Sep

19

We find that variables implied by the structural credit modeling framework have significant forecasting power for corporate default rates. In contrast, lagged corporate credit spreads have no predictive power for default rates. These results complement and extend (the findings of) others who also find that credit spreads are significantly influenced by factors that are difficult to link to credit fundamentals. Finally, we find that the state of the economy as measured by a recession indicator variable has little incremental predictive ability for corporate default rates.

Rocky Humbert writes:

On a quiet Friday afternoon (waiting for options expiry), I began to read this paper …as the abstract caught my attention. Putting aside the fact that they did a lot of data massaging, the whole thing came crashing down on page 22. When they wrote: "The variance of stock returns is positive and significant….the other structural variables are not significant [to the Structural Model.]

Ok…so let's cut to the chase. They found that the only significant good predictor of bond default is the stock price. That is, when the stock goes up, the company is less likely to default. And when the stock goes down, the company is more likely to default.

As homer simpson would say, "Duh!"

They found no relationship w/ economic cycles which is interesting and counter intuitive. But the nature of economic cycles has changed over the last 150 years…but even so, if the economy goes from Boom to Bust, this result is surprising.

Yet, the MOST interesting thing here is that they found: "ONLY THE STOCK PRICE MATTERS" for predicting defaults. And "THE ECONOMIC CYCLE DOESN"T MATTER" for predicting defaults.

Hence, they are basically telling us that the economic cycle must not matter to the stock market … which is the most interesting thing of all!

p.s. i stopped reading at page 22 when i realized what this all meant. (which is left as an exercise to the investor/reader).

Stefan Jovanovich retorts:

As "everybody knows" the country is suffering a slump brought on by a failure of aggregate demand; and that can be cured by an exercise of the sovereign's power to stretch the currency aka money supply. The authors, on the contrary, are heretical enough to think that balance sheets matter. They also think that risk assessments based on business cycle data are effectively useless as a method of determining the probability of whether or not someone is going to pay off the debt they owe, with interest. "The worst three-year period is the 1873-75 period in which the default rates total to 35.90 percent. The 1892-1894 and 1883-1885 periods resulted in total default rates of 18.69 and 16.06 percent, respectively. The 1933-1935 period during the Great Depression ranks a distant fourth with a total default rate of 12.88 percent. Thus, while the Great Depression may have been the worst economic period during the sample period, it is actually very far from being the worst credit event experienced in the corporate bond market. The fifth and sixth worst three-year periods have total default rates that are very similar to that of the 1933-1935 period. Observe that the 2000-2002 period is the only post-World-War-II period in the top twelve three-year periods listed in Table 3. We note that the year-to-date corporate nonfinancial default rate for 2009 as of the end of September is 2.46 percent (per Moody’s)."

The authors' point is that bond defaults are a function of leverage and that there is a correlation between leverage and diminished returns to both bondholders and shareholders. My point was that the elastic currency introduced by coordinated central banking has shifted the risk of leverage from businesses, which can no longer afford it, to sovereigns.

If one is looking for an explanation of business cycles, Eric Falkenstein's is still the best.

Rocky Humbert responds:

Stefan: You're operating at a higher level than me on this. I'm not quarreling with your comment below. However, because the study covers such a huge period of time, they had very limited data to study. For example, they used "high quality New England municipal bond prices" versus commercial paper prices for calculating the effect of the yield curve shape. Those two metrics broke down in 2008, and similar chaotic distortions may have occurred during other credit crisis periods making their analysis and conclusions suspect.

The other thing that the paper observed (which I think is interesting) is they found that owners of corporate (non financial) bonds who held through the entire study period had an excess return … even after defaults. They note that this is consistent with other studies which explain the excess return due to a liquidity risk premium.

If the authors included money supply in their analysis, I missed that. (Which is entirely possible.) But they used muni bond prices in the 19th century precisely because they recognized the gold-peg effect … which made govt bond prices effectively risk free (in gold terms that is.)

Stefan Jovanovich replies:

My comments are on a much lower level, Rocky. I am continuing my neo-political rant about how unification of the credit and monetary systems (along with the Federalization of legal authority) since the Progressive era has become its own form of leverage. Just as the authors of this paper had to search for evidence of default among the equity court records (no Federal bankruptcy), all our present calculations of sovereign leverage are incomplete because they do not include an assessment of the net obligation of all the guarantees and promises made. As Chris Whalen points out, with ZIRP those IOUs that are not traded cannot even be discounted. To be specific, California has no idea what it now owes its prison guards because it has no idea how much should be saved now for a payout to be made in 2040. That is fine with our legislature because they are the gang that never counts straight; but it means that the pools of investment capital dedicated to pensions are now sailing without either a map or a compass in terms of their future obligations. To me that forecasts a default rate in pensions that will mimic the post Civil-War collapse in railroad bonds that the authors studied.

Sinking further down. The reason Grant remains the least appreciated President in history even as he was the most popular one while still alive is that he had the wisdom to understand the common sense view of the ordinary citizen - that public money risk had to be completely separated from all credit risk (both private and public) in the financial system. Grant favored uniform regulation and clearing - the national banking system that the Federal Reserve effectively took over is one of his legacies - but he insisted that the government, because it had a Constitutional monopoly, must resume making legal tender exchangeable into specie - a commodity whose supply and demand could not be manipulated by the stroke of a pen or tap on a keyboard. The markets - left to themselves - could deal with fraud and deceit and failure and even banking default through effective discounting as long as the participants all knew that they had to resolve the matters themselves, without free money from the Treasury or manipulation of the interest rate.

As Rocky notes, the authors said absolutely nothing about the money supply, and they did not even mention leverage directly. They left those inferences to us drop-outs. Enjoy the weekend, all.
 


Comments

Name

Email

Website

Speak your mind

Archives

Resources & Links

Search