3
$\begingroup$

I'm referring to this interview with Michael Burry. He says:

Central banks and Basel III have more or less removed price discovery from the credit markets.

Why would Basel III cause this effect?

$\endgroup$
9
$\begingroup$

The argument about Basel is an adjunct to the central bank argument. There, a relatively price-insenstive CB either distorts credit by buying it directly (ECB, Fed wrt Agencies). Or distorts it indirectly by distorting the govvie market, pushing hitherto govvie holders into credit markets at a different price to what everyone would see without this effect. At least in the proverbial "good times".

The Basel argument proverbially burns the same candle from the other end, in terms of what ceases to happen in "worse times".

Tighter financial regulation makes bank balance sheet leverage a more expensive commodity (maybe no bad thing). It also increases the capital adequacy weighting of holding private sector credit risk (again, maybe not such a bad thing). But the effect of the two combined on the incentives for banks to warehouse this risk is chilling. Absent the same deep well of credit inventory to freely trade in size, absent (some degree of) price discovery.

Back in the bad old days, the banks would have been willing to step in and buy some juicy yield if prices gapped down. Or a very juicy spread if the bank's own funding was cheap and plentiful, as used to be the case. These days, the bank's risk officers will just as likely tell the traders not to bother trying, because they'll jack up the risk price as much as the pickup in yields/spread. So credit's traditional buyer of last resort isn't incentivised to step into weakness anymore.

That's what Burry (and myriad others, for a while now) are worried about.

$\endgroup$
4
$\begingroup$

Assume, for a moment, that all of the math behind modern portfolio theory is incorrect. See, for example, this video on deriving the distribution of returns. Without the normality assumption, indexing and methodologies like the Basel accords are not neutral.

The Gaussian distribution is the result of a convergent process over time. The system has to have a sink in it. Capital is a source, not a sink. It is the result of a divergent process.

I can think of a couple of distortions in the equity markets that were obvious. I am less active in the credit markets data, however, he is correct. There is, in fact, a far more serious pricing error than those in the credit markets, currently. A historical example may help here.

Back in 2008, when General Motors announced that its accounting records were a sham and were completely disconnected to reality, its price should have approached zero but it was a member of the Dow and other index funds. What happened, instead, is that as its price cratered it drove down the PE of the Dow as a whole, undervaluing the other components pushing money into Dow funds driving up the price of GM because 1/30th of the money went into GM stock. It also triggered options activities on the indices that had the effect of propping up GM stock.

By proper valuation, GM should have been worthless and had the United States government not rescued GM it wouldn't exist today.

Indexing creates a rigidity among the relationships between assets that are not rational. The NASDAQ, in particular, is counter-rational. Once you drop normality, the NASDAQ will tend to be composed of the most over-valued securities and the capital weight is strongly influenced by over-valuation. In a sink-based system, that is long-run harmless because a covariance matrix exists. The distributions involved with returns cannot have a covariance matrix so rigidities are intrinsically sub-optimal at the personal level. Burry's argument is that there is a macro effect. I strongly agree.

His argument around Basel is essentially the same. There are two perspectives on risk one could take. One is piecewise, the other is systemic. Basel homogenizes classes to make them tractable. Again systemic versus idiosyncratic risk depends on a covariance matrix. Without it, it isn't obvious that these can be parsed. While Basel does not create credit index funds, per se, it creates the same effect. Of course, credit index funds came into existence since then, in part, to ameliorate the cost of risk accounting.

If you would like to think of it another way, think of it in terms of Keynes's "sticky prices." Basel III, unintentionally, causes rigidities because there is an assumption that the strategy in use will not impact the incentives in the system. For indexing or homogenizing systems like Basel to work, they cannot impact the supply or demand curves for the specific items. That is nonsense.

Now, this is not to say that system-level risk does not exist. Trump's tariffs are doing real, permanent damage to American agriculture. There will never exist a route to the glide path it was on before Trump. Even if Trump resigned tomorrow and Pence announced some sudden "win," on the trade war, the system has adjusted worldwide in ways that build in the damage done. There are other industries that will end up there as well, most likely. Agriculture though can't be fixed anymore. Trump broke it. System-level risk is real and indexing does not ameliorate it without a covariance matrix.

This is a basic discussion of fixed versus flexible prices. It also, as in the video above, assumes there is a deep and fundamental error of mathematics in Modern Portfolio Theory.

I asked a mathematician once how I could explain to other economists the underlying error of mathematics in Modern Portfolio Theory. He had reviewed my work. His response was "That is easy. It violates the laws of general summation." I told him that I couldn't tell economists they do not know how to add correctly. Nonetheless, Burry makes sense only if there is a terrible flaw in economic modeling and econometrics.

There is.

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.