A new paper by three economists shows how far finance has come from the days when banks took in deposits and used the money to make loans.
Today that same bank might augment those deposits by borrowing on the repo market and by selling credit default swaps, the paper notes. Then it could use the money to buy mortgage-backed securities or Treasuries, make a repo loan to a hedge fund, rehypothecate the repo collateral on the repurchase market, buy a Spanish residential mortgage-backed security denominated in Euros and hedge its Euro exposure.
In “Risk Topography,” Markus K. Brunnermeier of Princeton University and Gary Gorton of Yale University – whom Federal Reserve Chairman Ben Bernanke has recognized as leading thinkers on the systemic danger of the financial crisis of 2007-2008 – are joined by Arvind Krishnamurthy at Northwestern University to claim that the data regulators collect from financial institutions is dangerously inadequate and out of date.
They liken the situation to conditions in the 1930s, when Presidents Hoover and then Roosevelt had to design policies to combat the Great Depression “on the basis of such sketchy data as stock prices indices, freight car loadings, and incomplete indices of industrial production. “
They propose a new approach that could help regulators and the Office of Financial Research detect systemic risk before a crisis occurs, they said.
From the paper:
The financial crisis is a strong reminder that measurement is at the root of science. The measurement systems that we currently have are outmoded, leaving regulators, academics, and risk managers in a dangerous position.
(Hat tip to economic journalist David Warsh at the Economic Principals blog April 10 for calling attention to “Risk Topography.”)
The economists would require financial institutions to answer a series of questions designed to get the firms to reveal whether they have too much debt and too many illiquid holdings that would be hard to convert to cash in a crisis.
For example, one question might be: What is the capital gain or capital loss to your firm, and how would your liquidity position change, if house prices fall by 5 percent, 10 percent, 15 percent, and 20 percent, and what if they rise by the same increments? The answers to these questions would be influenced by such things as whether the firm owns mortgage-backed securities directly or through derivatives, whether it is funded using short-term debt like repos, long-term debt, or equity, and what are the collateral arrangements in the repo and derivative transactions undertaken by the firm.
The questions would be posed monthly or quarterly, answers would be aggregated to protect proprietary information, and they would be made public. Regulators could cross-check answers to test for “truth-telling,” the authors write. Larger firms should have to supply more data. Over time the data could help regulators, analysts and others spot the build-up of risk, the authors believe.
This approach would be superior to judging a financial institution’s systemic importance by its size, according to the three economists, because size is based on the assets that a bank lists on its books, and that excludes large off-the-books exposures at today’s financial institutions.
In their paper, the authors give five examples of what financial institutions today might do with $20 of equity and $80 of debt. “The examples are simplified in the extreme and so they are clearly not realistic, nor are they intended to be,” write the authors.
Benchmark: Consider a firm with $20 of equity and $80 of 5-year debt with a coupon rate of 4½ percent. The firm makes loans to two different firms, each for $50 for one year at an interest rate of 5 percent.
Liquidity Mismatch: Consider a firm with $20 of equity and $80 of debt as above, but now half the debt is overnight repo financing at 1 percent and the other half is 5-year debt at 4½ percent. The firm buys one agency mortgage-backed security for $50 (which is financed via repo at a zero haircut) and loans $50 to a firm for one year at an interest rate of 5 percent.
In this scenario, the firm has introduced new risk. For example, what if it can’t renew the repo financing and has to sell some assets to repay the repo loan?
Suppose the mortgage-backed security is private-label (not guaranteed by agencies like Fannie and Freddie) instead of agency backed. The private label will be harder to sell.
Rehypothecation: The bank lends $100 to a hedge fund for three days and receives a bond with a market value of $100 as collateral. The bank then uses the bond as collateral to borrow $100 in the overnight repo market.
In this scenario, the bank is lending long (three days) by making a repo loan to the hedge fund, and it is borrowing short (overnight), also on the repo market. What if the bank can’t renew the overnight repo financing and has to repay that repo loan before the hedge fund repays its three-day repo?
Synthetic Leverage: Consider a firm with $20 of equity and $80 of debt; half the debt is overnight repo financing at 1 percent and the other half is 5-year debt at 4½ percent. The firm buys $100 of U.S. Treasury securities and writes protection (using credit default swaps) on a diversified portfolio of 100 investment-grade U.S. corporates, each with a notional amount of $10; so there is a total notional of $1,000. The weighted-average premium received on the credit default swap is 5 percent.
In this scenario, the firm has the same repo risk as the earlier examples, in that the repo lender could refuse to renew the loan and the firm would have to find the cash to repay.
In addition, what if several of the U.S. companies that issued the investment-grade corporate securities failed and lost half their value, perhaps because of a recession? The firm would have to pay off on the credit default swaps.
Or what if the credit default swap contracts call for the firm to post more collateral when the risk to its CDS-buyers increases, such as when the value of the original collateral declines or when the rating of the CDS-selling firm itself declines?
Cross Scenarios: Consider a firm with $20 of equity and $80 of debt; half the debt is overnight repo financing at 1 percent and the other half is 5-year debt at 4½ percent. The firm buys a Spanish residential mortgage-backed security denominated in Euros, for the equivalent of $50, and lends the other $50 to a U.S. firm. The firm does not hedge its Euro exposure.
In this scenario, if Spanish house prices fall, the firm takes a loss on the Spanish mortgage-backed securities.
If the Euro becomes worth less, the firm takes another loss because it gets Euros on the Spanish securities but has to repay its repo financing and 5-year debt in dollars.
If the firm tries to hedge its Euro exposure by agreeing to swap Euros for dollars, what happens if the company on the other side of that swap falters and can’t pay up?
Or what if the Euro becomes worth more, and the terms of the swap require the firm to put up more collateral?
All of these examples happened in the financial crisis. Most of these risks are not captured by today’s required reporting, according to “Risk Topography.”