Wednesday, October 26, 2011

The problem is not bankers' incentives, the problem is opacity

For those of you who have not read Andy Haldane's Wincott Annual Memorial Lecture, I think you will find it very interesting as it is the exception to what is Andy's reliably excellent work.

Mr. Haldane provides a broad history of the development of banks from small firms where the owners put their entire net worth on the line in the mid-19th century to today's financial behemoths were owners have limited liability.

What is remarkable about his discussion is that he does not mention regulators until the last sentence of the speech despite the fact that banking has been regarded as a highly regulated industry since the Great Depression.

This omission is significant as the regulators were in a position to restrict the growth of these financial behemoths.  For example, they could have disapproved of their merger plans.

Instead, he discusses how along with this transformation in the structure of banks came a change in bankers' incentives and it was these incentives that led to our current banking sector problems.

Simply put the change in incentives was from when your entire net worth is on the line, you tend to error on the side of less risk to when it is only your bonus on the line, you tend to error on the side of taking too much risk.

According to Mr. Haldane, this change in bankers' incentives lead directly to changes in the riskiness of the assets that banks held and to the degree of leverage in the banking system.  By increasing both, bankers were able to earn more money.

To support this thesis, he observers that there was a steady increase in leverage or decrease in equity to assets from the mid-19th century to today.  Mr. Haldane attributes this increase to the incentives of bankers since increasing the leverage of the bank was an easy way to meet your earnings target and trigger your bonus.

I would be interested in what happened to bank leverage after the introduction of deposit insurance in 1934.

According to the paper, leverage was approximately 6x in 1900 and by the 1950-80 period had increased to slightly over 15x.  Was leverage still around 6x shortly before the introduction of deposit insurance?  After the introduction of deposit insurance, has leverage moved in lock step with an explicit or implicit cap set by the financial regulators?  If there was no cap on leverage, did bank leverage jump to 15x after the introduction of deposit insurance?

The reason I am interested is that the financial regulators had to know from day one that if they permitted banks to increase their leverage deposit insurance made it easy.

Prior to deposit insurance, the depositor had to be concerned with the liquidity and solvency of the bank.  This concern may have acted as a brake on growth in leverage.  Given the risk to the investor of losing their deposits, maybe the accepted rule of thumb was not to put deposits into a bank with greater than 6x leverage.

With deposit insurance, the depositor was guaranteed both liquidity and the equivalent of 100% equity backing for the deposits.  As a result, any increase in the bank's leverage did not make the bank riskier to depositors. Therefore, bank leverage became irrelevant to the depositors.

Deposit insurance had the impact of removing the depositors' "cap" on leverage.  This freed depositors up to make a choice of where to put their deposits on issues like convenience.

While I agree with Mr. Haldane that bankers' compensation has a bias towards leverage, it is not clear that deposit insurance and the financial regulators did not facilitate this.

Second, Mr. Haldane observes there was a steady increase in the use of debt in the bank capital structure.  This was another way for bankers to increase leverage and meet their compensation targets.

Mr. Haldane notes that using debt in principle is not a bad thing because with it should come the disciplining effect of debt.  In theory, debt holders should restrict increases in risk by the bank by either increasing the cost of debt or reducing its availability.

As he points out, debt was an effective discipline in the 19th century with the potential for a run on the bank by depositors rewarding bankers for maintaing liquidity and conservative investments.  This was still true through the Chicago banking panic of 1932.  As he notes, the effectiveness of debt discipline became patchy to non-existent after that.

The introduction of deposit insurance in 1934 would cause the effectiveness of debt discipline from depositors to completely vanish.  Why would depositors engage in a run on the bank when even if it becomes insolvent the depositors can still get their money back?

With the introduction of deposit insurance, the role of deposit debt discipline was taken over by the government since it and not the depositors was now in a position to lose money if the bank was badly run.

As a result, there was only one source of discipline on banks...their financial regulator.

Third, Mr. Haldane focused on an increase in risk in the asset mix held by banks.

With bank runs no longer a fear because of deposit insurance, a change in the industry's asset mix should have been expected to the extent that the government allowed or encouraged it to happen.

For example, to get the economy going again in the 1930s, the government might have pushed for a decrease in liquid assets, think government bonds, and an increase in illiquid assets, think loans.  We see something like this happening today in the UK with Project Merlin requiring banks to originate a specific volume of loans.

It is hard to disentangle how much of the shift in asset mix was driven by the bankers' incentives and how much by government policy.

Finally, Mr. Haldane looks at uninsured debt and sees a lack of effective debt discipline.  In particular, he looks at the cost of credit default swaps prior to and after the beginning of the financial crisis and sees significant mis-pricing of risk.

The mis-pricing of risk is also consistent with uninsured debt holders, who did not have access like the government does to a bank's current asset and liability details, relying on the representation by the government that the banks had a low risk of insolvency (this seems reasonable particularly when the Fed was cheerleading about how structured finance had reduced risk in the financial system).

Mis-pricing of risk is not a function of bankers' incentives, but rather of opacity.

Market participants cannot accurately assess the risk of a bank if they do not have access to its current asset and liability details.  If they cannot accurately assess the risk, market participants are not able to properly exert debt discipline.

When the role of deposit insurance and the financial regulators is added to Mr. Haldane's work, it changes the conclusions dramatically.

Given bankers' incentives, all market participants should have access to each bank's current asset and liability detail and not just the financial regulators.  This eliminates reliance on the financial regulators and brings market discipline to the banks.

No comments: