…and the Financial and Economic Crisis
I don’t read very many books. At least not during the academic year. But I have read two books recently that are quite germane to thinking about the buildup to the financial crisis, and thinking about how to respond to the current economic downturn. The first is Akerlof and Shiller’s Animal Spirits. The second one is actually not yet out — it’s Justin Fox’s The Myth of the Rational Market (I got a prepublication copy; here’s a hint of it). They are both important books, well worth reading.
As one can guess from the titles of these two books, neither text is a paean to the predictive power of the neoclassical view of the world. I’d expect that most readers trained in this tradition would then skip this blogpost. But before you do, and go back to reading your financial industry newsletter, you might consider where these seemingly neutral phrases such as “risk appetite” come from. Why did “risk” seemingly disappear in 2005-06, only to reappear in 2008? Why did asset prices (including house prices) climb so much? Think about the traditional asset pricing (Gordon Growth equation):
Pt = Dt/(ke – g)
Where P is the stock price, D is the dividend today, ke is the equity discount rate, and g is the deterministic growth rate of dividends; and ke equals the risk free rate plus the equity risk premium.
Where does this equity risk premium come from? Well, it could come from the covariation of real equity returns with the ratio of the marginal utility of consumption. Or it could come from waves of excess optimism and pessimism. Or it could come from both – although one would need to take a stand on the relative weights of the two effects.
I’m sure there are readers out there at this moment saying “But what about the finding that stock prices follow a random walk; this is consistent with the efficient markets hypothesis.” But as Fox writes:
Fama had proposed that the way to test the efficient markets hypothesis was to see if stock price movements obeyed the dictates of the capital asset pricing model, but this was only a relative test. It might reveal whether stock price movements made sense in relation to each other and the overall market, but it was no help in showing whether the overall market was correctly price or not. (p. 184)
Digression 1: To place this in technical terms, the maintained hypothesis is the statistical null hypothesis of no predictability. Failure to reject to no-predictability null is consistent with the random walk hypothesis, and hence the weak form efficient markets hypothesis (EMH). However, it is very difficult for most statistical test to differentiate between no predictability and little predictability. Jeffrey Frankel has called this econometric approach “the zen of perfect nothingness”.
Digression 2: Larry Summers (J. Finance, 1985) showed how a financial market pervaded by strong and persistent deviations from the fundamentals (“fads”) would take five thousand years to have 50% chance of distinguishing it from a truly efficient market, defined in the statistical sense Fama defined it.
To the extent that these waves of optimism and pessimism — heck why not call them animal spirits — are a real world force, then this argues that self-regulation of the financial industry is not likely to be sufficient. It’s not a prima facie case in support of government regulation (especially if regulation can be captured, a la Simon Johnson‘s regulatory capture view). But I think we should at least try to regulation and/or other means of slowing down the excess movements (financial taxes could in principle work this way, as in the Tobin tax).
Akerlof and Shiller conclude Chapter 11 thus:
…financial markets require regulation. And sometimes, when these regulations fail, because of all the feedbacks between financial markets and the real economy, there is also room for thoughtful, careful policies of financial insurance. Rededication to protecting the financial consumer must be one of our highest economic priorities.
In an emergency, as a backup, when we do get into a recession, there is a monetary and fiscal policy. But we know that there are limits to such policy and to its effectiveness. It is now time to redesign financial regulations to take account of the animal spirits that often drive the markets, to make the markets work more effectively, and to minimize the extent to which we will need after the fact bailouts to get us out of the hole.
See also the oped in yesterday’s WSJ.
By the way, the authors of both books agree that the point is not to dispense with price theory. That point is made on page 2 of Akerlof-Shiller, and in a quote of one of the fathers of behavioral economics, Richard Thaler, regarding the Shanghai apartment market: “Maybe it’s just supply and demand” (p. 286). Rather, the point is that we need to augment price theory with a more realistic depiction of human behavior if we are to avoid a repeat of the current situation.
I leave with two figures from Robert Shiller’s paper “Do stock prices move to much to be justified by subsequent changes in dividends,” American Economic Review 71(3) (June 1981).
(Hint on interpretation: The dashed line should be more variable than the solid line, under the EMH and assumptions regarding stationarity. See Engel (2004) for updated discussion.)
I went through some of the write up, not all, and I likely missed some points. My initial concern was that we assume all “estimators” are using the aggregate economic discount rate, and thus are all investing over the same horizon.
It would seem to me, the correct model would be a two state estimator. Keep an ongoing estimate of the equity premium of the investor group and an estimate of stock price.
Under a two state formulation the state model relationship between the equity premium and stock price would be difficult to define, and if not defined by the state model, the estimators themselves would have a natural unit root.
What am I missing?
“I don’t read very many books.”
I would suggest you start with Hume, Smith, Hayek, Keynes, and Bastiat. If you want modern thinkers, go for Mandelbrot, Taleb, and Kahneman (and perhaps, John Gray).
I never take advice from people who don’t read, and am particularly skeptical of academics who only read academics from their narrow discipline. It’s a big ‘ol, complex world out there…
Russell, don’t confuse not reading many books with not being well read (or well published). I learned a lot more from Menzie’s post than I did yours (does Taleb really belong in your list?)
I think digression 2 gets at the heart of it. There is so much volatility and uncertainty in financial markets, that it seems obvious that intervention should help.
Interesting post, a couple of points
1)I always had a problem with variance bound tests. We only observe the actual path of dividend payments. In pricing the market, all potential future paths of dividends should contribute to the calculation of expected returns. I have always found the Shiller approach, while interesting, misspecified.
2)Didn’t Jim’s wife publish a piece that showed Shiller’s work suffered from small sample bias that once corrected for changed the result.
3) The financial industry is probably one of the most regulated. From first hand experience, I can say that the mess of rules and regulations make little sense. Many of the rules in place have never been enforced. For instance, naked short selling and the abuse of the DTC clearing process has been going on for decades regardless of who controls Congress or the Whitrehouse.
4) While conceptually interesting, the EMH and the informational content of prices has always left me cold. I have never understood why the profession doesn’t just focus of arbitrage and drop the other stuff. I think the law of one price driven by arbitrage a much cleaner story.
5) A little off topic, but I have also never understood the complaints about rationality. I always though that the assumption of rationality was really nothing more than the assumption of revealed preference. We assume that individuals make decision according to their preferences, which allows one to infer things about utility functions from observed behavior. I have never though rationalty was a very strong assumption.
In the spirit of having to augment any model derived from limited data and sloppy logic, here are two quotes from the Fed’s Stress Test release. My contention is that a year ago, few economists would have taken issue with either statement (I’m assuming the Monte Carlo simulation uses a normal distribution):
For other consumer loans and for commercial lending (including various types of commercial real estate lending), the agencies estimated loss rates using techniques such as regressions of historical charge-off or default data against macroeconomic variables such as home price appreciation.”
Supervisors evaluated firm loss estimates using a Monte Carlo simulation that projected a distribution of losses by examining potential dispersion around central probabilities of default.
One day, statements such as these will be seen as heresy rather than dogma. Right? Maybe not…
mattyoung: I’m not sure I’m following your argument, but in any case, I’m sure there are many potential refinements to the asset pricing model I sketched that would yield some sort of multi-state behavior.
russell: Gee, I didn’t say I didn’t read a lot. I just said I didn’t read a lot of books particularly during the academic year. Someday we can sit down together and compare the breadth and amount of topics in our respective cumulative reading lists. I won’t to presume the answer ahead of time in the absence of additional information.
Nelson: (1),(2) Indeed, the Shiller and LeRoy papers sparked an entire literature. That’s why I included the link to the 2004 Engel paper, so you could see that (like almost every other statistical paper), the hypothesis testing is joint in nature.
(3) I agree that the financial sector is probably the most regulated on around the world. That doesn’t mean that it’s at the otpimal level, nor does it mean those are the right regulations.
(4) There is a focus on arbitrage conditions (of course, that means something about your assumptions about risk aversion…). You should see Summers’ great parable about ketchup economists.
(5) I make a distinction between the efficient markets hypothsis (agents make unbiased guesses about future values, etc.) and rationality (e.g., agents use available information up to the point where the marginal cost of acquisition and processing equals the marginal benefit).
Presumably Taleb is on Russell’s reading list because they are both arrogant [edited for content – mdc] ?
My point is he rise and fall of the stock market changes the equity risk premium, via a circular path which is not modeled. Stock traders measure the resultant change in risk premium after the lag. Agents cannot model this “wealth effect” except as unrevealed information, an output. So the delay in risk change tomorrow from a price change today is always with us, and always creates cyclic behavior in price.
I did my undergraduate and half my graduate education in a Commonwealth country. Our political studies community tended to emphasise cultural and psychological factors and disparage rational actor explanations. Now I am studying in the US, where the field is called political science and rational actors are everywhere. Perhaps, then, US books about animal spirits as key explanatory variables should give me a warm nostalgic feeling, but I am hesitant. Is it not more accurate to say that in this case individuals did behave rationally, but that the sum of their conduct produced undesired outcomes for the system as a whole? For instance, the people selling mortgages to NINJA borrowers were not overcome by fits of madness or eccentricity; they rationally responded to demand. That demand came from people who believed they could manage the risk by packing mortgages together and selling slices. Much of this confidence was founded on sophisticated mathematical models that hardly anyone understood. Was it irrational to have faith in these oracles? Rational actor theories almost always carry provisos of bounded rationality, and I think that hedge works here, too. The subject was too complex to be properly understood in full, so it was necessary to model and simplify reality to interact with it. The people who could do the modelling were not the same people who lent the money or ran the banks, but it was rational to divide the tasks between those best-equipped to do them. (This is rational in the same way it is rational for me to get someone else to fix my computer, even though I do not really grasp how she does it.)
Essentially, the people close to each little bit of the reality knew the risks were inappropriate, but also knew their self-interest meant making loans anyway. The people actually getting the risk had reason to believe they were immune. They were not right about this, but were they really irrational?
This does not mean advocating regulation is wrong-headed; I am just arguing rational actor theorem need not be abandoned in favour of animal spirits in this particular case. If this is true, though, regulators might not have been able to stop the madness, because they would have faced lots of smart-looking people happy to give them very complicated reasons why it was not madness after all.
Excellent posting by DCF. Geanakoplos also makes the point that there was nothing irrational about individual behavior in his well-received paper about leverage cycles at the recent NBER conference.
Menzie wrote:
As one can guess from the titles of these two books, neither text is a paean to the predictive power of the neoclassical view of the world. I’d expect that most readers trained in this tradition would then skip this blogpost. But before you do, and go back to reading your financial industry newsletter…
My oh my, has someone touched a nerve?
Since the start of the Progressive era government has attempted to thwart the discipline of the market. At the beginning of this period businesses were producing more, better, and cheaper products, but government wanted a bigger piece of the action.
The 20th Century was the bold experiment in government intervention (confiscation) resulting in the Great Depression and the Great Inflation. The only time the creeping cancer of government intervention was slightly turned back was during the Reagan administration followed by 20 years of prosperity.
Well, government intervention has returned with a vengeance. We have an administration that is Teddy Roosevelt, Woodrow Wilson, FDR, and Richard Nixon all in one package.
There have been apologists for a return to mercantilism from the time real scientific, liberal economics came on the scene, but when they have acquired power they have always created economic disasters. J.M. Keynes is perhaps the most destructive not because his theories are anything new since the 17th and 18th century mercantilists, but because they came at a time when the folly of government intervention was being exposed by the Great Depression. The politicians were desperate to hang on to their power and so Keynes became the politically accepted theorist giving them the cover they needed.
Pushing Keynes into the economic departments of higher education – right beside Marx – was not difficult for those holding the purse strings.
So, today the academic purse strings are almost totally controlled by those who exercise power and anyone who would stray from the approved doctrine finds themselves on the outside, pushed out of academia.
But that is actually a good thing. Those people have moved into the world of business where they actually produce and many have shared their wisdom in “financial industry newsletter(s)” because academia is still more interested in bleeding the patient.
The battle is being waged and the demand side stimulators through the actions of George Bush and especially Henry Paulson have given the supply side a severe blow, and we have reaped what they have sown, the current economy tells the tale.
Will the demand side stimulation of housing and credit markets win or will the supply side of product creation and prosperity win? The answer will be found in how much liberty and freedom is finally lost or gained.
Mish on government intervention Japanese style.To see the graph follow the link below.
http://globaleconomicanalysis.blogspot.com/2009/04/money-multipliers-velocity-and-excess.html
Japan Government Debt vs. Economic Recovery
By weakening the private economy, government borrowing is not an inflationary threat. Much light on this matter can be shed by examining Japan from 1988 to the 2008 and the U.S. from 1929 to 1941. In the case of Japan government debt to GDP ratio surged from 50% to almost 170%. So, if large increases in government debt were the key to economic prosperity, Japan would be in the greatest boom of all time. Instead, their economy is in shambles. After two decades of repeated disappointments, Japan is in the midst of its worst recession since the end of World War II. In the fourth quarter, their GDP declined almost twice as fast as that of the U.S. or the EU. The huge increase in Japanese government debt was created when it provided funds to salvage failing banks, insurance and other companies, plus transitory tax relief and make-work projects.
In 2008, after two decades of massive debt increases, the Nikkei 225 average was 77% lower than in 1989, and the yield on long Japanese Government Bonds was less than 1.5% (Chart 6). As the Government Debt to GDP ratio surged, interest rates and stock prices fell, reflecting the negative consequences of the transfer of financial resources from the private to the public sector (Chart 7). Thus, the fiscal largesse did not restore Japan to prosperity. The deprivation of private sector funds suggested that these policy actions served to impede, rather than facilitate, economic activity.
DCF hits it right on the head. Everyone in the mortgage story acted rationally. Its nothing more then a giant principal agent problem. All the decision makers in this mess were non-owners being paid fees to make decisions on owners behalf. Ownership has been so diluted at all levels of business as to be meaningless, and then there is always implied government guarantee against tail risk that makes people complacent.
DCF, PE and dave: I grant you there is an element of the moral hazard issue, as well as the “insurance” issue (as in Dooley, and examined in Chinn, Dooley and Shresta 1999, discussed in this post). But I still think it is important to distinguish the efficient markets hypothesis, which is a statement about the unbiasedness of expectations in a steady state equilibrium, from rational behavior, which might exhibit learning, or be better described as bounded rationality (which is in turn related to the menu cost approach to pricing behavior).
mattyoung: Interesting hypothesis. Can you write down the model and test it?
Menzie wrote:
But I still think it is important to distinguish the efficient markets hypothesis, which is a statement about the unbiasedness of expectations in a steady state equilibrium, from rational behavior, which might exhibit learning, or be better described as bounded rationality (which is in turn related to the menu cost approach to pricing behavior).
Menzie,
I would say that liberty and freedom are a much more important issue to discuss. The US is not the strongest economy in the world because we adopted 17th Century mercantilism.
Our economy is strong because men like Hamilton and Jefferson and Madison were reading men like Adam Smith and John Locke. Our entire society, including the economic system, was based on both individual freedom and personal freedom with the constitution intentionally (anti-federalists) limiting the intrusion of government into our lives. The 20th Century changed all that.
Today we are the strongest economic power in the world much like Great Britain at the end of the 19th Century from shear momentum.
You can put makeup on a pig but it is still a pig. You can dress mercantilism in equations and graphs but it is still mercantilism.
I have almost completed the paper by Phillip Swagel concerning the inner workings of the Bush Treasury Department and virtually every page describes some mistake or bad assumption made by the decision makers, not because the economists were bad at what they were doing, but because the nature of the political system and the limitation of knowledge (Hayek) works against efficiency. Because of the number of participants in the market it will always have an enormously greater knowledge of the economy than a handful of the brightest academics. The best academics imaginable simply do not have the knowledge to make decision that can apply to all circumstances in all places in the country.
“Expectations in a steady state equilibrium” are certainly not perfect, but markets are far, far better at managing economic issues than the “philosopher kings” locked in a government office. The analysis should be relative not absolute.
Ahh… the momentum of Intel, McDonalds, Google, Amgen, Paulson & Co etc…
LOL. The best book on the subject of the crisis is Thomas Woods’, Meltdown. He explains why the crisis was inevitable and shows that the majority of Austrian School economists saw it coming and was able to explain it long before others figured out that there was something wrong. Just because we may be skilled in mathematics is no reason to go down the wrong track by using models that will not work and have no predictive power.
Vangel: Er, would this be the carefully reasoned and deeply considered analysis of Thomas Woods, who writes:
Life must be so simple for people who want a monocausal view of complex phenomenon.
Well sounds interesting. I’ve been always fond of reading books specifically those that concerns economics. Maybe I should check this out