets o/w net goodwill o/w gross goodwill Financial fixed assets WCR Net assets on divested activities Total equity o/w shareholders' equity Provisions Net debt Per share data (CHF) Shares outstanding (millions) Diluted shares (millions) Reported EPS Adjusted EPS Goodwill Cash flow Net dividend Payout ratio Book value Financial ratios Personnel expenses (M) Personnel expenses (% of sales) Operating margin Adjusted operating margin Effective rate of tax Net margin ROE ROCE Capital employed (M) Interest cover (x) Debt/EBITDA (x) Gearing WCR (% of sales) Goodwill (% of book value) Net investments (% of sales) Investment ratios EV/Turnover (x) EV/EBITDA (x) EV/Adjusted EBIT (x) P/BV (x) P/CF (x) Adjusted P/E Reported P/E Net yield Free cash flow yield (%)
Rating changes for Roche in the last 12 months Date 07/21/2010 04/16/2010 02/03/2010 Rating Neutral Buy (*) Add (*)

2008 33,664 7,000 7,000 3,821 6,387 53,822 44,479 6,732 -16,682 2008 862.6 862.6 10.40 11.21 0.00 15.18 5.00 48.1% 51.56 2008 9,401 20.6% 30.5% 30.5% -23.4% 23.8% 20.2% 25.3% 40,051 na -1.0 -31.0% 14.0% 13.0% -7.8%

2009 31,963 7,000 7,000 4,123 3,720 9,414 7,366 6,525 23,867 2009 862.6 862.6 9.02 12.18 0.00 18.65 5.40 59.8% 8.54 2009 9,402 19.2% 30.6% 30.6% -16.8% 17.3% na 30.7% 35,683 7.3 1.3 253.5% 7.6% 74.4% -6.6%

2010e 32,500 7,000 7,000 4,000 6,103 14,072 11,572 7,800 20,731 2010e 862.6 862.6 12.33 13.13 0.00 17.27 6.30 51.1% 13.41 2010e 9,402 19.3% 33.0% 33.0% -23.0% 22.3% 91.9% 30.4% 38,603 8.1 1.1 147.3% 12.5% 49.7% -6.7% 2010e 2.9 7.3 8.7 10.3 8.0 10.5 11.2 4.6% 8.1%

2011e 32,300 7,001 7,001 4,000 5,570 19,488 16,488 7,500 14,881 2011e 862.6 862.6 13.86 14.66 0.00 18.66 7.00 50.5% 19.11 2011e 9,402 18.4% 33.8% 33.8% -23.8% 23.8% 72.5% 33.3% 37,870 13.3 0.7 76.4% 10.9% 35.9% -6.9% 2011e 2.6 6.6 7.7 7.2 7.4 9.4 10.0 5.1% 9.8%
Price CHF143.00 CHF177.50 CHF180.60

2012e 32,700 7,002 7,002 4,000 5,721 25,499 21,999 8,000 8,922 2012e 862.6 862.6 15.05 15.85 0.00 19.94 7.50 49.8% 25.50 2012e 9,403 17.4% 34.5% 34.5% -25.4% 24.5% 59.0% 35.4% 38,421 20.7 0.4 35.0% 10.6% 27.5% -7.0% 2012e 2.4 5.9 6.9 5.4 6.9 8.7 9.2 5.4% 10.5%

CAGR 09/12 0.8%

39.4%

CAGR 09/12 0.0% 0.0% 18.6% 9.2% 2.3% 11.6% 44.0% CAGR 09/12 0.0%

2.5%

Previous Buy Add (*) Buy (*)

Pharmacie

I 70

S EC TOR R EPOR T

6 December 2010
143. 1 123. 1 103. 1 83. 1 63. 1 43. 1 23. 1 dc -07 j ui n-08 j anv -09 aot -09 f v r -10 sept -10

Pharmaceuticals

Germany

Stada
STAGn.DE / SAZ@GR
St ada Rel. DJ STOXX Small 200

Buy

The recovery is on track

24.15 29.00 20.1% Stada is the worlds fifth-largest generics group. Its strategy consists of stepping up its presence in eastern Europe, notably in Russia. For the time being, it is investing discreetly in biosimilars via an unconsolidated specific project financing structure, in which it owns a 15.9% stake, Bioceuticals. In the event of success, Stada has a buy option enabling it to acquire all of the shares as of 2011, at a predetermined price. In 2009, Bioceuticals generated sales of 26.7m (16.1m in 2008) and earnings of 0.9m (vs. -14.3m in 2008). Among the treatments developed by Bioceuticals, Epo-Zeta (Retacrit/Silapo, anti-anaemic) is the only one on the market (launched at the start of 2008). Hospira obtained exclusive distribution rights in the EU (with the exception of Germany and certain countries in eastern Europe, where the US group only holds semi-exclusive rights, with Stada). Thanks to this partnership, over the next few years Bioceuticals will receive from Hospira milestone payments of up to 14m. This amount will notably depend on the possible marketing of Epo-Zeta in the USA and Canada. The other developments are more laborious: research on the immunostimulant Filgrastim (Phase I) was suspended in 2009. As for the monoclonal antibodies, the clinical tests will not start until Bioceuticals has found partners. Stada seems to have emerged from its recovery phase and is aiming for sustainable growth. The targets set for 2014, namely a doubling in the space of five years of net profit and annual growth in sales of 7%, look credible. Stada should be able to offset the price falls in Europe by cost savings (restructuring costs between now and 2013: 50m). We therefore assume an improvement in margins between now and 2014, even though we remain cautious (operating margin of 14.5%e in 2014 vs. 13.5% in 2009). We are confident as regards the recovery in earnings, and we have a Buy rating on Stada. Our target price of 29 corresponds to a reduction in the discount relative to the sector in terms of PE 12 from 30% to 15%.

Source: Natixis

Price Target Upside

12/03/2010

Performance Absolute Sector DJS Small200 12-month high/low DAX 30 DJS Small200 Market capitalisation Free float Daily volume

1m 9.9% 1.0% 4.1%

12m 0.6% 8.1% 21.1%

1 Jan -0.2% 3.7% 18.2%

32.10 / 20.70 6947.7 177.0 1.4bn 100.0% 7m

Analyst(s) Philippe Lanone (33 1) 58 55 05 03 philippe.lanone@natixis.com Batrice Muzard (33 1) 58 55 05 13 beatrice.muzard@natixis.com Equity Markets Bloomberg access equity.natixis.com NXSE

Distribution of this report in the United States. See important disclosures at the end of this report.

Close on 31/12 2009 2010e 2011e 2012e

Turnover (m) 1,568.8 1,590.0 1,690.0 1,830.0

Reported net profit (m) 100.7 65.0 132.0 151.0

Adjusted EPS () 1.97 2.14 2.33 2.57

Chg. EPS (%) 0.3 8.5 8.7 10.2

PE (x) 12.2 11.3 10.4 9.4

EV/EBIT (x) 11.0 10.4 9.6 8.8

P/CF (x) 7.4 7.1 6.6 6.2

Net yield (%) 2.3 2.5 2.7 2.9

Pharmacie

I 71

S EC TOR R EPOR T

Company profile
Stada is the worlds no.5 generics group with sales of 1.570bn in 2009. Sales and net income growth over the 2006/2009 period came to 8% and 3%. The company has been on an intensive acquisitions drive (taking over the UKs British Forum Bioscience, Russias MAKIZ in 2007 and Serbias Hemofarm in 2006). It boasts key positions in Europe, especially on the most developed market, Germany, where it is ranked 3rd behind Sandoz and Ratiopharm in terms of market share. Moreover, Stada is ranked 2nd in Russia, a market posting very fast growth (Stada 2006/2009 sales growth: 30%/year).

Recent events
Operating trends are on the right track: recovery in sight in Serbia and in Germany, good performances in Russia (sales +22%, 9M 10/9M 09). The management expects sales growth in 2010e (Natixis estimate: 1,590m, +1.4%). The group will probably not have to set aside any further provisions as it is booking only solvent sales. This good top-line performance is reflected in margin trends. The operating margin (adjusted) was 14.1% in 9M 10 (vs. 12.6% in 9M 09).

2009 sales by division

2009 reported operating profit (m)

156 Branded 25% 106 56 6 156 75

Generics 71%

Others 4%

-44 -94 -39 Others Generics Branded

Opportunities / Strengths
A fast-growing market, not subject to any regulatory

Weaknesses / Threats
Overexposed to the German market (34% of sales

or patent risk. A significant market share in emerging countries in Eastern Europe, especially Russia (the companys second most important country in terms of sales). Speedy adjustment of the cost base in Germany (number of medical reps reduced by 150). Earnings growth estimated at 19% between 2007 and 2012, picking up after an expected blip in H2 08. Stada is a potential target in the current wave of consolidation.

in 2009), where generic selling prices are under a great deal of pressure. Solvency problems of payer organisations, especially in Serbia. Stada seems to have emerged from its recovery phase and is aiming to meet the targets set for 2014: doubling net profit in five years, with annual average sales growth of 7%. Relatively stretched gearing level (almost 1x equity). External growth operations, some of which could prove dilutive. 2010 earnings Q1 11 earnings AGM Q2 11 earnings Q3 11 earnings

Share ownership
Free float Shares Votes 100.0% 100.0%

Diary dates
03/30/2011 05/12/2011 06/16/2011 08/11/2011 11/10/2011

Pharmacie

I 72

S EC TOR R EPOR T

Financial Data 31/12

Breakdown by activity (m) Turnover Generics Branded Others Reported operating profit Generics Branded Others Reported operating margin Generics Branded Others Profit & loss statement (m) Turnover Change Organic growth EBITDA Change EBIT Change Adjusted EBIT Change Operating margin Financial items Pre-tax profit on ordinary activities Exceptional items Corporate tax Goodwill amortisation/ impairment Equity associates Minority interests Net profit on divested activities Reported net profit Change Adjusted net profit Change Cash flow statement (m) Cash flow from operations Nets Investments Decrease (Increase) in WCR Free cash flow Acquisitions Dividend Capital increase Divestments Miscellaneous Increase (Decrease) in cash Net debt Gearing 2008 1,646.6 1,155.0 368.9 122.7 175.7 136.7 53.0 -14.1 10.7% 11.8% 14.4% -11.5% 2008 1,646.2 4.8% 255.4 -11.8% 176.4 -18.1% 219.0 -11.8% 13.3% -68.4 108.0 0.0 -28.5 -0.8 78.7 -25.1% 115.7 -21.2% 2008 150.4 -113.9 21.1 57.5 -56.7 -41.6 0.6 27.3 -43.8 -56.7 1,015.8 121.0% 2009 1,568.8 1,115.6 392.6 60.6 191.9 156.3 74.9 -39.3 12.2% 14.0% 19.1% -64.8% 2009 1,568.8 -4.7% 280.1 9.7% 191.9 8.8% 211.1 -3.6% 13.5% -50.1 141.8 0.0 -40.8 -0.3 100.7 28.0% 115.8 0.1% 2009 191.2 -97.2 59.3 153.3 0.0 -30.5 0.0 27.3 -3.0 147.1 899.0 103.4% 2010e 1,590.0 1,095.0 420.0 75.0 157.0 125.0 71.0 -39.0 9.9% 11.4% 16.9% -52.0% 2010e 1,590.0 1.4% 300.0 7.1% 157.0 -18.2% 220.0 4.2% 13.8% -52.0 105.0 0.0 -40.0 0.0 65.0 -35.5% 126.0 8.8% 2010e 200.0 -100.0 -45.0 55.0 0.0 -32.3 0.0 3.0 0.0 25.7 873.3 96.4% 2011e 1,690.0 1,160.0 450.0 80.0 232.0 165.0 95.0 -28.0 13.7% 14.2% 21.1% -35.0% 2011e 1,690.0 6.3% 326.0 8.7% 232.0 47.8% 235.0 6.8% 13.9% -47.0 185.0 0.0 -53.0 0.0 132.0 103.1% 137.0 8.7% 2011e 215.0 -110.0 -35.0 70.0 0.0 -35.3 0.0 0.0 0.0 34.7 838.6 83.9% 2012e 1,830.0 1,260.0 485.0 85.0 245.0 190.0 100.0 -45.0 13.4% 15.1% 20.6% -52.9% 2012e 1,830.0 8.3% 346.0 6.1% 250.0 7.8% 252.0 7.2% 13.8% -43.0 207.0 0.0 -56.0 0.0 151.0 14.4% 151.0 10.2% 2012e 230.0 -115.0 -40.0 75.0 0.0 -38.2 0.0 0.0 0.0 36.8 801.8 72.3%

Stada
CAGR 05/08 5.3% 4.1% 7.3% 12.0% 8.5% 6.7% 10.1% -4.6%

CAGR 05/08 5.3%

7.3% 9.2% 6.1%

13.4%

14.4% 9.2%

CAGR 05/08 6.4% 5.8% -21.2% 7.8%

Pharmacie

I 73

S EC TOR R EPOR T

Balance sheet (m) Net fixed assets o/w net goodwill o/w gross goodwill Financial fixed assets WCR Net assets on divested activities Total equity o/w shareholders' equity Provisions Net debt Per share data () Shares outstanding (millions) Diluted shares (millions) Reported EPS Adjusted EPS Goodwill Cash flow Net dividend Payout ratio Book value Financial ratios Personnel expenses (M) Personnel expenses (% of sales) Operating margin Adjusted operating margin Effective rate of tax Net margin ROE ROCE Capital employed (M) Interest cover (x) Debt/EBITDA (x) Gearing WCR (% of sales) Goodwill (% of book value) Net investments (% of sales) Investment ratios EV/Turnover (x) EV/EBITDA (x) EV/Adjusted EBIT (x) P/BV (x) P/CF (x) Adjusted P/E Reported P/E Net yield Free cash flow yield (%)
Rating changes for Stada in the last 12 months Date 06/08/2010 Rating Buy

2008 1,307.5 400.0 450.0 105.4 569.1 839.7 827.4 126.4 1,015.8 2008 58.8 58.8 1.30 1.97 0.00 2.56 0.52 40.1% 14.07 2008 10.7% 13.3% -26.4% 9.5% 8.3% 1,927 3.2 4.0 121.0% 34.6% 47.6% -6.9%

2009 1,309.9 380.0 450.0 96.7 480.7 869.7 861.1 118.6 899.0 2009 58.8 58.7 1.71 1.97 0.00 3.26 0.55 32.1% 14.68 2009 12.2% 13.5% -28.8% 11.7% 8.3% 1,861 4.2 3.2 103.4% 30.6% 43.7% -6.2%

2010e 1,360.0 360.0 450.0 130.0 469.2 905.8 890.8 180.0 873.3 2010e 58.8 58.8 1.11 2.14 0.00 3.40 0.60 54.3% 15.15 2010e 9.9% 13.8% -38.1% 7.3% 8.4% 1,919 4.2 2.9 96.4% 29.5% 39.7% -6.3% 2010e 1.8 9.4 12.1 1.5 8.5 12.8 14.2 2.6% -6.6%

2011e 1,380.0 360.0 450.0 140.0 518.2 999.6 984.6 200.0 838.6 2011e 58.8 58.8 2.24 2.33 0.00 3.66 0.65 29.0% 16.75 2011e 13.7% 13.9% -28.6% 13.4% 8.6% 1,988 5.0 2.6 83.9% 30.7% 36.0% -6.5% 2011e 1.5 8.2 9.6 1.5 7.0 9.6 13.4 2.9% 0.0%
Price 29.79

2012e 1,380.0 361.0 451.0 145.0 601.3 1,109.5 1,094.5 215.0 801.8 2012e 58.8 58.8 2.57 2.57 0.00 3.91 0.70 27.3% 18.61 2012e 13.7% 13.8% -27.1% 13.8% -3.3% 2,071 5.9 2.3 72.3% 32.9% 32.5% -6.3% 2012e 1.5 9.5 11.1 1.7 9.4 12.3 18.6 2.2% 4.1%

CAGR 05/08 1.8%

8.5%

CAGR 05/08 0.0% 0.1% 14.5% 9.2% 6.3% 8.4% 8.2% CAGR 05/08 -

3.6%

Previous Neutral

Pharmacie

I 74

S EC TOR R EPOR T

Sanofi-aventis : The investment research supervisor(s), or a Senior Manager(s) of Natixis and/or of its subsidiaries is/are on the managing board of the issuer. Rating changes for Roche in the last 12 months Date 07/21/2010 04/16/2010 02/03/2010 Date 06/08/2010 Rating Neutral Buy (*) Add (*) Rating Buy Previous Buy Add (*) Buy (*) Previous Neutral Price CHF143.00 CHF177.50 CHF180.60 Price 29.79

Rating changes for Stada in the last 12 months

* see the methodology applied to ratings before 05/01/2010 and indicated in the disclosure below

Reference prices are based on closing prices.

The information contained in these publications is exclusively intended for a client base consisting of professionals or qualified investors. It is sent to you by way of information and cannot be divulged to a third party without the prior consent of Natixis. It cannot be considered under any circumstances as an offer to sell, or a solicitation of any offer to buy financial instruments. While all reasonable effort has been made to ensure that the information contained is not untrue or misleading at the time of publication, no representation is made as to its accuracy or completeness and it should not be relied upon as such. Past and simulated performances offer no guarantee as to future performances. Any opinions offered herein reflect our current judgement and may change without notice. Natixis cannot be held responsible for the consequences of any decision made with regard to the information contained in those documents. Natixis has set up due procedures for the separation of activities, notably in order to prevent conflicts of interest between the research activities and its other activities. Details of these information barriers are available on request from the head of compliance. On the date of those reports, Natixis and/or one of its subsidiaries may be in a conflict of interest with the issuer mentioned herein. In particular, it may be that Natixis or any person or company linked thereto, their respective directors and/or representatives and/or employees, have invested on their own account in, or act or intend to act, in the next twelve months, as an advisor, provider of liquidity, market maker, or corporate banker (and notably for underwriting transactions, placements or connected transactions), for a company discussed in this report. This research may be disseminated from the United Kingdom by Natixis, London Branch, which is authorised by the ACP and subject to limited regulation by the Financial Services Authority. Details about the extent of regulation by the Financial Services Authority are available from the London Branch on request. The transfer / distribution of this document in Germany is done by / under the responsibility of Natixis Zweigniederlassung Deutschland. NATIXIS is authorized by the ACP and regulated by BaFin (Bundesanstalt fr Finanzdienstleistungsaufsicht) for the conduct of its business in Germany. Natixis is authorised by the ACP and regulated by Bank of Spain and the CNMV for the conduct of its business in Spain. Natixis is authorised by the ACP and regulated by Bank of Italy and the CONSOB (Commissione Nazionale per le Societ e la Borsa) for the conduct of its business in Italy. Natixis, a foreign bank and broker-dealer, makes this research report available solely for distribution in the United States to major U.S. institutional investors as defined in Rule 15a-6 under the U.S. Securities Act of 1934. This document shall not be distributed to any other persons in the United States. All major U.S. institutional investors receiving this document shall not distribute the original nor a copy thereof to any other person in the United States. Natixis Bleichroeder LLC a U.S. registered broker-dealer and member of FINRA is a subsidiary of Natixis. Natixis has no officers or employees in common with Natixis Bleichroeder LLC. Natixis Bleichroeder LLC did not participate in the preparation of this research report and as such assumes no responsibility for its content. This research report has been prepared and reviewed by research analysts employed by Natixis, who are not associated persons of Natixis Bleichroeder LLC and are not registered or qualified as research analysts with FINRA , and are not subject to the rules of the FINRA. In order to receive any additional information about or to effect a transaction in any security or financial instrument mentioned herein, please contact a registered representative Natixis Bleichroeder LLC 1345 Avenue of the Americas, New York, NY 10105.

Pharmacie

I 75

S EC TOR R EPOR T

Natixis
Paris 30 avenue Pierre Mends France 75013 Paris France
A public company with a board of directors and 4,653,020,308.80 capital 542 044 524 RCS Paris

Natixis Bleichroeder Inc.

Postal address BP 4 75060 Paris Cedex 02 France New York 1345 Avenue of the Americas New York, NY 10105 USA
Member of the NASD and SIPC

Natixis
Frankfurt Im Trutz Frankfurt 55 D-60322 Frankfurt Germany Sarah Schmitz Tel. (49 69) 971 53 312 sarah.schmidtke@de.natixis.com

Office address 47 quai dAusterlitz 75013 Paris France

Chris Thompson Tel. (1 212) 698-3400 chris.thompson@us.natixis.com

Head of Research Sixte de Gastines Tel. (33 1) 58 55 06 87 sixte.degastines@natixis.com Head of Equity Sales Continental Europe Philippe Denoyelle Tel. (33 1) 58 55 05 91 philippe.denoyelle@natixis.com Head of Equity Sales UK Chris Thompson Tel. (1 212) 698-3400 chris.thompson@us.natixis.com Head of Sales trading Christian Nucci Tel. (33 1) 58 55 05 76 christian.nucci@natixis.com Head of Corporate broking Cdric Richard Tel. (33 1) 58 55 90 60 cedric.richard@natixis.com Chris Thompson Tel. (1 212) 698-3400 chris.thompson@us.natixis.com London Natixis London Branch LTD Cannon Bridge House 25 Dowgate Hill London EC4R 2YA UK