Arnold Kling  

Michael Pettis on Financial Fragility

PRINT
The Ethics and Etiquette of St... Business Dynamics and Macroeco...

A while back, I mentioned that I was going to get his book. It arrived yesterday, and I have some excerpts below the fold.

Meanwhile, James Hamilton asks whether risky assets were undervalued in February of 2009.

When the price of an asset falls dramatically, it is natural to ask whether this can be explained by some shock to fundamentals. If not, was the asset overvalued prior to its fall, did it become undervalued after its fall, or both?

I can think of many instances in which the fall in asset price appeared to exceed what should have taken place given any change in underlying fundamentals. The stock market crash of October 19, 1987 is my favorite example. The "flash crash" of May 6 this year may be another example, but it is almost too bizarre to classify.

My view is that for the most part, although not always, sudden declines in asset prices indicate that the asset was previously overpriced. Also, I think that a majority of the time, sudden declines in asset prices take assets below their fundamental value. However, if crashes always took asset prices below fundamental value, you could make a sure profit by buying after crashes, and that is not the case. (You can always make a profit by buying at the bottom, but a crash is not always the bottom--after a crash, sometimes asset prices continue to drift downward. This strikes me as a plausible scenario for U.S. real estate right now, for example.)

Returning to James Hamilton's question, my own bet would be that risk premiums got too low and the price of risky assets got too high during the financial boom. It is possible that at some point in September of 2008 and later the market was over-reacting in the opposite direction, charging exorbitant risk premiums and undervaluing risky assets, but that is less clear. I think that relative to the Fed, Robert Hall, and many others of the Insider persuasion, I feel more strongly that asset prices were underovervalued before the crisis than that they were undervalued afterward.

One slight indication that risky assets were undervalued in February of 2009 is that I behaved as if they were undervalued in choosing my investment portfolio. I invested in a "junk bond" mutual fund, and I weighted my portfolio more highly in stocks than I do now.

The main point of The Volatility Machine is that some financing strategies for less-developed countries are attractive in the short term but fragile. For example, borrowing in dollars can reduce the interest rate, but if adversity strikes it is more difficult to repay loans denominated in dollars than to repay loans denominated in the home currency.

One key argument is that large capital inflows to LDC's often take place not because of changes within the LDC but because large money centers sometimes experience a surge in liquid assets. (Think of the securitization phenomenon in mortgages as creating a large increase in money center liquidity.)

On p.30, Pettis says that security prices do not reflect fundamental values during a crisis.


There were no investors who were willing to buy an asset that had become obviously become "cheap," because no investor was confident enough about his ability to interpret information "inside" his time horizon.

On p. 47, he says that instead of economic liberalization leading to capital inflows, it can be the other way around.

Capital inflows can permit government officials to implement reform by providing resources and profitable business opportunities to the entrenched elite who might otherwise oppose the reforms.

On p. 85, he argues that there are two types of LDC crises. The first is a major liquidity contraction in money center countries. The early 1930's offer one example. The current crisis (which comes several years after his book) is probably another. The second type of crisis is more like a panic centered on an individual LDC or group of LDC's. Such panics are less damaging and less long-lasting.



COMMENTS (3 to date)
Alex J. writes:

I gathered at the time that nobody was buying mortgage backed securities, because the sellers were holding out waiting for the government to pay too much for them. If true, both the sellers and potential buyers could correctly identify the situation and not meet at a price. Or possibly, the information that they couldn't correctly interpret is what bailout strategy the government would choose: government chaos induced stasis.

Vladimir S. writes:
I feel more strongly that asset prices were undervalued before the crisis than that they were undervalued afterward.

Overvalued?

fundamentalist writes:
risk premiums got too low and the price of risky assets got too high during the financial boom.

Why would people make such mistakes? The role of the market is to provide accurate prices so that people can make wise investments. Why is the market sending the wrong price signals? It's the Feds! Not only is it Austrian theory, but there is considerable empirical support for the fact that when the Feds reduce interest rates below the market rate, people increase their risk tolerance and invest in more risky assets. Banks loan to riskier businesses. And the rising money supply causes all prices to rise at the same time and makes bad businesses look good. When the crash happens, the credit contracts, which in turn causes prices to fall.

We shouldn't take the quantity theory of money too literally, but it's basic MV=PQ. Holding V and Q constant, if M rises, so will P. That's what happens during periods of low interest rates. But when the boom hits the brick wall of the shortage of capital goods (causing the Ricardo Effect to kick in), credit (and money) contracts. M falls, so must P fall.

If you don't have a basic understanding of the quantity theory, these price movements will seem mysterious and irrational.

Comments for this entry have been closed
Return to top