Monday, January 28, 2013

Safe Assets and Financial Crises

Mark Thoma has shared a link to a new working paper by Gary Gorton and Guillermo Ordoñez called "The Supply and Demand for Safe Assets." The paper brings to mind a once-confidential document written by economists in the Clinton Administration called "Life After Debt" which was recently made public by the team at NPR's Planet Money. The report notes:
In the year 2000, the U.S. Treasury began actively buying back the public debt; we should all appreciate the tremendous achievement this represents for the Nation as a whole... We must realize however, that a sharp reduction in Federal debt and the possible accumulation of a Federal asset raises at least three important issues. First, investors looking for an asset free of credit risk can no longer count on an abundant supply of U.S. Treasury securities, and Treasury securities may no longer provide a reliable benchmark for other interest rates. Second, the Federal Reserve may have to change the mechanisms by which it conducts monetary policy. Third, continued surpluses after the public debt has been paid off will require the Federal. government to acquire assets; either directly or though the Social Security Trust Fund. This raises issues about what kinds of assets might be acquired, and the best way to manage this task.”
Gorton and Ordoñez's paper is relevant to the first of these issues. The Clinton Administration report elaborates on this issue, saying:

US Treasuries are considered free of default risk by investors the world over...The remarkable liquidity of Treasuries is also a result of the full faith and credit of the United States Government.  Holding a liquid asset is valuable because it affords an assurance of convertibility, and thus fast and easy access to capital.  Private investors, the Federal Reserve and many foreign central banks have used Treasuries to fulfill their need for a riskless, performing asset with liquidity second only to currency.
Gorton and Ordoñez note that the share of safe assets in the U.S. economy has remained constant since 1952. However, the composition of these safe assets varies. Safe assets consist of both U.S. Treasuries and privately-produced substitutes, so when the supply of Treasuries declines, the share of private subsitutes rises. What can be a private substitute for Treasuries? Typically, asset-backed securities. Collateral is key.

In Gorton and Ordoñez's model, for simplicity there is just one type of collateral: land. Land can be either "high quality" or "low quality." While the average land quality is known, there is no public information about which land is high quality and which is not. Borrowers can use their land as collateral to finance investment projects, and lenders don't know the land quality unless they pay some cost to find out. This is a type of financial friction: it is inefficient for the economy as a whole if lenders pay a cost to learn about collateral quality, because that cost does not result in any production.

In the model, there are normal times and crisis times. In normal times, the average land quality is high enough that lenders are better off NOT paying to check the quality of the land. The inefficiency from the financial friction is avoided. However, there can be shocks to the average quality of land. Land quality may get low enough that  lenders need to check the land quality before they accept it as collateral, resulting in economic ineffiiency and a financial crisis. This is where Treasuries come in. Government bonds can also be used as collateral, and they don't suffer losses in value like land does. In short:

Since bonds can be effectively used as collateral, a larger fraction of bonds buffers the economy from potential shocks to the expected value of land that may reduce its role as collateral, inducing a lower probability that such shock translates into a financial crisis. This is consistent with the empirical findings of Krishnamurthy and Vissing-Jorgensen (2012a); an increase in Treasury debt decreases the probability of a financial crisis. In our setting this is because bonds can be used as superior substitutes for private collateral – they are independent of shocks to land.
Of course, they are not just advocating for the government to run up a huge debt.The model also includes taxes, and they make the important additional note:
But, if taxes to repay bonds are distortionary, it may be optimal for the government to issue debt in times of crisis, but not in normal times. 
"Land," remember, is a modeling simplification, and really encompasses all types of private collateral. Before the recent financial crisis, there was a surge in the creation of "safe" private assets using "pools" of collateral including loans, bonds, and mortgages. Josh Koval and Erik Stafford explain:
The essence of structured finance activities is the pooling of economic assets like loans, bonds, and mortgages, and the subsequent issuance of a prioritized capital structure of claims, known as tranches, against these collateral pools. As a result of the prioritization scheme used in structuring claims, many of the manufactured tranches are far safer than the average asset in the underlying pool. This ability of structured finance to repackage risks and to create "safe" assets from otherwise risky collateral led to a dramatic expansion in the issuance of structured securities, most of which were viewed by investors to be virtually risk-free and certified as such by the rating agencies. At the core of the recent financial market crisis has been the discovery that these securities are actually far riskier than originally advertised.
For a time, the AAA-rated top tranches of these manufactured assets were considered really safe, and it was like the "normal times" in the model when lenders trust that on average, collateral quality is good enough that they don't need to pay the extra cost to check on it. But then it became apparent that the average quality was much lower, and these assets became less effective collateral, and the financial crisis began. There are at least some claims that the Clinton surplus kicked off the rise in mortgage-backed securities issuance. (I included two graphs below, made using data from FRED, in case you want to evaluate the claims for yourself.) If you decide to read "The Supply and Demand for Safe Assets," please do also look at Krishnamurthy and Vissing-Jorgensen's empirical counterpart. Or, for something lighter, listen to Planet Money's episode "What If We Paid Off The Debt? The Secret Government Report."


  1. In normal times, the average land quality is high enough that lenders are better off NOT paying to check the quality of the land. The inefficiency from the financial friction is avoided.

    A wornderful theory---where is the factual basis for the assertion?

    In fact, doesn't have to be false for any maturity of more than 5 or 7 days?

  2. They check it themselves or just ask an employee to check over said land...

  3. Do you think that the model (implicitly) distinguish non-safe asset, privately-produced safe asset and gov't bond?

  4. We have to be ready with this kind of cases. Financial and investment crises is something we can avoid, I can relate this kind of scene with real estate and investment. We have to be vigilant if things got out of hand.

  5. Excellent article here. Just wanted to add that if anyone is interested in learning more about similar topics, including enterprise data model tools, be sure to visit the Modern Analytics website for some excellent resources.

  6. Thanks for writing this amazing article, i would love it if you also write about Mutual Funds in Pakistan 2019.


Comments appreciated!