Magnifying the Credit Fallout

From The Wall Street Journal March 6, 2008 page A2

Even as financial firefighters try to douse the flames, the search is under way for the cause of the fire.

How could a mortgage-market meltdown -- losses of perhaps $400 billion, less than 2% of the overall value of the stock market -- cause so much of a disturbance and do so much damage to the U.S. economy? "After all," Federal Reserve governor Frederic Mishkin observed last week, "a 2% decline in stock-market prices sometimes happens on a daily basis, and yet leads to hardly a ripple in the U.S. economy."

And is the fire being fanned by the way commercial banks are required to keep their books and decide how much capital to hold as a cushion against bad times?

The short answer to the first question is leverage. The short answer to the second is yes. Leverage is borrowing money to make bigger bets. Invest $1, borrow $9, buy something for $10. If its value rises $2, you've tripled your initial investment. It works great when the market is on the way up. On the way down, it amplifies losses.

Banks are highly leveraged. That's how they make money. Peter Fisher of money manager BlackRock recalls giving a talk to a bunch of bankers and asking rhetorically: "What's the difference between a hedge fund and a bank?" Before he could answer, someone in the audience said, "Banks are more highly leveraged."

A bank has a set amount of capital. Based on regulatory rules and management's judgment, it borrows some multiple of that sum, either in the markets or by taking money from depositors. It puts that money into loans or securities, its assets.

Banks today find their assets worth less than anticipated, the consequence of a real-estate bust and falling market prices of securities. These losses erode a bank's capital cushion. With less capital, banks shrink their balance sheets; they lend less.

How much less? That's where leverage comes in.
Say for every $1 less in capital, a bank lends roughly $10 less. At a conference last week, sponsored by the Brandeis University and University of Chicago business schools, two Wall Street economists and two academics estimated that about half the mortgage losses, or about $200 billion, will be borne by banks and other leveraged financial institutions. That will lead them to shrink their balance sheets by about $2 trillion by lending less and selling assets.

With money cheap, banks gorged themselves with leverage in good times, making not only risky mortgages, but increasing leverage by investing in securities that rested on the riskiest slice of those mortgages. Bank balance sheets now are on a forced diet.

The impact on the rest of us depends on how much new capital banks raise and how much they reduce leverage. The four economists estimate all this will translate into $900 billion less in loans to households and businesses, and say that will reduce economic growth over the next year between one and 1.5 percentage points. Ouch.

That's the "what" and part of the "why." But the current approach to bank capital and accounting is exacerbating the ups and downs. When times are good, loans and securities look less risky, and banks increase leverage to make more money without building capital. When times are bad, they do the opposite.

"When asset prices rise, so does the value of collateral, which makes financing easier, increasing the demand for assets," Jaime Caruana, then Spain's central banker, observed in 2002. "That, in turn, pushes asset prices upward. In the downturn, as the value of collateral drops, financing possibilities decline, as does thus credit growth, a process often reinforced by financial institutions pursuing much more cautious credit policies as they are incurring losses or making smaller profits. ... Tighter credit policies reinforce recessionary forces and provoke additional reductions in asset prices."

For that reason, Spain, unusual among its peers, tweaked its rules in 2000 to require banks to set aside more capital when times are good so they have bigger cushions when times are bad.

The relatively new practice of requiring banks to value their loans and securities to market prices -- instead of assuming, unless there's good reason, the borrower will pay back the loan -- pushes in the same direction. It was a well-intentioned response to the mistakes of the past when Japanese banks and U.S. savings-and-loans were allowed to lie about the true value of their loans and collateral for years.

But, as economist Hyun Song Shin of Princeton University noted at last week's conference, when banks valued assets at what they paid for them, they had reason to sell when market prices rose and buy when market prices fell.

That welcome stabilizing effect has been lost. Banks today have incentives to buy more when prices are high, and are forced to sell when they are low.

As Mr. Shin and Tobias Adrian of the Federal Reserve Bank of New York wrote recently, "The expansion and contraction of balance sheets amplifies, rather than counteracts, the credit cycle." Capital standards and mark-to-market rules add to euphoria in good times and despair in bad.

That isn't smart.