Why government guarantees are a doubleedged sword
12 Feb 2009
Among many factors that contributed to the subprime crisis, a factor as important as any other was leverage – the astonishingly high leverage at which banks operated prior to (and in some cases, even during) the crisis.
Why were banks so highly levered? Professors Viral Acharya and Julian Franks have sifted through the list of possible culprits and highlight the role of regulatory guarantees of bank debt. The scope of these guarantees has only enlarged as part of the recent rescue packages, covering not only a larger fraction of deposits but also extending to uninsured public debt. At this delicate stage of the crisis, it is prudent to reflect on the role of such guarantees in encouraging banks to take on high leverage. While the world is rescued, we cannot sow the seeds of another crisis.
Bank debt is remarkable. It can be as simple as deposits and inter-bank loans or as complex as contingent liabilities through over-the-counter derivatives. Bank leverage tends to be large relative to assets, when compared to balance-sheets of non-banking corporations. Most remarkably, its market cost is extremely low in good times and remains so over a wide range of leverage ratios. Banks with leverage to assets ratios of 80 per cent are able to borrow in good times at just 0.20 per cent (20 basis points) above what governments can borrow at and so can banks with leverage ratios of 90 per cent. In contrast, the cost of debt for unregulated, non-banking corporations appears to rise steadily with leverage ratios.
What explains this difference between banks and other corporations? The answer lies in government guarantees.
While a large part of bank deposits is guaranteed, explicitly so for commercial banks, much of the remaining debt is implicitly guaranteed too. The presence of guarantees means that as bank leverage and default risk increase, the true cost to the provider of the guarantee (i.e., the government) rises, but the cost to the bank does not.
As a result, banks have a free option to increase leverage to extraordinarily high levels. This would of course not be such a problem if government "marked to market" the price of guarantees, effectively getting banks to pay a fair price commensurate with their risk taking, but this is typically not done. Compounding the problem, the high leverage that guarantees encourage inevitably leads banks to gamble on expansionary monetary policy by going down the quality curve.
Rocks and hard places
Of course at some point the low quality loans default, banks make losses and lose a part of their capital. Bank share prices fall reflecting their high leverage ratio and vulnerability to small changes in the value of assets. In turn, uninsured depositors and wholesale creditors start to worry about the likelihood of the government extending guarantees, and the market cost of bank debt spirals upwards.
A case in point here is the "bank run" at Northern Rock in the United Kingdom, and the fact that its problems caused a contagion primarily for those banks (Alliance and Leicester, Bradford and Bingley and HBOS) that had greater reliance on uninsured commercial paper. These banks were accorded higher valuation multiples by markets in good times, but once risks rose, the conditions were reversed. Banks like HSBC that had greater reliance on deposits have been much less affected, and have not even felt the need to participate in the U.K. government bailout.
Might it be that bankers, by taking on such high leverage, were simply maximising shareholder value, given there are mispriced government guarantees, and that they simply got unlucky? We do not think so. Instead, we believe that capital budgeting at banks is in fact broken.
First, the relatively flat cost of debt in good times gives bankers the illusion that their cost of capital in bad times will also be low, or in other words, that their funding costs will not rise even with extreme leverage and high business risks.
Second, the extremes of leverage encourage bankers to place little weight on the cost of equity. When leverage is as high as 95 per cent or more and its cost relatively flat, the cost of equity hardly matters. But this is true only until banks find themselves in a crisis and must issue equity at punitive dilution costs.
Third, there is excessive risk-taking at banks also induced by myopic compensation packages tied to past year (quarter)'s accounting profits rather than long-term return on assets. No bank (board) wants to deviate from such packages as it fears losing employees to competitors. It is essentially a "race to the bottom" in the governance of a highly competitive sector.
In absence of guarantees, banks taking on excessive leverage and risks would face steep costs of funding, be it debt or equity. Creditors, for instance, would monitor and discipline bank management in good times rather than leaving that task to regulatory supervision. Alas, such monitoring largely disappears when so much of bank debt is explicitly or implicitly guaranteed, and market discipline only imperfectly replaced by mispriced deposit insurance premium and coarsely designed capital requirements that bankers can easily game.
It is thus ironic that the very guarantees that induced banks to take on excessive leverage and risk, and endanger our jobs and savings, must be sharply increased to get us out of the current mess.
Regulators ought to remind themselves that guarantees are a double-edged sword. They are inevitable in systemic crises, for political reasons as well as for efficiency. But somewhat unfortunately, they linger even after crises abate and their pernicious effects on bankers' incentives remain unchecked.
The task of fixing capital budgeting at banks is not just for bankers. Only if regulators charge suitably for the guarantees will banks price them in to their loans and leverage decisions. Resolution of the most severe financial crisis of our lifetimes may otherwise soon turn out to be a Pyrrhic victory.
Viral V. Acharya (firstname.lastname@example.org) is Professor of Finance at London Business School and New York University - Stern School of Business. Julian Franks (email@example.com) is Professor of Finance at London Business School.