Sei sulla pagina 1di 3

Joshua Younger Henry St John Global Research

(1-212) 270-1323 (1-212) 834-5669 J.P. Morgan Perspectives


joshua.d.younger@jpmorgan.com henry.stjohn@jpmorgan.com 21 February 2020
Munier Salem
(1-212) 270-0317
munier.salem@jpmorgan.com

affect Reserve Account holdings for Libra and other words: why hold negative-yielding stablecoins when
stablecoins. This is not an isolated problem: looking your old private fiat money is offering zero or above?
across major currencies, only USD is a material source Alternatively, this could lead to the imposition of traction
of positive-yielding free float short-term government costs, which would similarly reduce the incentive to hold
securities (Figure 6). Despite their recent assurances to stablecoins. Thus, negative yields, particularly if they
the contrary, there are no guarantees that the Fed will not occur unexpectedly, introduce risks that for a
eventually adopt a negative interest rate policy—in fact, stablecoin with global scale would constitute a
markets are pricing a reasonably high likelihood of just significant threat to financial stability. Suddenness in
such an eventuality (see Interest Rate Derivatives, US the move to negative policy rates is, in fact, the rule
Fixed Income Markets Weekly, J. Younger et al., 27 rather than the exception thus far—as evidenced by the
September 2019). experience of Europe and Japan (Figures 7 and 8).

Figure 5: Thanks to large-scale asset purchases across Figure 7: The options market in Europe strongly discounted the
developed markets, the free float of G4 central government debt risk of negative rates until a few months before they emerged …
has not grown in more than five years … EUR 1Yx1Y zero-strike receiver swaption deltas compared to EUR
USD, EUR, JPY and GBP central government securities separated into those 1Yx1Y swap yields and NIRP risk inferred from ATMF 1Yx1Y
held by central banks (FX reserves as well as QE-related purchases) and free swaptions; %
float (LHS; $trn) and the fraction that is free floating (RHS; %), all converted
0.8 NIRP-risk from zero-strike receiver delta
into US dollar equivalents
NIRP-risk from ATMF pricing + normal distribution
30 90%
0.6 1Yx1Y swap yield
25 80%
20 0.4
70%
15 0.2
60%
10
0.0
Held by central banks 50%
5 Free float
% free float (RHS) -0.2
0 40% Jan 14 Apr 14 Jul 14 Oct 14 Jan 15 Apr 15 Jul 15 Oct 15
03 04 05 06 07 08 09 10 11 12 13 14 15 16 17 18 19
Source: J.P. Morgan, Haver Analytics, NYFRB, ECB, BoE, BoJ Source: J.P. Morgan

Figure 6: … and Treasury Bills and short nominal coupons are the
only clear source of positive yield short-term government securities Figure 8: … and in Japan, the risk was strongly discounted right
up until the breach occurred
Government securities with <1yr remaining maturity by currency, split into
JPY 1Yx1Y zero-strike receiver swaption deltas compared to EUR 1Yx1Y
positive and negative yield free float, central bank holdings (FX reserves and
QE-related holdings) and money market fund assets as of August 2019; $bn swap yields and NIRP risk inferred from ATMF 1Yx1Y swaptions; %
5000 Positive yld free float MMFs 1.00 NIRP-risk from zero-strike receiver delta
NIRP-risk from ATMF pricing + normal distribution
4000 Central Banks Negative yield free float 0.75 1Yx1Y swap yield

3000 0.50

0.25
2000
0.00
1000
-0.25
0
USD JPY EUR GBP KRW CAD -0.50
Note: We assume 15% of FX reserve holdings across currencies are <1yr remaining Jan 14 Jul 14 Jan 15 Jul 15 Jan 16 Jul 16
maturity coupon securities. ECB holdings based on J.P. Morgan estimates, while NYFRB,
Source: J.P. Morgan
BoE, and BoJ data are more granular.
Source: J.P. Morgan, Crane’s, NYFRB, ECB, BoJ, BoE, IMF COFER
One way to avoid negative yields on Reserve Account
One solution would be to pass losses along to token assets is to focus on high-quality bank deposits rather
holders. This would, however, provides a strong than government bonds. As noted above, private
disincentive since bank deposits have generally not money is generally exempted from negative policy rates.
followed policy rates into negative territory. In other However, this is easier said than done. The difficulty lies

55
Joshua Younger Henry St John Global Research
(1-212) 270-1323 (1-212) 834-5669 J.P. Morgan Perspectives
joshua.d.younger@jpmorgan.com henry.stjohn@jpmorgan.com 21 February 2020
Munier Salem
(1-212) 270-0317
munier.salem@jpmorgan.com

primarily in bank liquidity requirements. In particular, Figure 9: Among G4 economies, large commercial banks in the
systemically important banks are required to hold high- US are most liquidity constrained…
Aggregate LCR by jurisdiction as of 2Q 2019; %
quality liquid assets (HQLA) to cover potential stressed
150%
outflows. These outflows are assigned different weights
depending on their characteristics, with stable retail 140%
deposits assumed to be fairly sticky (~3% runoff) but
operational (i.e., those used for cash flow management 130%
transactional activity; ~25%) and especially non-
operational wholesale deposits (i.e., those not used 120%
primarily for operations; ~60%) are assumed to be much
more flight-prone. 110%

100%
We think it likely that stablecoin reserve deposits at US Euro Zone Japan UK
traditional banks would fall into the non-operational Note: Includes US GSIBs, the 10 largest Euro Zone banks, the four largest public
wholesale category. That means consumer inflows into Japanese banks, and the three largest U.K. banks.
new tokens would constitute a conversion of stable Source: J.P. Morgan, Pillar 3 Disclosures
retail funding into non-operational wholesale deposits.
The additional HQLA required could have a material Figure 10: …which is reflected in their strong preference to
impact on bank profitability. Imagine, for example, a source new funding from retail and secured wholesale sources
while specifically avoiding non-operational wholesale deposits
canonical commercial bank with a typical funding mix: Change in unweighted outflows by category for US GSIBs, 3Q 2019
40% retail, 20% operational wholesale, and 15% non- versus 2Q 2017; $bn
operational wholesale deposits, with the remaining 25% $500
coming from secured wholesale sources. Were a quarter YTD 2018
of those depositors to move their savings into a new $400
stablecoin, and those inflows were rolled back into the 2017 Total
$300
banking system rather than government securities, the
bank in question would need to increase their holdings $200
of HQLA by roughly 40%—presumably at the expense $100
of wider spread lending such as loans. To the extent this
$0
becomes a systemic shift it would likely contract
overall availability of credit (by ~8% in our example) -$100
and weigh on bank profitability. Unsecured Non-op Operational Retail Secured
debt deposits deposits deposits wholesale
Source: J.P. Morgan, Pillar 3 Disclosures
In practice, such punitive treatment is a strong
disincentive to adding—even retaining—non-
That is not to say stablecoins could not find jurisdictions
operational wholesale deposits. This constraint is
in which it is easier to source deposits for reserves. For
particularly acute in the US, where large banks are
example, Japan and the U.K. are much less constrained
closest to the minimum required holdings of HQLA
by liquidity requirements than the US and Europe. This
given their funding mix (Figure 9). As a result, though
all suggests a successful stablecoin would likely be
US banks have grown meaningfully over the past two
backed by a mix of short-term USD government
years, most of that funding has come in the form of
securities, which are positive yielding with significant
retail and secured wholesale sources (Figure 10). There
free float, as well as JPY and GBP bank deposits. This
has in fact been a concerted effort among GSIBs to
is in principle a scalable proposition, but it does limit the
actively shed non-operational funding for precisely this
extent to which a multi-currency reserve could be
reason (see Deposit non grata, A. Roever et al., 27 Feb.
diversified in practice.
2015). Given these regulatory constraints, we think
stablecoin issuers may find it hard to source deposits
at banks for which liquidity constraints are or could
become binding.

56
Joshua Younger Henry St John Global Research
(1-212) 270-1323 (1-212) 834-5669 J.P. Morgan Perspectives
joshua.d.younger@jpmorgan.com henry.stjohn@jpmorgan.com 21 February 2020
Munier Salem
(1-212) 270-0317
munier.salem@jpmorgan.com

Conclusions
What are we to conclude from this exercise? The world
is absolutely ready for private money—most of what we
think of as fiat currency is already privately issued.
However, the likely regulatory and compliance required
for stablecoins to, for example, be accepted as payment
for liabilities to the central government would be
significant. Sourcing collateral for a global stablecoin
will likely be a challenge, but is absolutely achievable,
and will likely require a preference for bank deposits
over government securities in some jurisdictions. That
said, the risk that USD policy rates in particular turn
negative is a significant risk to the stability of asset-
backed designs. Energy requirements, however, remain a
potentially significant limitation. Absent substantial and
ongoing improvements in efficiency, it will be very
difficult for truly distributed stablecoins to achieve
global scale, in our view. Reliance on a central authority
for validating transactions and maintaining the integrity
of the ledger is a possible solution, but does not offer the
same benefits as a true DLT.

Joshua Younger AC
joshua.d.younger@jpmorgan.com
J.P. Morgan Securities LLC

Munier Salem
munier.salem@jpmorgan.com
J.P. Morgan Securities LLC

Henry St John
henry.stjohn@jpmorgan.com
J.P. Morgan Securities LLC

57

Potrebbero piacerti anche