The stress tests for which the Fed announced results this week have received ample publicity. The publicity seems all to be about which banks passed and which did not, as well as whether the tests were sufficiently stringent to ferret out whether the banks were strong enough to survive bad economic times. Most of the commentary missed the most important consequences of the stress tests, which are twofold:
(1) To create a true risk-based capital system for the first time. And
(2) Effectively to induce the banks to use capital allocations to their different businesses. This will cause banks to pay closer attention to real risks of each business, and quite likely will change the way they conduct many types of lending to customers and trading with counterparties.
Long-term, this development will make bank earnings more stable, will make bank lending less pro-cyclical, and will be good news for investors in bank preferred stock and subordinated debt.
Stress tests were begun in 2009, admittedly with the goal of reassuring the public about the condition of American banks. The tests now have been institutionalized, and the banks themselves will in future have to conduct them, using methodology that they must convince the regulators is sound and sufficiently stringent. The banks also will have to create and publicly announce capital plans for any shortfalls that the tests reveal.
A consensus formed around the idea of risk-based capital in the mid-1980s. That resulted in the Basel I standards that were adopted by the BIS in 1988. The risk ratings used in Basel I were pretty primitive, and they were immediately criticized on that ground. But how to create better risk ratings remained a subject of contention. Eventually, in 2004 the BIS adopted Basel II to replace Basel I for large banks. Basel II permitted the banks themselves to model their portfolios and businesses and to determine their needed capital levels for themselves, subject to regulatory review of their methodologies. Not surprisingly, the banks ran rings around the regulators under Basel II, with the consequence that when the Great Recession came a-calling, the banks and investment banks were significantly undercapitalized. Basel III, a better version of Basel I, now has been adopted and is in the process of implementation. The Fed's stress test regime basically leapfrogs the Basel III process and, in my view, makes it unnecessary for American banks.
The Reserve for Loan Losses
The stress test methodology also cures one of the biggest holes in American bank safety methodology: The historically based allowance for loan and lease losses (ALLL), better known as the reserve for loan losses. The ALLL now can be whatever it is. The meaningful number will be the projected losses that result from the stress test, since it is those losses that will set the bank's capital effective requirements.
I do not know what the SEC is going to make of this new situation. In the past, the SEC would not permit the banks to use forward-looking tests to determine how much to reserve in the ALLL. The SEC made this ruling because it said that the forward-looking data tended to smooth bank earnings, which, as I have said many times, is exactly what the ALLL is supposed to do. Now the ALLL-and indeed substantially all of a bank's backward-looking historical financial statements-will be taking a back seat to the forward-looking stress tests, and it will be the disclosure of the stress test results that principally will influence both management and investor conduct.
All that may seem like technical yada-yada to many investors. But it is anything but yada-yada. It is changing bank management styles. It is changing the way banks influence the economy, both on the upside and on the downside. And it should make the banking world much safer for dividend-investors, including preferred stock investors.
The stress tests tell banks where their real risks lie. That is, if the stress test shows that a particular form of lending will take significant losses under stress, the new rules will require a capital plan to address the consequences of such losses. That in turn shows the bank that it needs to allocate more effective capital to that business or figure out how to reduce the risk. The recently-announced stress tests showed, for example, that credit cards will incur substantial losses in a severe downturn. That is very likely to induce managements to be more conservative in their credit card lending in order to avoid the large capital charges. Indeed, as I reported a few weeks ago, that is already happening. Several major banks are cutting back on the size of their credit card portfolios in favor serving good customers rather than customers on whom they make money by earning late fees.
Lord Turner, the U.K.'s leading financial regulator, has called for "minimum haircuts and margins for 'repo' and other secured lending to minimise their pro-cyclical effect in falling markets." Lord Turner is worried about the "shadow banking" system and wants to regulate it. But would it not be better for stress tests to prod the market to see the real risks and police itself on that basis? I believe that will happen in the U.S. and that the effect will, indeed, be counter-cyclical.
I anticipate that the skeptics in the audience will assume that the banks will game the stress test regime just as they have practically every other regulatory initiative. Why is the stress test regime any different from Basel II, they are wondering? My Panglossian view will be shown up, over the longer term, as naïve, they are thinking. No, I aver. I began studying U.S. bank regulation in 1966, and this is the first time in at least 20 years that I have thought something actually was going to work. And it is going to work because it utilizes the securities laws to reinforce the regulatory regime, and it is avowedly forward-looking based on fairly specific economic scenarios. Basel II permitted the banks to use their own assumptions and metrics whereas the stress tests do not.
I do not know whether the Fed knew what it was doing when it designed the stress test process, but for the first time, it puts bank managements at risk under the securities laws if they do not make honest assessments under the stress tests. I believe that the relative ease of enforcing the securities laws, as compared with the bank regulatory laws, will make a vivid impression on bank managements. It will make bank managements more conservative in their lending and risk-taking.
The impact of the stress tests, over time, therefore, will be to depress economic booms and, as a consequence of depressing the booms, to tend to prevent economic busts.
The macro-economic and management impacts of the stress tests are so beneficial that probably they should be required of any form of financial business that has regulatory capital requirements. I might even go further to suggest that the SEC might require that non-bank financial companies stress test their capital positions, as well. Sunlight, they say, is the best disinfectant.
If I am correct, bank earnings will be less prone to booms and busts. And steady earnings benefit dividend investors. Steady earnings also protect mezzanine investors, such as owners of preferred stock and subordinated debt. I am not going to recommend a particular issue of preferred stock or subordinated debt, since there are so many, but I am recommending that investors with a modest appetite for credit risk but a fairly healthy tolerance for interest rate risk should consider the preferred stock and subordinated debt of the major banks that passed the stress tests. These include JPMorgan Chase (JPM), Goldman Sachs (GS), Wells Fargo (WF), US Bancorp (USB), and even Bank of America Corp (BAC). The returns run from the 8% area for the riskier BAC issuances to 6% for the less risky WF issuances. In an era when money markets earn nothing and long-term Treasuries earn maybe 3%, those are nice yields for companies that seem likely to maintain their ability to pay dividends even in times of "stress".