Fear of a downward credit spiral hangs over America’s economy. The script is well known. Rising losses from the mortgage mess make banks less willing to lend, which weakens the economy’s prospects, which puts further strain on the banks. But until now most of the estimates of the likely cost of the crisis look as though they have been plucked from thin air. A paper prepared for a recent meeting of the US Monetary Policy Forum, a gathering of Wall Street economists, academics and central bankers, is far more thorough.
The study begins by estimating the size of mortgage-related losses using three different methods. One extrapolates from the defaults seen in different mortgage vintages to date; a second calculates the losses implicit in the prices of credit derivatives linked to subprime mortgage-backed securities as measured by the ABX indices produced by Markit a third draws on the experience of foreclosures in previous regional housing busts in Texas, Massachusetts and California. Each method involves some heroic assumptions. With house-prices falling nationally, foreclosure rates may rise far higher than they have in the past. By contrast, the ABX prices may be reflecting underlying risk and illiquidity as much as expected losses. Strikingly, however, all three approaches yield similar results: that mortgage-credit losses are likely to be around $400 billion.
That is a large number, but it is no worse than the losses that can be suffered on a bad day on Wall Street. The reason that the macroeconomic consequences are likely to be much bigger is that many of these losses will be born by banks and other leveraged financial institutions that hold approximately half of all outstanding mortgage debt in America. If the losses are spread evenly (a big if), that suggests America’s banks are likely to take a hit of some $200 billion.
How much would such a loss dampen lending? The answer depends on how easily banks are able to rebuild their capital, and how far they scale back lending in response to weaker balance sheets. The authors reckon that the leverage ratio for financial institutions as a whole is approximately 10:1, so that one dollar of capital translates into $10 of lending. They assume that banks counter half their $200 billion losses with new capital infusions and reduce their leverage ratios by 5%. Estimating total assets at $20.5 trillion, the authors calculate that net lending to businesses and households would shrink by some $910 billion—several times the size of the credit losses.
The final piece of the puzzle is to estimate what effect such a drop in lending would have on the overall economy. If capital markets were perfect, there would be no effect. Instead of borrowing from banks, those in need of funds could issue securities. In the real world, however, a drop in bank lending does affect overall access to finance—all the more so in this crisis, when the lending drought coincides with a sharp drop in securitization. Historically, GDP growth has indeed been correlated with growth in debt, though it is not clear whether weak credit supply slows the economy or a weaker economy slows the demand for credit. The study tries to sort out the causality by using the Federal Reserve’s survey of loan officers as a proxy for the supply of credit. Based on these results, the paper estimates that a $910 billion drop in lending would drag down GDP growth by 1.3 percentage points over the following year.
That is a sizable hit, and it comes on top of the other problems facing America’s economy, such as the drag on consumer spending from falling housing wealth or a weaker job market. What’s more, this study looks only at mortgage losses (ignoring, for instance, losses from defaults on credit cards or higher corporate defaults). The news is already sombre: but the eventual costs could be a lot higher.