On disconfirming rational expectations

9 March, 2014 at 14:18 | Posted in Economics | 2 Comments

In his latest blogpost Noah Smith writes that

Rational expectations have not been empirically disconfirmed.

beat2Noah Smith is, of course, entitled to have whatever view he likes (it’s, to say the least, difficult to empirically disconfirm the non-existence of Gods …) — but for the rest of us, let’s see how rational expectations really fares as an empirical assumption. Empirical efforts at testing the correctnes of the hypothesis has resulted in a series of empirical studies that have more or less concluded that it is not consistent with the facts. In one of the more well-known and highly respected evaluation reviews made, Michael Lovell (1986) concluded:

it seems to me that the weight of empirical evidence is sufficiently strong to compel us to suspend belief in the hypothesis of rational expectations, pending the accumulation of additional empirical evidence.

And this is how Nikolay Gertchev summarizes studies on the empirical correctness of the hypothesis:

More recently, it even has been argued that the very conclusions of dynamic models assuming rational expectations are contrary to reality: “the dynamic implications of many of the specifications that assume rational expectations and optimizing behavior are often seriously at odds with the data” (Estrella and Fuhrer 2002, p. 1013). It is hence clear that if taken as an empirical behavioral assumption, the RE hypothesis is plainly false; if considered only as a theoretical tool, it is unfounded and selfcontradictory.

For more on the issue, permit me to self-indulgently recommend reading my article Rational expectations — a fallacious foundation for macroeconomics ina non-ergodic world in real-world economics review no. 62.

About these ads

2 Comments »

RSS feed for comments on this post. TrackBack URI

  1. Savvy traders know that it is rational to expect alternating periods of irrational behavior in markets driven by fear (bears) and greed (bulls), even very short term, and they seek to profit from excess and deficiency in adjusting price to value owing to cognitive bias. Behavioral finance pays a lot of attention to cognitive bias aka irrationality, or “animal spirits” (Keynes).

    Anyone who has operated in marketing, advertising, and sales knows that irrationality is fundamental in pricing as well as driving sales, e.g., impulse buying. Marketers even figure the expected return rate beforehand to determine the percentage of impulse sales that stick. Even if the return rate is fairly high, it’s usually profitable to push impulse buying.

    And, of course, advertising is mostly about manufacturing demand using cognitive bias. This was introduced by Edward Bernays by combining crowd psychology with the psychoanalytic theory of his uncle, Sigmund Freud. (His mother was Freud’s sister, and his father was the brother of Freud’s wife). After this persuasion became a profession and evolved into public relations, advertising and marketing, and propaganda.

    Playing on irrationality is very much like creating artificial scarcity using monopolistic and oligarchical practices. Position and branding, for example, is targeted at this, as Veblen noticed as conspicuous consumption of what economists now call Veblen goods, which are more in demand with rising prices making them more attractive as luxury goods.

  2. Continuing the above and applying it to the crisis, I believe that the crisis can be explained on the analogy of holiday impulse buying that has been engineered. It’s well known that a high percentage of retail sales are associated with Christmas buying, which used to begin in earnest on Black Friday, the day after American Thanksgiving. This year to spur sales it was extended to Thanksgiving and to Black Monday.

    The commercialization of Christmas was engineered in the US by commercial interests that turned what began as a religious holiday into a spending spree, largely driven by impulse buying using all sorts of sales and advertising gimmicks. Soon other holidays are either pressed into service, like Valentine’s Day by the jewelry industry, and the creation of Mother’s Day and Father’s Day by commercial interest. These are all very rational exercises in using irrationality for profit. However, it doesn’t always work the same way, since changing propensity to consume based on both actual factors like income and debt and also psychological factors like fads and peer behavior. Not only have economics factors after the crisis affected consumer behavior, but also the psychologically depressive effects of it.

    I think we can view the global financial crisis that originated in the US subprime RE market on this analogy, using Minsky’s financial instability hypothesis, as well as the aftermath.

    It’s becoming clear that the quality of credit counts as Minsky noted, and that quality deteriorates over the financial cycle as euphoria over the trend replaces prudence regarding risk. In this context, the crisis was engineered by lenders seeking to improve position by loading up on risk with the expectation of selling it on through securitization as well as a lot financial innovation supposedly to spread risk.

    The real reason seems to have been moral hazard, perverse incentives and “control fraud” (William Black). This led to a game of “musical chairs,” quoting former Citi chairman and CEO Chuck Prince. Prince was named as one of the CEOs of big banks that didn’t see the crisis coming.

    I was in California near ground zero at the time of the crisis was building and it was abundantly obvious that the game was rotten through and through, from the brokering, to the appraisals, to the underwriting. Traders knew and were just waiting for the signal to run for the door, but like everyone else, they were going to dance while the music was playing, figuring that they would be the smart ones getting to the door first. Some did and some didn’t. Rationally irrational?

    It was not until later than I learned what was going on inside the banks, though. The order was out from the top to sent on as many mortgages that could be underwritten, pretty much regardless. Risk was almost completely discounted. Why? There was a huge appetite for safe assets (AAA), and the rating agencies were complicit enough with the banks who paid them from their services to rate dreck triple A. Again, it was rational (perverse incentives, moral hazard) to be irrational (imprudent wrt to risk, and even violate the law).

    Of course, when the music stopped everyone ran for the door as they had been planning to, but the door opening was not only not big enough but quickly narrowed, creating a glut of real assets rather than a shortage of financial ones.

    This was entirely foreseeable as being based on “rational” irrationality as long as the music was playing and there being no science of bubbles, no one could tell how long the good times would last. The old hands that had seen boom-bust cycles in real estate were under no illusions about the eventual outcome, however. It was a repeat performance on a grand scale. All holidays rolled into one.

    What got me interested in economics was my astonishment at finding that economists not only had totally missed the crisis but also lacked an explanation for it. The “rates too low too long” was insufficient and CRA explanation of subprime as the culprit was ridiculous. Then I discovered Minsky’s financial instablity hypothesis, Fisher’s debt deflation theory of depressions, and the forensic analysis that revealed the rational underpinnings of irrationality in moral hazard, perverse incentives, sharp practices and, indeed, crime.

    Crime is the ultimate it rational irrationality, or perverse rationality. The problem with rationality as an explanatory device is that even some of the most extremely irrational behavior can be cast in a rational light. So “rational” has to be given a technical sense to be meaningful. This is the idea that randomly sampled behavior follows a pattern (normal distribution), which is defined as rational. But as Taleb points out, the tails matter. In a normal distribution there are no fat tails.


Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

Blog at WordPress.com. | The Pool Theme.
Entries and comments feeds.