Barclays analyst Jeffrey Meli has issued a report "European bank stress tests: A preview" in which he estimates that if properly executed, the Stress Test, whose results are to be announced on July 23, will require an infusion of €85 billion to replenish capital levels. Specifically, quantifying the amount of capital needed would include €36 billion for Spanish cajas, a number far greater than expected to date, €34 billion for the German landesbanks, €8.6 billion for Greek banks and €6 billion for Portuguese banks. To be sure, Barclays has a positive spin on things, probably no doubt correlated to the bank's existing holdings of sovereign debt (the same way RBS tried to discredit Zero Hedge when half a year ago we dared suggest that Greece is in far bigger trouble that expected, only to be proven subsequently that even we were too optimistic): "We have a bias to be long risk as the results of the stress tests are released. First, the capital needs we estimate are not insurmountable, particularly given the programs already in place to address them: FROB in Spain, the SoFFin in Germany, and the Financial Stability Fund in Greece. In each case, the total needs are within the potential scope of the programs. For the majority of banks, transparency alone may succeed in restoring confidence. Any market stabilization due to the stress tests would be beneficial to banks, particularly if it allowed them to issue term debt at lower spreads and move away from covered bonds and ECB funding that they have been forced to use recently." Yet even the admittedly optimistic Barclays highlights the following major risks: "Capital hurdles may be set too low by looking at T1 capital instead of core T1 capital. Bank books may be treated too lightly. Finally, spreads have rallied over the past two weeks, suggesting the bar is no longer set so low that any disclosure whatsoever will cause a rally."
Some other troubling observations in the Barclays piece revolve around the issue we have discussed ad inf previously, namely the insufficient provisioning for sovereign debt haircuts:
CEBS has not yet disclosed their assumptions concerning losses on sovereign debt. Media reports have suggested that banks will be stressed for a 3% haircut on Spanish debt and a 17% haircut on Greek debt. These losses are lower than those implied by CDS, and with the cash-CDS basis negative, are even more divorced from the losses implied by cash valuations.
Another caveat as we pointe dout yesterday, has to do with the isolation of sovereign stress merely to trading accounts. Obviously this is simply stupid.
Further, it remains unclear if the sovereign losses will be applied only to assets held in trading books, as opposed to those in held-to-maturity (HTM) accounts. The rationale for applying the stress to trading books only is that they reflect the mark-to-market losses of those exposures, but that sovereign defaults are unlikely in the near term after the creation of the EFSF. Therefore, applying the losses to HTM assets overstates the potential near-term risks posed by sovereigns.
In fact the decision is so stupid, Barclays had to spend two full paragraphs defending it.
Despite the market’s focus on this issue, there are legitimate reasons not to apply losses to sovereign debt in HTM books. First, capitalizing banks based on potential sovereign defaults has an element of double counting. The European authorities have already created a program to provide funding to peripherals, which, if successful, should obviate the need also to provide enough capital to banks to counter potential sovereign losses. Given the extent to which sovereign debt is held in the banking system, doing both essentially requires funding potential sovereign losses twice.
Second, accurately predicting the implications of a sovereign default is not as simple as haircutting the amount of sovereign debt held by each institution based on a recovery assumption, due to the cross-holdings of bank liabilities across the financial system. Ostensibly, some banks would fail should there be a sovereign default, so institutions that held their debt would be exposed to secondary risks, which would have tertiary risks for their creditors, etc. Estimating the potential fallout from a domino-like banking crisis is nearly impossible, in our view, and the severity of the possible near-term implications of a sovereign default speaks to the point of creating the EFSF in the first place.
However, there is one important reason to include HTM sovereign exposures in some form – market expectations. It is clear from press reports that investors have a laser-like focus on the sovereign loss assumptions, and including the HTM assets in some fashion would go a long way towards establishing the credibility of the stress tests. One possible solution is to apply losses to HTM assets with maturities beyond 2012 or 2013, on the premise that the EFSF-provided liquidity has taken near-term defaults off the table. This would limit the double-counting aspect (as the EFSF funds are sufficient to provide liquidity to the peripherals for 3-4 years only) and apply haircuts to the sovereign assets most likely to experience a loss.
Like we said, Barclays has a very obvious axe.
While the balance of the report is somewhat trivial, Barclays ends by recaping the circumstances of the US stress test, which resulted in the biggest bear market short covering squeeze since the Great Depression. Obviously the firm is trying to create the self-fulfilling prophecy, that is at the very base of the "Stress Test", that once these pass, the market has to surge. We only hope the ECB is as good at reinforcing market "upward bias" as the Fed was in early-mid 2009, when it did everything and anything to push stocks to ridiculous levels.
The US stress test was a significant positive catalyst for the market, and we believe it is worth reviewing what made it successful. Of course, when the US went through this exercise, “green shoots” were just beginning to arrive in the economy, and spreads in the credit market were wider, but the example set by the US process still provides a useful framework for comparison. In Figure 12, we provide a timeline of key dates for the US stress test and average 5y CDS spreads for the large banks.
The release of the US stress test results was a catalyst for spreads. The average 5y CDS of the large banks tightened 50bp in the two days leading up to the release and then tightened an additional 50bp in the two days following it. This was the beginning of a rally that lasted the rest of the year. Three factors that we believe contributed to the success of the U.S. stress tests were transparency, timely and credible capital plans, and a government capital backstop.
As can be seen above, total capital raised directly and indirectly to satisfy our own Stress Test Farce was far less than the $109 billion Barclays estimates will be needed in Europe. This is not a trivial amount, yet with the bulk of asset managers already positioned bullishly ahead of July 23, in hopes that the event will be a catalyst to a comparable bounce as seen in the US, we fear the contrarian play may, as usually ends up happening, be the proper one in this case.