What Would Higher Bank Capital Requirements Really Do To Lending?
In more depressing news, I see that international bank regulators continue to have learned nothing from the financial crisis that kicked off the Great Recession. While we can continue to hope that American regulators are tougher, I wouldn’t hold our breath. No individual jurisdiction is going to want to lose a lot of banking business by having tougher requirements than other major banking destinations.
Just to recap what the Basel decision was about: there are broadly speaking two issues at play: whether we want banks to hold more Tier 1 (equity) capital in general, and how we want them to calculate what their assets and liabilities are (how easy it is to move assets off-balance sheet and how much you can use derivatives to reduce what your “actual” net position is that you have to hold capital against).
The thing to remember about both issues is that the financial crisis was caused substantially by banks losing enormous amounts of money on very highly rated assets that, in many cases, were hedged with highly-rated counter parties. Specifically, banks held on to the AAA-rated senior position in securitized pools of loans of various kinds (particularly residential mortgages), and hedged those positions by purchasing a derivative contract from an insurer. This resulted in an income stream that was very small in percentage terms, but potentially very large because of the sheet size of the transactions, and potentially very lucrative because you had to hold almost no capital against the hedged position.
If we don’t want banks to pursue that kind of transaction, we need to charge enough capital against hedged positions to make hedging a second-best alternative to liquidating in most situations; we need to be skeptical of basing capital requirements primarily on ratings (or on historic performance, particularly for asset classes without a lot of history). The minimum leverage requirement was an attempt to cut the Gordian knot by saying, in effect, if you figure out how to game our other rules for calculating your capital requirement, at a minimum you’ll still need to hold at a certain minimum amount of equity capital, and you won’t be able to play games with derivatives to increase your leverage further by making it look like you have a smaller balance sheet than you do. The folks in Basel just decided that, actually, you should be allowed to play rather more games with derivatives than one might have hoped if one cared about actually avoiding a repeat of 2008.
I do want to tackle in a bit more depth the point Matt Yglesias makes here, about the effects of higher capital requirements on bank lending. Yglesias says:
The big lie of leverage regulations is that a strict rule could “curtail lending” and that laxer ones will “keep lending flowing to the economy.” This is simply not the case. The issue isn’t what kind of lending (or, more broadly, investment) banks can do. It’s how can they finance that lending—specifically, to what extent new lending will have to be financed with new borrowing versus cash. Despite their whining, banks do not actually face serious logistical difficulties in obtaining cash to fund investment. They can get it the same way most companies do: by recycling past profits. In a pinch, they can issue new shares to raise cash, as Facebook did last month. From a shareholder’s viewpoint, it’s better to have past profits flow directly into your pocket as dividends, rather than have them recycled as investments. But there’s no reason you would want a bank to actually forgo a profitable investment opportunity simply because it had to finance it one way rather than another way.
The only universe in which debt financing should lead to more investment than non-debt financing is a universe in which some of the lending being financed is unsound, and creditors are only willing to lend because they’re counting on bailouts.
That’s not exactly right, and it’s worth explaining why, and thinking about what higher requirements would do to lending.
Different tiers of capital demand different levels of return, because they expose the investor to different levels of risk. Equity investors demand higher returns than debt investors, and mezzanine debt investors demand higher returns than senior debt investors. Now, if you make banks hold more equity, that equity arguably becomes less risky, and that should prompt investors to demand a lower return. But investors have their own benchmarks to hit, and they don’t have to put money into banks – there isn’t any mechanism for “forcing” investors to do so. Banks need to generate returns that make them competitive with other uses for investor capital. Higher capital requirements will make it harder to generate those returns.
This is why people argue that higher capital standards will curtail lending. They assume, logically enough, that what banks will do in response to higher capital requirements is curtail any lending that, based on the new leverage ratios, doesn’t generate an adequate return on capital. So there will, in aggregate, be “less lending” going on. Another way of putting it is that average credit spreads will go up because banks need to get a higher return on their lending to maintain the same return on a larger equity base, and that for some borrowers the higher credit spreads will make their businesses no longer profitable.
Nobody is arguing that bankers simply won’t have the money to lend. They’re arguing that lending won’t be sufficiently profitable for their equity investors. The easiest way to make it clear that capital ratios could affect lending is to ask what would happen if banks could not borrow at all. Does anyone believe that a 100% equity requirement wouldn’t result in a massive curtailment of lending?
But all lending is not equal, as Yglesias points out at the end. His mistake is assuming that everybody knows what kind of lending is unsound, and that creditors engage in unsound lending simply because they expect to be bailed out. It’s not at all that simple.
The large banks did not build huge positions in mortgage derivatives because they expected to be bailed out. The built these huge positions because they were profitable and their own models told them they were taking reasonable and calculable risks. Those models were wrong, and all future models will be wrong in the same way because by definition risk of this sort is not stable.
But if you are terrified of taking on unknown and unknowable risk, then you are going to prefer the highly-leveraged investment whose risks you think you can measure and model to the less-leveraged investment whose risks are extremely hard to assess. Assuming both trades have a similar return, you will prefer to buy AAA-rated mortgage-backed securities, hedge them with a AAA-rated insurer, and lever the income stream 1000-to-1, rather than to lend to a Rwandan shoe factory and hold the position without leverage, because you can assess the risks of the former more readily and quantitatively than the latter.
This is basically what the global banking industry did throughout the 2000s. To a considerable degree, it’s what they are still doing – and the regulators continue to encourage them to do this, because regulatory capital is assessed based on models that privilege assets that can be modeled, notwithstanding that we all know how those models fail catastrophically, over and over again, in major financial crises.
The hope is that a firm minimum leverage ratio would not reduce lending across the board, but would specifically reduce the attractiveness of these sorts of apparently super-low risk trades whose risks are, in fact, hidden way out in the tail of the distribution where financial crises lie. The loan to the Rwandan shoe factory, a presumably unrated company, would already require so much equity capital against it that the minimum leverage requirement wouldn’t have any effect. But the 1000-to-1 leveraging of the mortgage-backed securities would no longer generate an adequate return on capital at a 10-to-1 leverage ratio.
So the larger hope is that, in a new environment where you can’t just use leverage on the same old assets to generate higher returns, banks will have to go looking for those Rwandan shoe factories, and will have to take more risks that are harder to assess but that clearly lie closer to the middle of the distribution rather than way out on the tail. The actual asset pool that banks hold, in other words, would look more risky – because the only way to generate return is to take risk. But the hope is that it would also be tilted more toward “real” businesses. Overall lending may indeed drop. But the hope is that productive lending will actually go up.
Do we know that the Rwandan shoe factory is a good risk? No, we do not. We also didn’t know that all the broadband communications infrastructure that was built out in the 1990s was a good bet. Some of it wasn’t – WorldCom turned out to be a fraud, after all, and a lot of its competitors went bankrupt when the market realized they were benchmarking to a fraudulent competitor. But at least we got new broadband communications infrastructure out of that particular bubble. What did we get from the mortgage bubble? A lot of houses in California, Nevada and Arizona that nobody wants to live in.
The internet bubble was largely equity-financed. Debt played only an ancillary role. Better bank capital rules won’t prevent speculative bubbles. The hope, though, is that they’ll redirect lending out of unproductive areas and towards areas that could have a real positive impact on the global economy.