There are serious proposals to force banks to fund themselves with considerably less debt and far more money from their shareholders. This would protect the rest of us, by leaving more of the risk with shareholders and reducing the potential need for taxpayer bailouts. However, there is a trade-off for the greater safety; loans would become more expensive and the economy would slow.
The added safety is well worth the cost when raising equity levels from the risky pre-crisis levels to those being mandated by global regulators under the “Basel III” rules. It might be good to go somewhat further, but not to the extreme levels advocated by some. My fear is that drastic actions may be taken in this area because some argue that it would be economically costless to do so. This idea is wrong in the real world, even if it makes sense under very specific theoretical conditions. There is only space in this column for a high-level discussion of this complex topic. Please see my recent paper on bank capital requirements for a somewhat more detailed explanation.
US banks currently fund about 5% of their assets with money from their common shareholders (“common equity,” one part of the safety buffers known as “capital”), with the rest coming from depositors, bondholders, and a few other sources. This is more than double the pre-crisis levels and is only modestly below the Basel III requirements. Some have called for increasing the level to as much as 30%, a drastic change that would be costly for the economy.
At first blush, it seems obvious that selling stock to investors who want returns of 10-15% a year would increase a bank’s costs, and therefore its loan rates, as compared to borrowing from bondholders or depositors who charge far lower rates. However, the economists Modigliani and Miller won the Nobel Prize in part for showing that, under idealized conditions, it does not matter what proportion of a firm’s funding comes from equity rather than debt. Adding more equity makes a firm less risky and reduces the cost of each unit of equity or debt by an amount that exactly offsets the switch to an otherwise more expensive mix of funding.
This fundamental theory of finance is the core reason some theorists and their followers argue that there is no economic cost to forcing banks to fund themselves much more through common stock. However, there are at least 6 differences between the real world and the idealized conditions necessary for Modigliani-Miller to hold. Taken together, these imply substantial societal costs to mandating extreme levels of equity.
Tax advantages for debt. Modigliani and Miller ignored corporate taxes in their initial work. In reality, interest payments on debt and deposits are tax deductible, while dividends to shareholders are not, creating a major incentive for banks and other firms to fund with debt. Miller later showed that tax advantages at the investor level for owning stock could work in the opposite direction, and would fully eliminate the corporate tax effect under very specific conditions that are not met in the US tax system. The actual offset in the US is perhaps a 50% reduction, maybe less, still leaving taxes as a big factor. This does mean tax collections would be higher, so the net effect on economic growth would depend on what was done with the extra money.
Many advocates of extreme levels of equity call for the abolition of interest deductibility. The same relative effect could be achieved by giving banks a tax deduction on their dividends, as Belgium does. For better or worse, neither of these things is likely to happen, so bank funding costs would go up and some or all of this would be passed on to borrowers. Advocates of extreme capital ratios should offer their proposed back-up plans if interest deductibility is not abolished.
Deposit guarantees and other backstops. Bank deposits are guaranteed up to certain limits, and some argue that federal policies provide protection to uninsured deposits and bank debt through implicit guarantees. Guarantees of debt and deposits block the key mechanism of Modigliani-Miller, since there is little reason for funders with guarantees to lower what they charge as banks become safer. A perfect risk-based pricing system for guarantees would offset the behavioral effect, but we do not have this in practice and are unlikely to achieve it, for both political and technical reasons.
Issuance costs. Modigliani-Miller ignores the transactional costs of raising funding. In practice, the direct issuance costs for equity are much higher than for debt or deposits, although still not huge in the grand scheme of things. More importantly, investors insist on a significant price discount if a firm wants to sell them stock, out of a fear that management knows of reasons why the share price should be lower and therefore is seizing an opportunity to “sell high.” Modigliani-Miller ignores both effects. Recognizing this, some advocates of very high equity levels are willing to allow banks to meet the requirements very gradually through retaining all profits, in exchange for a ban on dividends and share buybacks. This largely eliminates the problem of issuance costs, but would create major market distortions that would potentially last for decades, as some banks would build up their equity levels faster than others and therefore operate with a different, and more expensive, cost structure. There could also be substantial disincentives to increase lending, if doing so would require equity issuance to avoid lowering the equity ratios. If there are no such requirements to maintain equity ratios, then there would be the opposite incentive to increase lending sharply to restore the bank’s lower preferred equity ratios, undoing the effect of setting higher requirements.
Skepticism by investors. Many investors and equity analysts have made clear their skepticism that adding equity increases bank safety as much as the theory says it would. Actions by managements or mistakes by regulators could counteract the positive effects, at least partially. Banks are always going to be “black boxes” to some extent, so there may be a limit to how much investors are willing to drop their required return. Nor is there clear historical evidence to refute the concerns about a partial offset due to investor skepticism. As long as significant numbers of investors are skeptical, the price of equity and debt will not go down to the extent that Modigliani-Miller assumes as banks raise more equity. This will put pressure on financial institutions to avoid operating with the higher equity levels.
Shadow banking. The higher costs that would be imposed on banks because of these real world issues would create strong market pressure to move business out of the highly regulated banking system into various forms of shadow banking. Dodd-Frank has given regulators some powers to deal with shadow banking, but nothing like the authority that would be needed to counteract this level of market pressure. In practice, fully counteracting this pressure may be impossible without rigid government controls that would harm the economy in themselves. Few, if any, analysts believe we would be better off with a massive shift of banking activity into shadow banks. A financial system that relied primarily on shadow banking would be much more vulnerable to crises that would shake the wider economy.
Transition issues. As already noted, there are a host of issues of how to get from here to there without damaging a still fragile economy.
In sum, higher equity levels at banks increase the safety of our financial system in important ways, but we should not overshoot, as there are real costs that we must balance against the benefits.