A conundrum: How increases in 'risk aversion' lead to higher stock prices - 12-Mar-07 - Generational Dynamics

Generational Dynamics Web Log for 12-Mar-07

A conundrum: How increases in 'risk aversion' lead to higher stock prices

Maybe because the global financial markets are increasingly "accident-prone."

As one pundit after another explains why the markets are in wonderful shape, and why last week's fracas was a momentary blip, and why the markets are poised to reach new high after new high, with no end in sight, it's always interesting to listen to the one or two mainstream pundits who are saying something different.

Michael Metz, chief investment strategist at financial investment firm Oppenheimer & Co., was Friday's token "bear" at CNBC. Here's what he said:

"You have margin debt at an all-time high. You have a great speculative boom in almost every equity market in the world. You have money dirt cheap - people leveraging.

It's going to result in some sort of accident - we've had one accident already. How serious an accident the next one is I don't know, but it's an accident prone system."

He's saying what should by now be completely obvious to regular readers of this web site -- with bubbles pushed to all-time highs around the world, thanks to massive amounts of private debt (not to mention public debt), all it takes is some kind of "accident" to bring things down. It's an interesting way of putting it.

But here's a strange question: Why have these bubbles continued to get bigger and bigger? As we've said, investors have been getting increasingly risk averse. So why are they investing in larger and larger bubbles? Why has the stock market has continued to go up for several years now?

A partial answer is given an interesting piece by Morgan Stanley economist David Miles, The piece is entitled, "What Happens When We All Get More Pessimistic and Think There Is More Risk?," and it's an attempt to analyze how the increasing risk aversion of investors is going to affect the marketplace.

Warning: The reasoning is complicated and even mind-boggling, but I'm going to go through it here step by step for two reasons: So that I can understand it better, and so that readers who are interested in macroeconomics can understand it better. Those interested in further information about integrating generational dynamics concepts into mainstream macroeconomics should read my October article, "System Dynamics and the Failure of Macroeconomics Theory."

Miles starts by reminding us that "Equities have fallen quite a lot pretty much everywhere across the world," and says, "The reason? Overwhelmingly the favourite answer is: a combination of a bit more pessimism about economic fundamentals and a lot more worrying about risk."

But he points out that it's a lot more complicated than that.

Miles: "If we are more pessimistic on growth, the likely increase in corporate earnings is lower — this is a negative for stocks." [Call this PRINCIPLE #1.]
This is the obvious reasoning that people have been using. If I'm pessimistic about the economy, then I believe that corporate earnings are going to be lower. If I think that the corporate earnings are going to be lower, then I'm not willing to pay as much for that company's stock, and if I believe that for the economy as a whole, then the entire stock market goes down.

Miles: If we're more risk-averse, then "we get a fall in the safe real rate of return...."
Now this is a lot more subtle, and requires a more lengthy explanation.
You may recall that early in 2005, Fed Chairman Alan Greenspan identified a "conundrum": Long-term interest rates around the world were lower than short-term interest rates. Why were long-term interest rates so low? That was the conundrum.
It turns out that economists started working to figure out an answer to Greenspan's conundrum, and they came up with an answer late in 2005 -- a controversial answer, as we'll see.
Now, Miles refers to the "safe real rate of return," which he defines that as the rate of return (interest rate) on investments in "inflation-proof government bonds." These interest rates are very strange these days, because the yields (interest rates) are inverted.
In "normal" times, long-term bonds pay higher yields than short-term bonds, but just the opposite is true today. So, the Fed overnight rate today is 5.25%, but a 3-month Treasury bond pays 4.92%, and a 10-year Treasury bond pays 4.57%. This is the conundrum -- in "normal" times, longer investment periods carry higher risks, so interest rates should be higher, not lower. In fact, all of these rates should be above the Fed Funds rate of 5.25%.
So why are they lower? That's the point: When investor risk-aversion increases, "We get a fall in the safe real rate of return ..." In other words, the interest rates on bonds falls -- in today's world, very, very sharply.
Explanation: When investors become risk averse, then they move their money into safe investments (government bonds). This increases the demand for government bonds, which means that their purchase price goes up (law of supply and demand), which means that their yield (interest rate) goes down. The higher the price of a bond, the lower the yield. [Suppose a bond pays $1000 ten years from now. If you buy it for $100 then you'll get a higher annual rate of interest than if you buy it for $90 -- the higher the price, the lower the yield.]

Next, Miles says that the lower interest rates on government bonds "is a positive for the value of all financial assets." Incredibly, this means that stock prices go UP. [Call this PRINCIPLE #2.]
This is the mind-blowing step, the one that you may have to sleep on, as I did.
If the "safe real rate of return" goes down, that means that if you want to make an investment that pays decent rate of interest, then you have to invest in something "non-safe," or rather, less safe. This is the way it works in "normal" times, too: Safe investments pay lower interest rates than non-safe investments.
So you have investors who want a decent return on investment to purchase stocks instead of safe bonds that pay little or no interest.

So now we have two factors pulling stock prices in different directions: Principle #1 says that stock prices go down, while Principle #2 says that stock prices go up. Which of these factors is dominant?
We have the answer empirically: Stock prices go up, and so Principle #2 is the dominant one. We can see that today, just by noticing that the stock market keeps going up.
So why is Principle #2 the dominant one? For this, Miles refers to a 2005 paper by Robert J. Barro, a Harvard economist, called "Rare Events and the Equity Premium."
By "rare event," Barro means things like World War I and II and the Great Depression. Barro builds a macroeconomic model that takes into account investors belief that there may be a "rare event" coming.

Barro finds that if investors believe that there's a 1.0% probability of a "rare event" within the next year then, under this assumption, then Principle #2 dominates by a substantial amount, so that stock prices go up. He shows this through his macroeconomic model, and he also supports it empirically by examining what happened in about 20 countries when they suffered these "rare events."
Barro quantifies this by identifying two values for a given set of conditions: The risk-free interest rate (the yield that government bonds pay) and the "spread." The spread is the additional effective yield (interest rate) that an investor can expect to earn by investing in "risky" stocks rather than risk-free bonds.
If p=.01 (i.e., p=1.0%) is the probability of a disastrous event in the next year, then the risk-free interest rate is 2.3%, and the spread is 3.6%, and so the expected yield from "risky" investments is (2.3%+3.6%) = 5.9%.
The following table lists the values of interest when the perceived probability of a "rare event" goes from p=0=0% to p=0.015=1.5%:
```
        | Perceived probability | risk-free |        |  "risky" |
        |     of "rare event"   |   rate    | spread |    rate  |
        | --------------------- | --------- | ------ |  ------- |
        | p = 0.0%              |  9.3%     |  0.1%  |   9.4%   |
        | p = 0.5%              |  5.8%     |  1.9%  |   7.7%   |
        | p = 1.0%              |  2.3%     |  3.6%  |   5.9%   |
        | p = 1.5%              | -1.2%     |  5.4%  |   4.2%   |

        Note: Column 4 = the sum of columns 2 and 3.
```
As you can see from this table, when there is no perceived danger of a "rare event," the spread is close to 0, which means that there is little difference between investing in risk-free and risky investments.
As an aside, a near-zero spread is the predicted result of macroeconomic models that don't account for rare events. That's Alan Greenspan's "conundrum," and why this situation has puzzled economists at least since the 1970s. No one could account for the fact that in the "real world," the spread was nowhere near 0, when the models predicted it should be close to zero.
As the table above shows, when the probability of a "rare event" increases from 0% to 0.5%, 1.0% and then to 1.5%, the risk-free rate falls, even going negative, and the spread increases. Thus, the more fearful people are of a stock market crash, the more likely they are to invest in stocks.
Huh? Did I really just type that? That's the conclusion of Barro's model, and it's also supported by the data he collected from 20 countries over the last century.
It's also what's happening today. Many people have been getting more concerned about a stock market crash, and yet the stock market keeps going up, and "risk-free" bond yields stay comparatively much lower.
We'll have a lot more to say about this conclusion below.

Now let's return to David Miles' original article. He says,

"On the other hand, we might just stop and wonder a bit whether what appears to be a nice logical "explanation" for falls in stock prices -- like "people got worried about risk and more pessimistic about growth" -- really does explain very much at all about what just happened."

Miles is making the point that Barro's model doesn't explain what happened in the last two weeks: Why did the stock fall, instead of continuing to go up?
That, of course, has a generational explanation, which we'll get to below.

From the point of view of Generational Dynamics, Barro's article is very important because it comes very close to recognizing generational theory.

Of course, it doesn't mention anything generational, but once you start talking about and analyzing "rare events," then the next step is to wonder how often "rare events" really occur, and from there, the only answer is generational theory.

I've said many times on this web site that pundits, analysts and economists seem congenitally unable to recognize generational effects, even the most obvious ones. Barro mentions in his paper that the "rare event" theory that he's developing, based on work done by T.A. Rietz in the 1980s, has received a very unenthusiastic reception from other economists. That's exactly the same as the unenthusiastic reception that Generational Dynamics gets.

Barro's paper could potentially be important to Generational Dynamics for other reasons. It provides a model for today's stock market debauchery, and this model could be enhanced with generational concepts to become a very powerful explanation of what's going on in the world today.

The following notes show how generational theory should be incorporated into Barro's work. Some of these notes are corrections, but most of them indicate improvements, including answers to questions that Barro himself asks.

There's an apparent contradiction in Barro's results. There are layers and layers of analysis and formulas to show that the risk-free interest rate goes down and the spread goes up, with the result that both bond prices and stock prices go up.
And there's a contradiction: In order for bond prices to go up, you need people to be moving their money from stocks to bonds (for increased safety). In order for stock prices to go up, you need people to be moving their money from bonds to stocks (for increased return). So you need people to be moving in both directions simultaneously, which is impossible.
However, it becomes possible when you realize it isn't that people have to move in both directions. What's needed is a lot more money in the system, buying BOTH stocks and bonds. In order for bond and stock prices to go up, you need demand for BOTH to increase, which means you need more money in the system.
When is there more money in the system? Well, the drop in bond prices automatically means that there's more money in the system, because the fall in bond prices means an increase in the supply of bonds from the Fed.
But more important, you need more liquidity in the system through credit -- through higher margin or other credit mechanisms.
So, for example, people bought those "tulip futures" in the 1630s for the Tulipomania bubble. A tulip future was a certificate that you bought in the fall. It gave you ownership to a particular tulip that would be grown in the spring. People must have borrowed money to purchase these certificates, secure in the "knowledge" that they'd end up making money, because the tulip would be even more valuable, which it was, until the bubble burst. But before that, the circulation of borrowed money would have increased liquidity, lowering interest rates.
Use of credit has to be added to Barro's model for it to make sense. Also, use of credit is a generational variable, since the willingness to use credit varies by generation, based on whether or not the generation lived through the previous "rare event" debt bubble crash.

In fact, Barro does consider the role of credit, but he discards it. He adds leverage rates (margin rates) to the model -- 20% and 40% -- and finds that they make little difference to the final results. He concludes that leverage (or margin or credit) is not an important factor.
This is a mistake because you need credit to resolve the contradiction just described above.
Furthermore, a moment's thought reveals that the use of credit is intuitively crucial to Barro's model.
The model is based on the "perception" of the probability of a rare event such as a stock market crash. What would cause investors to increase that probability? Obviously widespread use of credit would do it.
Barro's results may make sense with leverage values at 20% or 40%, but we're way beyond that today.
The hedge fund industry is a worldwide pyramid scheme (or Ponzi scheme), and many investors realize this, even if they won't say so out loud. And today's leverage amounts in the hedge fund industry are way above 40% -- well above 100% and probably over 1000%.
In fact, there's no obvious way to find out whether the "p=1.0%" probability figure makes sense, or whether the value should be should be measured in some way. It may turn out that the best way to measure it is by means of the use of credit, and of course the willingness to use credit is clearly a generational variable.

Barro uses economic data of most of the last century from 20 countries around the world in order to test his model and support the model's results. In interpreting the data, he makes several errors that he would have handled correctly if he understood Generational Dynamics:
- In analyzing each of the two World Wars, he mixes together all countries that fought in that war without distinguishing whether it's a generational crisis war or non-crisis war. For almost all countries, either WW I or WW II was a crisis war; there are no countries for which both wars were crisis wars.
- In analyzing data over a long period, Barro mixes completely independent data series together, which makes as much sense as adding together basketball scores and bowling scores.
  As far as I know, there are only two long-term cycles in economic data: The generational cycle (70-80 years) and the Kondratiev cycle (40-50 years), which appears to be technology related. These two cycles are completely independent of one another; sometimes they reinforce each other, and sometimes they cancel each other out. Today they're reinforcing each other.
  Beyond that, the only cycles are short-term. In the US, there is a four-year cycle synchronized with the US Presidential elections.
  Every researcher I've seen on these subjects makes the same damn mistake -- they add together all economic data without separating out the independent series. Without separating out the the different series, the results are gibberish. If Barro recomputes his results and separates the independent series, then I believe that he'll get very significant results.
Several paragraphs back I wrote, "Thus, the more fearful people are of a stock market crash, the more likely they are to invest in stocks." The model seems to support this. Why is this remarkable result true?
Here's how Barro explains it:

"Intuitively, a rise in p motivates a shift toward the risk-free asset and away from the risky one—this force would lower the equity price. However, households are also motivated to hold more assets overall because of greater uncertainty about the future. ... In any event, the risk-free rate falls by more than the risky rate, so that the spread increases."

(Recall that "a rise in p" means that the perceived probability that stock market crash will occur is rising.)
Barro's explanation isn't very satisfying. In the face of increased belief that a stock market crash is coming, why would "households [also be] motivated to hold more assets overall because of greater uncertainty about the future"? That doesn't make sense to me.
In fact, this is another apparent contradiction: Once the stock prices have gone up, if you're expecting a crash, then why wouldn't you sell your stocks at that point? The answer, of course, is that you expect them to keep going up. If you're in a bubble, then you expect the bubble to keep expanding.
This is a subject that I keep railing about on this web site. I keep expressing shock and surprise that investors are so stupid as to keep investing in a bubble. Don't they know they're going to lose everything in a crash -- a crash that they expect?
Well, Barro's model tells us that what is happening today is what we should expect. The giddy craziness of today's global financial system is exactly what we should expect. But why?
The only way to make sense of this is to incorporate a generational view of the willingness to use credit. In the decades immediately after a crash, a society's use of credit is very limited, because of fear of another crash; but as the survivors of the last crash retire or die, the use of credit increases.
Now, as we said a few paragraphs ago, it may be that the best way to measure the value of p (the perceived probability of a stock market crash) is through the use of credit; as society uses more and more credit, resulting in bigger and bigger bubbles, there's a two-way pull on investors:
- Investors increasingly fear a crash (p rises), and so they're motivated NOT to buy more stocks;
- Credit is increasingly available (along with p), so investors are motivated to use credit to buy more stocks.
This is another example of a "push-pull" kind of thing that can be modeled, once the generational cycles are sorted out and put into the model. It should then be possible to show that the second of these two factors dominates the first one.
There's only one more thing that has to be added: People stay in the market, even though they fear a crash, because they believe that they can see the crash coming and get out in time.
In fact, here's a twist: You may wonder, as I have, why investors didn't learn from the Nasdaq crash in 2000. The answer is, I believe that they're kicking themselves because they should have seen what's coming and didn't get out in time, and having lived through that crash, they think that they know how to beat the next one.

Barro says that to test his model, you should analyze different values of p (along with another of his disaster variables, q):

"However, the results also show the effects from permanent changes in any of the parameters, such as the disaster probabilities, p and q. In this and the following sections, I use the model to assess the effects from changes in p and q. However, in a full analysis, stochastic variations in p and q -- possibly persisting movements around stationary means -- would be part of the model."

His suggestion of using "stochastic variations" means that random values of p would be chosen in the model, as it moves forward through time. This is the usual naïve technique, used by someone with no concept of generational cycles (which means just about everybody).
The most effective model would be one that shows p having a low value shortly after after a stock market crash, and then rising gradually through the entire 80-year period. Furthermore, the value of p should be correlated with the use of credit (or leverage).

Barro describes objections to his type of model posed by economists R. Mehra and E.C. Prescott. Here is his summary of their objections (where Rietz is an earlier researcher on the "rare event" theory):

"For example, the perceived probability of a recurrence of a depression was probably high just after World War II and then declined. If real interest rates rose significantly as the war years receded, that would support the Rietz hypothesis. But they did not."

Barro says that "I am skeptical that this probability varied over time in the way suggested by Mehra and Prescott," and indeed he should be. There is absolutely no way that the "perceived probability of a depression was ... high just after World War II."
At that time, the survivors of the Great Depression knew exactly why it had occurred: It was caused by the evil of buying stocks on margin, or using credit. And after World War II, almost no one used credit. Anyone who understood the use of credit would not have expected a recurrence of the depression.

Barro uses economic data from 20 countries to support his model, but the way in which he interprets that data is very confusing to me. Here is one paragraph where he discusses this:

"Changing probabilities of a depression would likely isolate the effect of changing p, but the analysis depends on identifying the variations in depression probability that occurred over time or across countries. From a U.S. perspective, the onset of the Great Depression in the early 1930s likely raised p (for the future). The recovery from 1934 to 1937 probably reduced p, but the recurrence of sharp economic contraction in 1937-38 likely increased p again. Less clear is whether the end of World War II had an effect on future probability of depression."

He's correlating values of p to economic expansion and contraction, but he admits it's only guesswork. In fact, one could just as easily argue the opposite: When the stock market rises, someone might expect a crash, but when the stock market begins to fall again, someone might think that the danger is past.
In fact, I believe that, in the end, it will turn out that p increases almost monotonically in the 70-90 year period from one crash to the next, and then falls sharply when the crash occurs, beginning the cycle all over again.

Finally, let's get back to David Miles' original posting. In that posting, he's looking for an explanation of the stock market fall in the last couple of weeks. He concludes with this:

"What all this means is that a pretty standard economic model can generate the prediction that more pessimism about average future growth and greater uncertainty about future growth will take stock prices a bit higher.
All this may seem largely academic and of little value to practical people who have to make money in financial markets. It may remind some of the old joke made at the expense of the head-in-the-clouds academic who on seeing some real world phenomenon observed that it was all very well in practice but it shouldn't work in theory.
On the other hand, we might just stop and wonder a bit whether what appears to be a nice logical "explanation" for falls in stock prices --- like "people got worried about risk and more pessimistic about growth" --- really does explain very much at all about what just happened."

As Miles points out, Barro's theory explains why stock prices keep rises, but it doesn't explain why the recent falls occurred, or why crashes occur at all.
The answer to that is simply that, sooner or later, the bubble has to burst.
Barro's theory, when generational theory is added, is really quite a remarkable explanation of why a bubble occurs in the first place. The fear of a "rare event" (stock market crash) is more than offset by the availability of credit that pulls investors into a larger bubble.
There certainly are people today who recognize the danger. At the beginning of this article, we quoted Oppenheimer's Michael Metz as saying that he's expecting "some sort of accident" in what is "an accident prone system."
In an article I posted last week, I quoted Douglas A. Kass of Seabreeze Partners Management as saying, "I think that there's a very good chance that the market, rather than test its low, crashes in the next several months."
What happened in the last couple of weeks is that the "perceived probability of a rare event" has really begun to catch up with investors.
Many investors believe (or, at least, fear) that a crash is coming, and must also believe that they're going to be especially clever and get out in time, because they'll see in advance all the signs that they didn't bother to see during the Nasdaq crash in 2000. Thus, when there's some kind of scare (caused by a chaotic, unpredictable event), all of these investors immediately believe that they've seen the sign, and they want to get out right away.
Well, of course, when a real crash happens, the huge numbers of sell orders will swamp all the computer systems of all the banks, investment houses, and exchanges, so that few people will actually get out in time. The rest will lose pretty much every penny they own, and most will be left with a mortgage debt and credit card debts they can't pay. Since many companies will go out of business, many people will become unemployed, bankrupt, and homeless. The ones who owe child support will all be sent to jail for failure to pay. Others will simply be dumped on the street to live under bridges and in cardboard boxes.
Those are the missing pieces that David Miles was looking for. Barro's theory explains what happens up to the point of the crash, and then full-scale panic explains the crash itself.

From the point of view of Generational Dynamics, Barro's paper presents an exciting theory that integrates very well with the rest of Generational Dynamics, and explains a great deal.

Those interested in further information about integrating generational concepts into mainstream macroeconomics should read my October article, "System Dynamics and the Failure of Macroeconomics Theory." (12-Mar-07) Permanent Link
Receive daily World View columns by e-mail
Donate to Generational Dynamics via PayPal

	Generational Dynamics
	Forecasting America's Destiny ... and the World's

\| HOME \| WEB LOG \| COUNTRY WIKI \| COMMENT \| FORUM \| DOWNLOADS \| ABOUT \|

Generational Dynamics Web Log for 12-Mar-07 A conundrum: How increases in 'risk aversion' lead to higher stock prices

Web Log Pages

Generational Dynamics Web Log for 12-Mar-07
A conundrum: How increases in 'risk aversion' lead to higher stock prices