The True Purpose of Bank Capital
The objective of most banking regulation is straightforward: Taxpayers should never have to choose between enduring a deep recession and having to bail out money-losing financial firms.
The actual rules, however, are extremely complicated. Just try reading the proposals from the Basel Committee on the assets banks hold and the ways banks fund those assets. While every approach has its drawbacks, the current framework is far too complicated and ripe for gaming. If nothing else, it's a needless subsidy for finance lawyers.
I'm reminded of all this because, on Tuesday, the Federal Reserve voted on new rules governing the minimum required ratios of bank equity to bank assets as part of its effort to become compliant with its Basel commitments. The complete document is 972 pages. Few have read the whole thing -- I personally gave up after about 100 pages -- but summaries prepared by the Fed are available here and here. (The Federal Deposit Insurance Corp. and the Office of the Comptroller of the Currency will issue their own rules next week.)
Banks now have to keep track of lots of different ratios. There is the minimum ratio of common equity tier 1 capital to risk-weighted assets (4.5 percent), the ratio of total tier 1 capital to risk-weighted assets (6 percent), a ratio of total capital to risk-weighted assets (8 percent), a ratio of tier 1 capital to "average total consolidated assets" (4 percent), and, for banks subject to the Basel III "advanced approaches," a separate leverage ratio of 3 percent.
The point of these rules, as Bloomberg View's editors have explained, is that equity can absorb losses without the need for bankruptcy or complicated debt restructurings. More equity also means smaller subsidies for "too big to fail" institutions.
Bankers, however, dislike equity funding because of the way they are compensated. They argue that every increase in equity capital requirements restricts credit growth and holds back economic activity. It's hard for me to see how this could be true. (Some critics make the more sophisticated point that a reduction in bank debt will cause the supply of broad money to contract, but this problem can be solved with more government debt.)
For policymakers, the hard part is figuring out the right amount of equity needed to protect the rest of us from mistakes made by financiers. I've previously argued that a 100 percent equity/assets ratio is the best way to ensure that the financial system can't blow up the economy. Among the advantages of this proposal is its relative simplicity.
Policymakers, however, have been reluctant to embrace the radical insights of Irving Fisher, Milton Friedman and James Tobin, instead preferring to make as few tweaks as possible to the existing system. This may be because they genuinely believe the arguments of the banking apologists, who love to speak of ambiguous "tradeoffs" between stability and growth, or simply because policymakers often view the perfect as the enemy of the good. Whatever the reason, the unfortunate result has been the complicated mess known as the Basel Accords.
The biggest debates aren't about the required ratios of capital to assets but the definitions of "capital" and "assets." Capital is supposed to be the part of the balance sheet that can absorb losses without causing creditors to have a panic attack. A straightforward analysis would lead many people to conclude from this that capital should only be defined as tangible common equity. Yet many other types of funding can be counted as capital, including preferred shares (which guarantee the amount of dividends investors get paid, although not the timing), subordinated debt securities and bonds that "convert" into equity when the regulatory capital ratio falls below a certain threshold.
Defining "assets" is also pretty tricky. (And I'm not even going to get into the question of how you measure lending commitments that aren't on the official balance sheets.) Right now, bankers and regulators engage in a practice called "risk-weighting." This is based on the sensible intuition that certain assets are more likely to lose value than others. Since capital is supposed to absorb losses and protect secured creditors (depositors, repo counterparties, etc.), it's reasonable to think that "safer" assets ought to require relatively less equity funding than "riskier" ones.
The problem, however, is in the details of implementation. No one can know which assets will be riskier than others in advance. Moreover, the assets often perceived as "safest" generally are the most vulnerable to sudden large declines. Just look at the two biggest casualties of risk-weighting over the past few years: subprime mortgage bonds and European sovereign debt. Both were considered low risk, so many firms bought as much of them as they could with as little equity financing as possible.
Amusingly enough, the Bank for International Settlements recently found that different banks calculate the same risks very differently. That could actually improve the stability of the broader financial system if it means that firms aren't all doing the same things. However, it also means that regulatory capital ratios are unhelpful at telling us whether a bank is sufficiently capitalized.
Hyun Song Shin, an economist at Princeton University, has identified a broader problem with risk-weighting: it is "procylical." Most banks' internal models calculate the regulatory capital needed to fund each position by looking at the volatility over the past few years. Since asset prices generally rise when volatility is falling, this means that, paradoxically, bank risk models encourage traders to take more risk when securities are getting more expensive. By contrast, falling prices are generally associated with rising volatility, which means that bank risk models often encourage excessive asset sales. Surely there must be a better way to do this.
The problem of risk-weighting could be overcome with a sufficiently high capital requirement, especially since the ratio of total assets to risk-weighted assets is usually between 3 to 1 and 2 to 1. Another alternative is to give every asset a risk weight of 100 percent. In this case, the ratio of equity to assets isn't called a capital ratio but a "leverage ratio." (Go figure.) This is appealing, especially as a supplement to the existing -- and complicated -- risk-weighted approach. However, it too can be more complicated than it seems.
Consider the case of derivatives. U.S. Generally Accepted Accounting Principles say that offsetting derivatives exposures should cancel each other out. Under this regime, banks can avoid counting these exposures when calculating their capital requirements. Other countries prefer to use the International Financial Reporting Standards, which add together offsetting derivative exposures rather than subtract them from each other. This difference between U.S. GAAP and IFRS accounting leads to substantial variations in the official sizes of bank balance sheets. (Personally, I lean toward the view that it makes more sense to net out offsetting derivatives than to add them together, but there are some good reasons to prefer the IFRS methodology.)
The rules around banking are complicated. Maybe they need to be this way, but I seriously doubt it. After all, the only people who seem to like the status quo are the banks and their lawyers.
(Matthew C. Klein is a contributor to the Ticker. Follow him on Twitter.)