Economics focus

Buffer warren

Oct 29th 2009
From The Economist print edition

Why are banks so averse to raising equity?













Illustration by Jac Depczyk








THE usual laws of corporate finance do not seem to apply to banks. Almost all big industrial companies—and decent analysts of them—are subject to a tight mesh of proven rules, backed up by decades of financial theory. Everyone agrees, for example, that accounting values are often flaky and that cashflow matters most when valuing a firm or trying to work out if it might go bust. The profitability of any activity, too, must be assessed before the magnifying effects of leverage are taken into account. In bank-land, however, anything goes. Accounting, not cash, is king. And most common yardsticks for measuring performance are all in some way distorted by leverage, not least return on equity (ROE).

The peculiarity of the banks is not some arcane matter. Regulators are furiously trying to find ways to prevent taxpayers picking up the tab for banking crises. The latest bill passing through Congress aims to hit the industry for the cost of bail-outs, for example. Their main weapon, however, is forcing banks to have bigger equity buffers. Bankers complain that equity is too expensive and will have a knock-on effect on the price of credit, damaging the economy. But this contradicts a cornerstone of corporate finance, set out by Franco Modigliani and Merton Miller in 1958, that a firm’s value is unaffected by its capital structure (at least in a perfectly efficient, tax-free world).

This theory says that although equity owners demand a higher return than creditors, their required rate of return on each unit falls as the amount of equity rises, since profits after interest become less volatile. The cost of debt falls too, since creditors have a bigger buffer beneath them. The firm’s blended cost of capital is unchanged, and is driven largely by the risk of the firm’s assets, not how they are paid for. At the extremes, it can be very low. A firm that owned only government bonds yielding 5% would have a cost of capital of just 5% even if entirely equity-financed.

There are quirks in the real world, not least the tax-deductibility of interest costs, which give debt an advantage. A lunatic fringe, populated mainly by private-equity types, thinks this is a licence to gorge on debt, but most serious companies are mainly equity-financed and were not tempted by the credit bubble. Why, then, are banks different? By their very nature some of their assets are funded by deposits rather than by equity. A century ago, when banks could fail, they carried equity buffers of about twice today’s levels in order to reassure customers. Since then deposits have become insured in government-backed schemes. Other creditors now enjoy a near-explicit government guarantee too.

It is not just guarantees which make deposits special. They are also priced off central banks’ short-term interest rates. Furthermore, with the promise of unlimited liquidity provision from governments, banks (unlike normal companies) do not have to worry much if they have loads of other short-term debt that constantly needs to be refinanced. Along with the guarantees, the very short-term nature of deposits and debt means their cost for banks is extraordinarily low when central banks’ base rates are low. America’s mega-banks typically paid a blended annual interest rate on borrowings and deposits of 1-2% in the second quarter. Equity cannot compete with that. Banks’ privileges mean the rules on cost of capital break down.

Because banks are too big to fail, it is too expensive for them to raise more equity to become safer. If that sounds circular, it is. Once society underwrites banks, the choice is between big equity buffers that push up the price of credit but make things safer, or smaller buffers and periodic bail-outs. Which is best? Some of the predictions made by banks about the impact of more equity on the cost of credit are silly. Take a bank that replaced long-term debt with equity in order to raise its ratio of equity to risk-adjusted assets from 10% to 15%. In order to keep its ROE constant at, say, 15%, it would have to boost the interest rate it earned on its assets by only 0.5 percentage points. And since a bigger equity buffer would make profits less volatile, investors should actually accept a lower ROE, mitigating part of the impact. What is more, the sheer havoc and expense of crises means it is worth paying to avoid them. A new study commissioned by the Financial Services Authority, Britain’s bank regulator, and conducted by the National Institute of Economic and Social Research, suggests more capital would marginally boost Britain’s wealth.

If you take me off life support, I’ll kill you

That all points to more equity, but what is the right amount? Ideally enough to absorb another meltdown while still retaining a decent buffer. Using the projections from the Federal Reserve’s “stress tests” in May, America’s biggest banks combined would have needed an 8% ratio of equity to risk-adjusted-assets going into the crisis to have avoided breaching the permitted floor of 4%. Most big Western banks are now at about 8%.

In a meltdown, however, a few outliers lose far more than the average. UBS would have needed a ratio of about 20% to have avoided breaching the 4% floor. Forcing all banks to carry enough equity to survive this scenario is one option. But there is a less dramatic alternative: to force these extra losses onto creditors of the outlier banks. Last time round this was impossible without causing the bank to default. Next time the solution may be to require banks to carry a layer of convertible debt, as proposed by the Squam Lake Group of American academics, which includes Raghuram Rajan, a former chief economist at the IMF. The layer does not need to be vast. If UBS had entered the crisis with an equity-to-risk-adjusted-assets ratio of 8%, it would have needed to convert about SFr33 billion ($32 billion) of debt into equity to have remained above the 4% floor, compared with total debt issued at the end of 2007 of SFr222 billion. But it does need to be distinct, so that regulators can force it to be converted into equity without causing the firm to default. That might not make banks look more normal, but it might at least make them safer.

Copyright © 2009 The Economist Newspaper and The Economist Group. All rights reserved.