Sei sulla pagina 1di 2

!

Forex Income Portfolios Currency Income Portfolios Built. One Client at a Time Time. Built. One Client at a .

Stefan Whitwell, CFA, CIPM Chief Investment Officer

March 26, 2013 MANAGER DATA, DUE DILIGENCE AND SELECTION PROCESS RISKS There are 10 specic risks that the portfolio manager needs to manage. 1. Using too many managers: The portfolio will produce Beta -- and investors will only want to pay commoditized fee levels associated with index funds; what investors need, however, is Alpha. 2. Using too few managers: Idiosyncratic risk will be excessive leading to big losses. 3. Believing the DDQ Narrative: Those of us who worked on trading desks on Wall Street can attest that most fund managers trade quite similar to most other fund managers, even though every single DDQ ever written makes specic claims of being different (hence the frequent use of the word proprietary). Note: often the core trading is similar, while relatively insignicant factors are in fact unique. If everyone is driving in the same lane in the same direction, it is not a material difference if you hold the steering wheel with one hand, two, left or right. 4. Managers appear different when in fact they are similar: Everyone claims to be different, but statistically we know this is not true. If we assume that the portfolio is diversied, when it is not, then the investors will likely face a nasty surprise. 5. Undisclosed policies: We have seen situations where, for example, a manager offered their program denominated in different G-7 currencies which implied that the difference in return between the different units was merely the monthly currency effect. It turns out that they do not have a systematic method for converting returns and improvise based on their market views, which presents another risk and furthermore, their method of currency conversion is not disclosed in any written materials. We have also seen examples where a historical return table published by a major investment bank failed to disclose that the volatility target for that program changed dramatically mid-stream and caused a big change to subsequent data. It takes a lot of work to make sure one really understands the advertised numbers. 6. Using non-homogenous data in analytical comparisons: The major FX platforms charge different fees and require the managers to trade at different volatility-levels so that they are incomparable. In addition, we have seen situations where a platform (run by a large investment bank) did not distinguish between actual returns and hypothetical returns for one or more managers, which raises a whole set of other questions. Furthermore, some managers disseminate gross returns but will provide net returns upon request; others use net returns. And lastly, different managers and platforms report their returns in different currencies. Many people in the business try and short-cut the work involved by paying for access to huge databases that provide data on large numbers of sub-managers. While convenient, those databases do not have complete disclosure, manager by manager, of how those data were calculated and unless you source original data, you have no means of verifying its accuracy. As they say in engineering, trash in, trash out. Analytical comparisons, such as manager selection analysis, requires homogenous data. Empirical Solutions, LLC
815-A Brazos Street | Suite 491 | Austin, Texas 78701 | 877.936.3372 | stefan@empiricalresults.net

Stefan Whitwell, CFA, CIPM Chief Investment Officer

7. Forming analytical conclusions based on gross data: Many of the graphs and much of the performance data published by platforms (manager summary sheets) use gross returns (returns prior to subtracting fees). This inates the returns and misrepresents the actual client experience, which is decidedly net of fees. Mathematically, the difference between gross and net returns can be substantial especially taking compounding into effect for longer periods of time.
There is a popular sub-manager, for example, that proudly advertises its Sharpe Ratio, which it calculates using gross returns. If you re-calculate the ratio on a net basis, the result is shockingly lower; so much so, we got in touch with the manager to make sure that our calculation was correct, since it was so much lower.

Therefore, relying on statistics produced by third party databases or by the managers themselves is risky business. It is important to calculate the ratios rsthand since it is the only way to know, with certainty, how the calculations were made. Due to the time it takes, however, and the work required, few do this. 8. Feeling safer with long track records: We recommend using the last ve or six years worth of data. That said, it is human nature to feel safer with managers that have long performance records. One well known manager comes to mind that manages several billion dollars. They proudly advertise a Sharpe Ratio in excess of 1.0 which is impressive and true. Since they are big and have a track record in excess of a decade, most people will believe that statistic at face value and never do the work to calculate it for themselves. When we did our own calculations, we discovered that the rst three years of data were spectacular and the last seven years was below average; as they say, do not judge a book by its cover. 9. Believing that big-name consultants advice is safe: One of the biggest currency platforms today uses as one of its selling points the fact that it hires a major and bigname consulting rm to do intensive due-diligence on managers before they are allowed to join their platform. This leads the users of such platforms to feel safe. Unfortunately, big does not necessarily denote better and in mid-2012 there was a spectacular example of the fallibility of this big rms analysis: one of the managers that was approved by both the investment bank and the consultant had a spectacular blow-up -- down over 50% in ve months (down 25% in one month). Clearly that manager did not have a sufcient risk-management system in place. 10. Over-reliance on Past Data: Historical data is useful -- from it we can learn a lot about the manager. A seasoned professional with rst-hand forex trading experience can also use it to verify gaps and holes in the narratives in written documentation that can identify risk issues not otherwise disclosed. However, in one of the great ironies of the investment management business, most professionals regularly use (sometimes required to be used by regulators) disclaimers stating something to the effect that past performance is no guarantee of future results. However, in practice, most portfolio managers and investment committees behave exactly the opposite -- even though there is plenty of data that shows the fallibility of this approach. Linear interpolation seems innate to human nature. Keeping this irrational habit in mind, it is vital that the manager selection process have a logical framework by which managers are picked, so that past data is used constructively, but not relied upon with false condence.

Empirical Solutions, LLC


815-A Brazos Street | Suite 491 | Austin, Texas 78701 | 877.936.3372 | stefan@empiricalresults.net

Potrebbero piacerti anche