Buttonwood’s notebook | The debt crisis

Wipeout

A book calls for the writedown of consumer debts and the resdesign of debt contracts

By Buttonwood

DEBT forgiveness, and the redesign of debt contracts to involve more risk-sharing, is the answer to the problem of recurrent financial crises. That is the argument of two economists, Atif Mian and Amir Sufi, in their book "House of Debt" (our Free Exchange column discussed the book here).

They argue that economic recessions often follow surges in household debt. They dub their thesis "the levered losses framework"; it is based on the idea that equity losses do not affect homeowners equally. The rich tend to have a lot of equity in their homes; the poorest homeowners tend to have slivers of equity (10-20%). When house prices fall, this wipes out their stake and may force them into foreclosure. The authors illustrate this with some neat statistical analysis of individual areas of America, showing that places where the average equity stake was low suffered significantly more than the rest. One can divide workers into those who serve their local area (shops, restaurants etc) and those who serve the national population (car manufacturers, for example). Places with high leverage saw big job losses in local businesses; areas with low leverage did not. Both areas suffered equally from the effects of the fall in national demand. Furthermore, poorer homeowners tend to have a higher marginal propensity to consumer than the rich; hence their losses had a bigger impact on aggregate demand.

While this case is convincing, the academics seem to skate over an obvious point; the poorest homeowners are not the poorest people in society. Some people have negative net worth; they owe money on credit cards or to car companies or even to loan sharks. They may be too poor to qualify for a mortgage. Even at the height of the housing bubble, home ownership reached 68% of the population. Almost a third of Americans were renting (or living with their parents). For those people, a collapse in house prices won't affect their personal finances (save through the knock-on effects on employment) and they may even be helped as the interest rate on their debt falls.

However, having myself written a book on the crisis that focused on the rapid rise of debt, I am naturally sympathetic to this argument. The authors also have a healthy reverence for the work of Charles Kindleberger, who did much to explain the cycles of debt, asset bubbles and recessions. Kindleberger and Hyman Minsky showed how the cycle works; the greater availability of credit allows more people to buy assets, which rise in price, making lenders more confident and encouraging further credit growth. When the cycle turns, credit is restricted and prices collapse, causing forced sales and so on.

The general point is that credit is not a zero sum game in which gains to creditors equal losses to debtors; both can lose as prices fall. The creditor has to write off part of the loan and the debtor loses their deposit (equity). The tweak here is that Mian and Sufi argue that losses fall disproportionately on the debtor.

The academics go on to argue that the crisis was misdiagnosed by the authorities, with too much emphasis being placed on rescuing the banks and not enough on helping homeowners.

Policy-makers have consistently viewed assistance to indebted households as a zero-sum game: helping home owners means hurting banks, and hurting banks would be the worst thing for the economy

Only 2% of the TARP (trouble asset relief programme) was spent on helping homeowners; 75% was spent on the rescue of financial institutions.

The authorities were right to insist that depositors were protected from bank failures but wrong, the authors say, to protect other creditors such as bondholders. While again one has broad sympathy for this argument, I think it is wrong to suggest this was solely down to the power of Wall Street special interests. Creditors were not protected in all cases, and the result was that there was nervousness about holding bank debt in general; this nervousness put all banks' bonds and shares under pressure and added to the sense of crisis, particularly via a run on money market funds. The fundamental problem is the geared nature of bank balance sheets; protect one class of creditors and the run will occur somewhere else.

There was also a misdiagnosis on the lack of bank lending, the authors say. The proportion of small businesses that cited access to finance as a problem never rose above 5% during the crisis, while nearly 35% cited weak sales as a problem. Large businesses were also quicker to lay off workers than small businesses, even though the former still had access to bank loans. The problem was not that banks didn't want to lend; the problem was that businesses didn't want to borrow, given the weakness of the economy. Monetary policy creates bank reserves but not currency in circulation; central banks cannot force commercial banks to lend (or more pertinently, consumers and businesses to borrow).

The long-term solution, they say, is to change debt contracts so that losses are more evenly distributed.

If financial contracts more equally imposed losses on both borrowers and lenders, then the economy would avoid the levered losses trap in the first place. This would force wealthy lenders with deep pockets to bear more of the pain if a crash materlialises. But their spending wouild be less affected and the initial shock to the economy would be much smaller.

The idea of a shared mortgage is that the homeowner's debt will be linked to house prices in their immediate area; if prices fall, then both the interest payment and the principal amount would decline (this structure has similarities with that of inflation-linked bonds). Should house prices rebound, then the interest payments and principal would follow the same path. Of course, this structure is more risky for the lender, so banks would want to charge a higher interest rate. To prevent this, the homeowner would have to commit to pay 5% of any capital gain on the house to the lender.

In practice, I can see some complications with this approach; price losses are most likely to occur in the early years of a mortgage but some owners may hang on to their house for 30 or 40 years, postponing the bank's capital gain. The authors have a partial answer to this - the homeowner would have to pay up when the house is refinanced - but one could foresee that some banks might not be keen.

Clearly, there would need to be some sort of government encouragement for this market to get off the ground - in Britain, there is a help-to-buy programme under which the government underwrites 20% of the loan. This structure has come under criticism and the authors are not great fans. But the take-up of the programme shows that equity-sharing could be popular. (It is striking, also, that the idea is central to Islamic finance while debt forgiveness or a jubilee is an old Jewish tradition.)

More broadly, the authors suggest that government debt could be linked to GDP so the burden could fall with recessions (Paul Woolley of the LSE had this idea some years ago.)

The ideas in the book are radical and are not going to be adopted overnight. But as I have said elsewhere, the choices facing us in the current position are stagnation, inflation or default. We don't want stagnation and inflation remains absent. So default may be the best answer. And Messrs Mian and Sufi's admirably clear book gives us an imaginative way of addressing the problem.

Discover more

So long, farewell

Three worries and three signs of hope in the final blog post

The flaws of finance

The sector is essential to the economy. But it is rewarded too highly and imposes wider social costs. The penultimate in a series of farewell blogs


Hope I save before I get old

Although we will probably spend 20 years or more in retirement, we don't think about it enough. The third in a series of farewell blogs