Od Citibank:
Global Economic Flash
Stress Tests Results: Still on the Treadmill
The stress test results for the top 19 US bank holding companies came out in
line with the most recent leaks: a total capital need of $75 billion. While we
take these results as reassuring, the details are not likely to calm skeptics for
at least three reasons.
First, the political constraint from Congress on additional financing for the US
financial institutions is at odds with the desire of US authorities to reduce
uncertainty in the aftermath of the crisis. Skeptics would maintain that capital
needs greater than the resources already pledged would not be announced.
Second, criticism continues about the degree of stress built into the “more
adverse” scenario used by authorities to calculate the needed capital buffer.
Third, assumed earnings and details on changes in loan loss reserves used in
the test over this year and next are unclear. Skeptics are maintaining that the
earnings are too optimistic. They also complain that the detail on loan
portfolios is not detailed enough to independently verify their regulators’ work.
Skeptics, though, are likely wrong for two reasons. First, the average
percentage losses on the loan portfolio are very aggressive—some 32% higher,
for example, than the percentage losses assumed by the IMF in its most recent
report. The average loan loss percentage is also higher than the worst
percentage of the Great Depression.
Second, despite the naysayers, the greater disclosure of firm-level information
should allow some convergence among independent assessments of the
aggregate losses that, in turn, should calm concerns about systemic vulnerability.
------------------------------------------------------------------------------
The stress test results for the top 19 US bank holding companies came out in
line with the most recent leaks: a total capital buffer need of $75 billion. The
individual company detail also matched leaks. Banks immediately announced
plans to raise capital to meet the additional requirements, eliminating some of
the uncertainty embedded in the original plan, which gave the banks 30 days
to come up with a capital plan for the needed buffer. Overall, these results
should help calm concerns about the solvency of the US financial system. The
details, though, may not calm skeptics for at least three reasons, leaving us on
a treadmill for longer.
First, the political constraint from Congress on additional financing for the US
financial institutions is at odds with the desire of US authorities to reduce
uncertainty in the aftermath of the crisis. Skeptics would maintain that capital
needs greater than the resources already pledged would not be announced.
Such an announcement would be at odds with the desire to foster stability.
This political constraint increases the need for disclosing the details of the
regulators calculations—much of which is included in the firm-specific tables
included in the report.
Second, criticism continues about the degree of stress built into the “more
adverse” scenario used by authorities to calculate the needed capital buffer.
When the downside scenario was framed in February, it was meant to be an
outcome with 10-15% probability of occurring. Since then, though, the
consensus growth and unemployment forecasts have become more bleak.
Expected real GDP growth for 2009 and 2010, using Bloomberg consensus
data, is now -2.5% and 1.8%, respectively. The supervisors used a base case
of -2.0% and 2.1%. The figures relevant for the calculation of the capital
cushion under the adverse scenario, though, are still well away from the
consensus at -3.3% in 2009 and 0.5% in 2010. The likelihood of hitting those
numbers, though, is certainly greater than 10-15%. The degree of adversity
embodied in the unemployment assumption is lower than that of GDP. This
year the adverse rate—8.9%—which matches the current consensus and the
rate we reached with today’s payroll report!
Third, the transparency of the underlying earnings, reserves and portfolio
information, while greater than initially planned, is still less then some are
demanding. For example, detail is not provided to allow others to separate out
assumed net earnings over the next two years from net charge-offs on existing
loan loss provisions or the assumed provisions needed to meet 2011 expected
losses. That said, our banking team assumed earnings over this year and next
that are some $95 billion higher than the pretax earnings net of changes in
loan loss provisions (net charge-offs and a cushion for 2011 losses) presented
in the Fed analysis1. That amount is some 20% of the assumed losses by
regulators, a healthy cushion that implies the earnings figures are likely
conservative.
Despite these objections from skeptics, the results will likely lead to greater
confidence in the US banking system, as regulators and policymakers hope.
First, the average percentage losses on the loan portfolio are aggressive—some
32% higher, for example, than the percentage losses assumed by the IMF in its
most recent report (see Figure 1). The average loan loss percentage is also
higher than the worst percentage of the Great Depression. These more
aggressive loss assumptions should offset some of the concerns about the
likelihood of the adverse scenario that underpins the capital requirement.
Second, the greater disclosure of firm-level information, though not complete,
allows independent assessments of the aggregate losses that should calm
concerns about systemic vulnerability. While it may take some time for this
analysis to be done, the combination of the greater information, the alacrity of
the disclosure of the capital raising plans of the banks and positive effects of
improving financial conditions on the economic outlook itself, should all prove
helpful to ultimately getting us all off the “stress test treadmill”.
)
