November 7, 2006 Letters to the Editor

Updated: November 17, 2006.

Short-Term Price Fluctuations

I received this letter from Rob Bennett.

The article "Short-Term Price Fluctuations" is helpful and important, John.

There are many who think of the risk that comes with investing in stocks at times of high valuations in a one-dimensional way. They see the risk as being the possibility of a wipeout, but decide to just hope that they can avoid the wipe-out, thinking that if that happens their investment choice will be proven "okay." It's not so.

There are all sorts of possible paths for stocks to follow at high prices, and all of them are bad. There could be a wipeout, which is obviously bad. There could be only small losses, but small losses that extend for far longer into the future than the wipeout would have (because it takes longer to return to fair value when you do it a little at a time). That's also bad. There could be years of dead money, with stocks staying at the same prices and inflation bringing about the gradual return to fair value. Also bad. There could be dramatic gains for a few years. That's ALSO bad. For the long-term investor, experiencing gains from high valuations means experiencing even larger drops in future days than you anticipated when buying into the asset class.

The best scenario of all is probably for valuations to remain for a long time at the same high levels while stocks experience price gains of 6.5 percent real year after year. That way you would be getting a return that beats that available from alternative asset classes for an extended period of time. Getting in at the sorts of prices that prevail today, however, you would need to see many years of 6.5 percent real returns to counter the negative effect of the huge price drop when it comes. The likelihood of this scenario playing out is exceedingly small, probably smaller than the likelihood of making a killing in Las Vegas. Putting down your money with the thought of seeing this sort of outcome is gambling, not investing.

For the long-term investor, it is never only the short-term that matters and it is never only the long-term that matters. The long-term investor must care about both, as the long-term is comprised of a series of short-terms. It is by surviving a string of short-terms that we become long-term investors not just in our minds but in the actual real world where money is earned and spent. The investor needs to look at the short-term possibilities, see whether he can realistically expect to be able to live through the worst of the short-term possibilities without selling, and then check whether the realistic long-term possibilities are to his liking.

The fact that one aspires to be a long-term investor does not mean that one should ignore short-term possibilities. It is by learning what the short-term possibilities are IN ADVANCE of putting down money that one develops the inner conviction in one's investment choices needed to become a long-term investor in the real world. Worst-case short-term scenarios do not come as a surprise to valuation-informed investors. Thus, valuation-informed investors are able to hold their stocks through the worst sorts of short-term scenarios that the market can dish out.

I believe that it is unrealistic for those who ignore the historical stock-return data to believe that they will be able to do this (the size of the price drops that follow huge bull markets are so large that it is not possible for the human mind to imagine them without reference being made to the historical record). This is demonstrated by the fact that there has never been a time when middle-class investors were able to successfully practice long-term buy-and-hold investing starting from the sorts of valuation levels that apply today.

HERE IS MY RESPONSE

Thank you.

You will be interested in my latest, which I added today (November 7, 2006).

Short Term Price Fluctuations
More about Short Term Price Fluctuations

I have constructed tables of conditional probabilities. They may be easier to understand. They show what happened to prices when P/E10 fell within specified ranges during the years of 1921 through 1980.

Short Term Price Fluctuations (Continued)

What is P/E10 ?

I received this question from Frank.

"What is P/E10?"

HERE IS MY RESPONSE

Thank you.

This is what I wrote in Why PE10?

"P/E30 is the current price (index value) of the S&P500 divided by the average of the most recent thirty years of (trailing) earnings. P/E10 is the same as P/E30 except that it uses the most recent ten years of (trailing) earnings. All terms include adjustments for inflation."

P/E10 is Professor Robert Shiller's preferred measure of stock market valuation. He includes it in his S&P500 database, which is the standard. He uses the latest month's index value (price) and the previous 120 months of earnings, where all values are adjusted for inflation. He excludes the current month's earnings from his calculation.

I base most of my calculations on the January values of P/E10. When I vary allocations in accordance with P/E10, I make changes only in January, not during the year.

Professor Shiller has stopped updating the P/E10 data at his regular Online Data page. These days, he maintains P/E10 at his "Irrational Exuberance" web page.

Why PE10?
Professor Shiller’s Web Site
Professor Shiller’s Online Data Page
Professor Shiller’s Irrational Exuberance Web Site

Demographics

I received this letter from Demographics Inquirer.

Harry S Dent has a free PDF update on his web site explaining his latest views on demographics. Harry thinks the market may turn down starting some time in late 2009, give or take a year or so. I was wondering what you think of his time table for a return of the market to historic P/E multiples.

Harry S. Dent Web Site

HERE IS MY RESPONSE

Demographics are a major influence in today’s market. Many times, what seems to be unusual turns out to be typical as soon as you take demographics into account.

Harry Dent attempts to predict what will happen over all time frames. He is very specific, even in the short-term. I do not attempt such precision, especially for the short-term. I am willing, however, to make some observations related to his article.

I do not consider it remarkable to expect that “the market may turn down starting some time in late 2009, give or take a year or so.” What is remarkable is Harry Dent’s strong assertion that the market will continue going up from here.

I would describe today’s market as being a Bull Market rally within a secular (long lasting) Bear Market. A two or three year upswing of 50% or 60% would be well within historical bounds, although somewhat unusual. Harry Dent is quite specific with specific numbers.

In his book Bull’s Eye Investing, John Mauldin mentioned an academic article related to demographics at the end of Chapter 1. It is Demography and the Long-Run Predictability of the Stock Market by John Geanakoplos, Michael Magill and Martine Quinzii. August 2002.

Demography and the Long-Run Predictability of the Stock Market

I draw attention to Figure 3, “The middle-young (MY) cohort ratio.” This ratio approximates the number of people at their peak earnings years (today’s Baby Boomers) divided by the number of people who are just starting out. The figure shows a peak in 2002 and a decline to a bottom around 2016 (with a broadly defined minimum extending from 2012 through 2020). The year 2009 is exactly half way between 2002 and 2016.

This result is consistent with how Harry Dent selects dates. He routinely starts out by selecting the midpoint of an “S-Curve” (cumulative probability versus time, ranging from Innovation to Growth to Maturity).

My own assessment is that we are currently in a secular (long lasting) Bear Market. It has far to go. It will have Bull Market rallies along the way. It is almost certain to hit bottom before 2020. In addition, I expect to see a recession within the next 5 years simply because we have not had one recently (business cycle). I am virtually certain that we will have a recession within the next decade.

There are other ways to look at dates. They produce numbers in the same ballpark.

In New Standards for Financial Reporting, I stressed the need for confidence limits. I credit Harry S. Dent in satisfying this need, at least, minimally. I interpret “give or take a year or so” as being an inner confidence limit, spanning the range from unlikely to likely. Doubling this range to plus and minus two years produces the outer confidence limits, spanning the range from highly unlikely to highly likely. In that sense, I define the outer confidence limits of Harry Dent’s estimate as 2007 to 2011.

New Standards for Financial Reporting

Not commenting on the rally before 2009, Harry Dent’s forecasts fall within what I consider a reasonable range. I make no estimate as to whether we have a rally earlier, either for or against.

Mark Hulbert has discovered that 20% of newsletter writers exhibit skill at stock selection and that a different 20% of newsletter writers exhibit skill at short-term market timing (as measured using the Sharpe Ratio, mainly through reductions of volatility, and regardless of the rationale behind such timing). In all cases, skill must be measured over a long time period. Even a highly skilled newsletter writer will underperform for several years in a row.

I do not claim any skill at short-term market timing. I am not in a position to assess Harry Dent’s short-term predictions. I will note that I disagree on some of his specifics, regarding facts. His impressions, whether right or wrong, may be far more important in the short-term if they capture the psychology of market participants. Perceptions are often more important in the short run.

Readers might be interested in my article about The Next Recession. You can be comfortable if you delay stock purchases, even today.

The Next Recession

Demographics Inquirer was kind enough to follow up with this:

Thank you for your insightful comments John. You have given me much to think about. I noticed that Mark doesn't track Harry's newsletter, so I can't figure out if Harry is in the top 20% of market timers with skill or not. I will ponder this some more.

Dent, Bogle, and Russell/Bennett

Rob Bennett adds some interesting comments. They help put everything into perspective.

Comments you put forward in your response to the recent letter about Harry Dent's demography-related predictions crystallized something in my mind.

You have on many occasions found fault with stock analysts for not including confidence limits with their predictions. I have a theory for why it became a general practice not to do so. Most people who read predictions don't understand the math stuff. They just want to know the bottom line. Since they don't understand the math, what they are really placing their confidence in is the expert. They have made a decision that they like this particular expert and are willing to put trust in him or her.

Which inspires more trust to someone who doesn't understand the math stuff -- a specific prediction or one with qualifiers and caveats and confidence ranges attached? The more specific number generally inspires more confidence for those who don't understand statistics (only those who understand statistics fully appreciate the benefit of confidence limits).

It may be that people like Harry Dent offer specific predictions without confidence limits in an effort to win the trust of readers who do not appreciate the added accuracy provided in predictions that include confidence limits. I'm not saying that this makes it the right thing to do. I am trying to explain the behavior.

Now, let's consider Bogle.

One of Bogle's breakthroughs was to advance what might be termed a "Know-Nothing" approach to investing. He says that you don't need to study the companies. He even goes so far as to suggest that you may be better off if you do not study the companies.

Bogle is gaining the confidence of his readers through use of a very different strategy than the Dent strategy. He is not putting himself forward as the all-knowing seer. He is saying "I don't know, but at least I know that I don't know." That claim possesses a good bit of appeal to those who have been burned by the inaccuracies of the predictions (lacking confidence limits) of the all-knowing seers like Dent. Many of those who don't understand the statistical realities are driven from putting too much faith in overly specific predictions (the Dent approach) to putting too much faith in "You never can tell" claims (the Bogle approach).

I believe that you and I are traveling a different path, a path that is a mix of these two failed approaches. (My view is that the failures of the Dent approach were pointed out by Bogle, and that the Bogle approach has been shown to be a failure during The Great Safe Withdrawal Rate Debate -- many of Bogle's claims cannot be defended in civil and rational discourse.) The reality is that we can know some things and not other things, and the effective investing analyst should be explaining to his or her readers which sorts of questions fall on which side of the line.

Dent is overstating his knowledge. Bogle is overstating his lack of knowledge. The reality is that there are some things that we can know with a high degree of confidence, there are some things that we can know with a modest degree of confidence, there are some things that we can know with a small degree of confidence, and there are some things that we cannot know at all.

HERE IS MY RESPONSE

I certainly agree with your assessment of Bogle’s knowledge. I think that he knows far more than he lets on. Otherwise, his books wouldn’t be nearly so clean when viewed from an engineer’s perspective. How else would Bogle identify the “right” issues, supply the “right” data, keep everything consistent and get everything “right” when it comes to numbers?

I suppose that my engineering background has a large influence on my treatment of numbers. Engineers always need to understand the reliability of numbers. Engineers always need to understand the applicability of their calculations. Engineers must always search for hidden flaws.

Experience teaches us that flaws don't remain hidden.

Letters to the Editor in 2006

Letters to the Editor in 2005