Credit Crisis Roundtable: Markus Brunnermeier

The finance department at Princeton recently had a round table discussion on the financial crisis, causes and solutions. The panelists were Markus Brunnermeier, Hyun Shin, and Alan Blinder. All 3 panelists had some very insightful comments resulting in a very lively discussion. I wanted to share a few of the important points that I took away with some additional comments of my own. After I started writing this, I realized that there is plenty to say so I am going to split into 3 parts, one for each panelist.

Markus Brunnermeier

The first panelist to present was Markus Brunnermeier who discussed the mechanism at work behind the crisis. His basic thesis is that the losses associated with the real assets that went sour is a drop in the ocean. By his back-of-the-envelope calculation on slide 11, he estimates that a bad case would see losses on the scale of $500 billion. Although a large number, this is quite small in the grander scheme of things as the stock market can lose or gain that much in a single day (less than 2%). He concludes that there needs to be some "amplifying mechanism" by which these relatively small losses create problems not commensurate with their size. I definitely agree with his assessment. Indeed, the mechanisms are precisely leverage that created the inverted pyramid we see currently and the real effect of credit tightening. He outlines these on slide 22.

1. Margin spiral. The losses, unlike in the case where the stock market loses 2%, are concentrated on certain sensitive players. These players tend to be highly leveraged. Even a small drop in assets causes large capital impairment, resulting in margin calls, distressed sales, and further capital impairment and margin calls. This is a theme I have discussed earlier so I will not beat it to death again.
2. Credit tightening. Prof. Brunnermeier refers to these as "lending channel effects." I mentioned these in the context of Lehman, but did not elaborate at the time. The basic idea is that the various financial institutions that were funding their long term assets with short term liabilities (commercial paper) cannot any longer pursue the activities they had previously been pursuing. First, their funding is drying up, as we have seen in the ABCP market; second, their capital is at significant risk of impairment resulting in precautionary hoarding; third, there is heightened counter party risk to worry about. The net effect is that banks and financial institutions must cut back on their activities and thus decrease their balance sheets.
   
3. Run on financial institutions. Just as banks can experience runs, various other financial institutions, such as hedge funds or SIVs, can as well. See slide 32.
4. Gridlock risk. I am actually not sure why this was discussed. It seems to me like if the situation ever arose there would be much larger problems at hand. It's kind of like if there's blood on your shirt, laundry is probably not your biggest problem. However, it is quite possible I am missing something so I refer the reader to slide 34.
   

The crux of these amplifying effects is the leveraged inverted pyramid structure that is built on top of housing as well as the real effects of money and credit. The latter is often hard to quantify as it involves unseen activities, but it is a very large and important cost. What I found most impressive about the discussion was that all the panelists, although they listed duration mismatch and risk shifting as central to the problem, were quick to point out that those are actually important fundamentally sound aspects of modern finance that should be taken in context. No point throwing the baby out with the bathwater. It was not discussed at much length, but I feel these are important points to be made:

1. Duration mismatch. This is an idea I have discussed earlier. Banks basically finance their acquisition of long term assets by issuing short term liabilities. Thus, they prefer an upward sloping yield curve. Much of modern finance is built around this theme. It is advantageous for all parties involved, including the average consumer. The crux of the modern economy is to eliminate the double coincidence of wants. I have discussed this problem in the context of money and interest rates on my other blog, but the idea extends to all aspects of economics. Well functioning capital markets serve a function that should not be underestimated. Unfortunately, the capital markets functioned in a skewed manner the last few years, but this has nothing to do with the duration mismatch and everything to do with the skewed incentives of various actors.
   
2. Risk shifting. The role of financial markets is to increase informational efficiency and allocate scarce resources in accord with consumer preferences. One such avenue is risk shifting. Prof. Brunnermeier mentions it on slide 7 as a "good" reason for structured products. In the case of structured products, various pension funds or other institutional investors who couldn't otherwise invest in the mortgage market were now able to because of the repackaging of risk. In addition, various actors (hedge funds) can assume precisely the risk they are looking for. The idea is fundamentally sound. In fact, the idea is no different from stocks, bonds, convertibles, or warrants, on the face of it. The problems arise due to agency issues. That the derivatives market has been so maligned is a reflection of the incentives within the market that resulted in deteriorating credit standards, and not derivatives themselves.
   

Although duration mismatch and structured products have taken a lot of flack recently, it would be wrong to accuse them of being the root causes of the credit crisis. The root cause was the structure of the regulatory environment that enabled the "originate to distribute" model and the highly misleading "ratings". As a result of the skewed incentives in the model, the house of cards built on the illusion of real wealth that is excess money flowing through the system, has finally been realized as paper wealth and not real wealth after all. Various mal-investments (level 3 assets) must now be liquidated (repriced) to borrow an analogy from the Austrian theory of the business cycle.

My Additions

Having discussed the topics Prof. Brunnermeier covered, I would now like to add some additional points that were not brought up, but that I feel are important.

1. Fed. The Fed kept interest rates too low too long and fueled one of the largest bubbles we have ever seen. Anyone who has played around with compounding can appreciate the difference a 1% versus 2% interest rate makes on the affordability of a mortgage. That interest rates were at historical lows for a large time enabled many buyers to bid up the price of houses considerably without increasing their monthly payments. But it doesn't stop there. The cash spigot created incentives for speculators and flippers to come in and further bid up prices. Banks in their eagerness to capture the liquidity the Fed was providing made it easier and easier for buyers to speculate on housing with deteriorating credit standards and financial "innovations". Just look at so called liar loans and NINJA loans. Further, the securitization and distribution of mortgages enabled capital to be freed up very quickly for originators to make more and more loans. Effectively, there was so much money in the system chasing all the various buyers that credit standards deteriorated and house prices were bid up furiously. When excess money is flowing through the system, it creates many skewed incentives along its path. I was surprised that Prof Brunnermeier failed to mention the role of the Fed in fueling the housing bubble.
2. Credit ratings. The major credit ratings agencies were the other players at the center of the problem. Their business model was built around originators paying them to rate the structured products that were being built. In order to get more business, they had to give more lenient ratings. In addition, institutional investors were pushing for higher yielding "AAA" rated securities that were within their mandates. With such incentives, is it a wonder that ratings were far higher than fundamentals would suggest. Although the structuring of the debt was a good idea, the ratings that were slapped onto it were unrealistic. Had more realistic ratings been provided, many of the investors who bought the repackaged debt would have been a little more wary. Unfortunately, as a result of the skewed incentive, structured products as a whole have received a bad rap.
3. Statistical models. One could argue that, even if there were skewed incentive to provide excessively high ratings, the ratings agencies still need to provide some objective justification for their ratings. Indeed this is true, but they did have those "objective" justifications in the form of statistical models. Statistical models are very useful, but need to be carefully understood and correctly employed. I have played around with the pricing methods for CDO's, CDS's, MBS's, and can attest first hand that the models are very sensitive to inputs. It is not very difficult to change a few inputs and create the desired effect. I discussed this elsewhere in the context of FASB 157. I am planning to write more detailed articles on the exact models soon.
   

Net I really enjoyed Prof Brunnermeier's presentation and the discussion that followed. I am sorry to race through such a vast topic. I wanted to put these thoughts down before I forget.