Banks, Securitization & Incentives: Are We Protected From Another Collapse?
The bank crisis was serious – $52 trillion in capital losses leading to the largest global recession since 1929. We should work hard to insure it does not happen again. But there is a problem: there is little agreement on why it happened. The litany of reasons includes: banks got too big, government pushing too hard on homes for everyone, inadequate regulations, and rating agency failures.
My view: everything started downhill when the primary incentive structure of banks changed. Until recently, banks lived on the difference between what they had to pay for deposits and their income from loans they held to maturity. That spread was their life blood. And because of this, they were extremely careful about whom they lent money to. That meant that throughout the life of their loans they stayed in close touch with their borrowers to insure that interest and principal repayments occurred on schedule.
In the last decade, this has all changed. Instead of holding loans to maturity, most banks now sell off at least a portion of their loans for commissions. This caused a fundamental change in bank incentives. Instead of being very careful about their borrowers, banks stopped caring: the more they lent the more commissions they could earn from selling them off.
The following graph traces the growing sale of collateralized loan obligations (CLOs) by banks. As indicated, there was little securitization of CLOs until 2004 when it increased dramatically until peaking in 2007 at $180 billion. Other forms of securitization followed the same time profile and brought the total into the trillions by 2007.
Source: Bord and Santos
Evidence of Incentive Effects
Until recently, the point about the change in bank incentives, although quite plausible, has been nothing more than conjecture. But two new empirical studies support the proposition that the ability to sell off loans does reduce the quality of bank loans. One looked at securitized sub-prime mortgage loan contracts and the other examined CLOs.
a. Securitized Sub-Prime Mortgage Loans (SSMLs)
Keys, Mukherjee, Seru and Vig just completed a study looking at the variation in the riskiness of SSMLs depending on the ease of securitization. They concluded: “the portfolio with greater ease of securitization defaults by around 10%–25% more than a similar risk profile group with a lesser ease of securitization…. Our findings suggest that existing securitization practices did adversely affect the screening incentives of subprime lenders.”
b. Collateralized Loan Obligations (CLOs)
Regarding CLOs, Bord and Santos found that “the loans banks sell to CLOs perform worse than comparable loans originated by the same bank. Loans sold to CLOs at the time of their origination underperform similar unsecuritized loans originated by the same bank….The difference in performance between CLO credits and non-CLO credits appear to have resulted from banks’ use of laxer standards to underwrite the loans they sell to CLOs.” The authors also found that selling off loans resulted in weaker bank ex post monitoring activities.
These conclusions are not surprising: when banks can make more money by selling off loans for commissions than holding them, they will make more loans. As a consequence, the riskiness of both the loans and banks will increase. The next question is whether changes since the bank collapse have rectified the problems that resulted in the banking collapse.
Are Banks Safer Today?
In response to the US bank collapse, the Dodd Frank Bill was passed. It sounds pretty good. I quote from a summary of its provisions:
“The Dodd-Frank Act implements changes that, among other things, affect the oversight and supervision of financial institutions, provide for a new resolution procedure for large financial companies, create a new agency responsible for implementing and enforcing compliance with consumer financial laws, introduce more stringent regulatory capital requirements, effect significant changes in the regulation of over the counter derivatives, reform the regulation of credit rating agencies, implement changes to corporate governance and executive compensation practices, incorporate the Volcker Rule, require registration of advisers to certain private funds, and effect significant changes in the securitization market.”
There are only two primary problems with the law:
- It did not reverse bank incentive structures: the Volcker Rule was not implemented: banks can still create, buy and sell loan packages. Congressman Barney Frank, the ranking member on the House Financial Services Committee, recognized the problem and tried to require banks to hold on to least some share of the loans they sold off. No luck: the provision was removed before the Act was approved.
- The lobbyists and lawyers of banks will have a primary say in how the Act is implemented. And they are significant players – Open Secrets reports that banks spent $61 million lobbying Congress in 2012.
Many believe that certain reforms will make the banks safer. We hear a lot about large banks, that some are so large we can’t allow them to fail. Keep in mind that more than 700 banks needed a TARP bailout, and many were not large. Many believe improved regulation is the answer. But is it? In recent years, as the riskiness and complexity of banks loans have increased, regulators have tried to keep up. They have tried to adjust the safe assets banks must keep on hand for different degrees of riskiness in a bank’s loan portfolio. This won’t work. The regulators cannot effectively regulate. I offer three pieces of evidence to support this assertion:
- Leading up to the Eurozone Crisis, regulators allowed banks to count sovereign debt as equivalent to cash as part of their asset base!
- Dodd Frank became law on July 21, 2010. Two years later – the summer of 2012, we hear that JP Morgan Chase (JPM) trading losses could reach $9 billion!
- Take a look at JPM’s annual report; in particular, skim over the 123 pages of “notes to the consolidated financial statements” (pp. 182 – 304) in the report. Can regulators understand what JPM is doing? No.
The Only Solution
We do not deposit money in banks so banks can gamble. We put money in banks for safe keeping. Depository institutions should not be allowed to trade any assets. Too risky. A simple solution – limit government bank insurance (in the US, FDIC insurance) to banks that manage their own loans to maturity and do not engage in trading on their own accounts.
Post Script on Rating Agencies and Incentives
Incentives are important to how any institution operates. And rating agencies are now paid by the institutions whose securities they are rating. Their ratings cannot be trusted.