The Information Matrix was the uninvited guest that everyone wished attended the 2018 Nobel Symposium on Money and Banking.

Thanks to John Cochrane, a Senior Fellow at the Hoover Institute who blogs under the Grumpy Economist, we know the high point of the conference was when the question was asked how is information-insensitive debt reconciled with the Economic profession’s traditional model of bank runs.

The honest answer:  not very well, but who cares as both ideas are fundamentally flawed.  But I am getting ahead of myself in the story.

Regular readers know the Information Matrix explains the flaws in both the concept of informationally insensitive debt and the traditional model of bank runs.

The Information Matrix provides the framework for understanding financial crises using behavioral economics.  Behavioral economics recognizes people like a good story.  In fact, people can like a narrative so much they behave in ways that are irrational.

Wall Street understands this.  Behind every investment product Wall Street sells is a narrative designed to make the product sound appealing.  However, every investment product narrative is not necessarily accompanied by the disclosure investors need in order to verify the story Wall Street told.

Information Matrix

	Does Seller Know What They Are Selling?
Does Buyer Know What They are Buying?		Yes	No
	Yes	Perfect Information	Antique Dealer Problem
	No	Lemon Problem	Blind Betting

Keep in mind, behavioral economic’s observation people like a good story operates in both the Perfect Information and Blind Betting quadrant.  The key difference between the two quadrants is in the Perfect Information quadrant the story can be verified and in the Blind Betting quadrant the absence of the necessary information means the story cannot be verified.

Whether the story can be verified or not results in a vastly different response when Wall Street’s valuation story is called into doubt.  In the Perfect Information quadrant, the story can be verified and the doubt dismissed.  In the Blind Betting quadrant, the story cannot be verified.  Not only is the doubt not dismissed, but the logical follow-up question arises:  is the investment worth anything?  This too cannot be verified.  Investors recognize this and, as behavioral economics suggests, naturally panic.  The result is a “run” to get their money back.

So now let’s turned to the fundamentally flawed bank run model.  The model assumes multiple potential causes of a bank run including sunspots.  It is the inclusion of sunspots that is telling.  It tells us there is a factor that unites all of the potential causes the model creators didn’t identify.  The Information Matrix shows this factor is opacity.

Why do we know this?

Because runs are not triggered in the Perfect Information quadrant by sunspots or anything else.  Investors have the information they need to verify Wall Street’s valuation story and dismiss any doubt raised about the valuation caused by sunspots.  Runs only occur in the Blind Betting quadrant.

One of the creators of the flawed bank run models, Douglas Diamond, spoke at the symposium.  He concluded

I agree.  And when you focus on runs in the financial system what you find is the necessary condition for them to occur is the opacity of the Blind Betting quadrant.  And when you recognize this is the necessary condition you understand why the global financial system is based on transparency and why it tries to force all transactions into the Perfect Information quadrant.  Transparency eliminates opacity which in turn eliminates bank runs.

However, Professor Diamond doesn’t understand the interaction of opacity and human behavior that causes runs.  For example, he observed

He focuses on the moment when risks/losses are revealed.  In fact, investors began their run from Lehman in the spring of 2008 as soon as its valuation story was called into doubt.  Like the Too Big to Fail banks, Lehman was opaque.  It didn’t provide the information investors needed to either dismiss doubts about its valuation or conclude the investment was worth anything.

[Please note, transparency does not prevent risk taking by banks.  Besides eliminating runs, what transparency does is subject the banks to market discipline.  When investors can see how much risk a bank is taking, they limit their exposure to the bank to what they can afford to lose given the risk of the bank.  As risk taking increases, investors reduce their exposure.]

Also speaking was Yale Professor Gary Gorton who is still pedaling his view of debt

as a generally information-insensitive, money-like asset.

Information-insensitive debt doesn’t exist.

This is worth saying again:  all debt is sensitive to information.

As shown by the Information Matrix, information-insensitive debt doesn’t exist.

Professor Gorton assumes investors are too lazy to look at the available information about the investment themselves or hire trusted third parties to do the assessment for them. [This would appear to be contradicted by retail investors use of professional portfolio managers and the growth of mutual funds.]

The only question is whether or not the information an investor needs to assess the risk of the debt is disclosed or not.

Mr. Cochrane summarized Professor Gorton discussion of what drives runs and in the process reveals Professor Gorton also doesn’t think information-insensitive debt exits.

Except after bad news, then debt gets closer to the boundary, and more importantly people who haven’t been paying any attention start to pay attention. Debt becomes illiquid, due to fear of asymmetric information, and everyone wants a limited supply of money. A crisis follows.

Wait a minute, how can debt be information-insensitive when it reacts to bad news?  Doesn’t that imply it was always information-sensitive?  [hint: YES!!!!!  As a practical matter, a fixed income investor, particularly in money market or short-term debt, doesn’t react to good news as their return is capped to getting back their investment plus interest.]

Mr. Cochrane is troubled by this question too.  As he states it:

I find this a compelling story for the financial crisis and crises in general. But it is quite different than the type of bank run Doug Diamond describes. Most importantly, there is a deep tension between Diamond and Rajan’s view and Gorton’s view. Diamond and Rajan say that debt is necessary, because it disciplines managers. Debt holders are constantly monitoring management, and running at the first sign of trouble. In direct contrast, Gorton’s debt holders are paying no attention at all most of the time, and then dump debt out of blind fear.

One weak spot of the conference was that everyone was being too polite. Well, everyone but me. Here we have a glaring difference in views. Which is right? I asked the question.

Rajan’s response was very informative: Yes, most retail debt customers are “information insensitive,” and likely even most corporate treasuries using repo as a cash substitute.  But among the New York banks who are funding each other very short term, yes indeed they are paying a lot of attention and will run when they see trouble. So the “discipline” story is narrow, for this class of lender and borrower. That seemed to me a nice reconciliation of dramatically opposing views that has troubled me for some time.

Rajan’s response is a nice reconciliation.  Unfortunately, it too is wrong.

Rajan also assumes debt investors, including corporate treasurers, simply trust and are too lazy to verify Wall Street’s valuation story before investing in debt securities.  There is no evidence this is true.  In fact, when it comes to handling the company’s cash, corporate treasurers are extremely risk averse.  Why?  There is a direct link between their losing their company’s money and their losing their jobs.

Rajan goes on to say there is a group who pays attention when investing and as long as they hold a debt security: New York banks.  Why does he limit this group to only New York banks?  After all, money market funds invest in unsecured bank debt securities so they too have the same incentives.

Here is another example of why the Information Matrix is needed.  It answers the question of how market discipline and bank runs fit together.

The “discipline” story isn’t narrow.  The “discipline” story only works when the debt is in the Perfect Information quadrant.  There is no reason to expect the “discipline” story to work when the debt is in the Blind Betting quadrant.

On the other hand, we know runs originate in the Blind Betting quadrant.

So if banks are opaque, then they are not subject to “discipline”, but they are prone to runs.