J. Lyons Fund Management, Inc. Newsletter

March 2012
Posted March 1, 2012

Dow 10,263

An (unsolicited) addendum to Jeremy Siegel's Dow 15,000 call

Trivia Question: On this Leap Day, February 29, 2012, the S&P 500 closed at 1366 (rounded). What was the S&P 500's closing price 3 Leap Days ago (February 29, 2000)?

Recently, Barron's published an article touting research by Wharton Business School professor Jeremy Siegel. The crux of this research was that since the poorest 5-year periods in the stock market are historically followed by above-average 2-year periods, these next 2 years should be very rewarding for investors, with the Dow Jones Industrial Average likely reaching 15,000 and quite possibly 17,000.

The article received much fanfare, thanks in part to the ubiquitous Mr. Siegel's countless financial TV appearances and the magazine's cover which read "DOW 15,000" in roughly size 15,000 font. The piece provided ample fodder for both the professor's media sycophants as well as many detractors of his buy-and-hold philosophy in the investment community. Well, Mr. Siegel didn't ask us (and furthermore, I was only a B-minus student at Wharton, not the Russell E. Palmer Professor of Finance) but we'd like to throw in our 2 cents as well. Besides, unlike many of the folks that have offered their boisterous opinions on the article, we have actually done quite a bit of research on the subject and we have a couple of points to make that may provide interesting perspective for investors.

First of all, while "Dow 15,000" is a real attention grabber of a headline, particularly with the fascination of big round numbers on the part of investors to some extent and the financial media to an embarrassing extent, it really is not that ambitious of a 2-year target. Secondly, at first blush, the evidence that professor Siegel offers is both bullish and compelling. However, it is always wise (and imperative when managing other people's money) to perform the due diligence for yourself. Therefore, we ran some historical tests of our own in order to substantiate his results as well as to ensure that we are examining the complete picture when deciding whether or not to consider his research an appropriate guide for the current market.

Trivia Answer: Amazingly, on Leap Day 2000, the S&P 500 also closed at 1366. (More amazing still, on Leap Day Eve 2008, the S&P 500 closed at 1367 -- so much for Wall Street and academia's buy-and-hold mantra!)

This brings up our final point and one which we emphasize often, explicitly so in last month's newsletter, "Liters or Gallons?". Specifically, when making historical market comparisons, one should do so versus historical markets occurring at similar periods during similar market cycles. So while the Leap Day trivia mainly makes for a mere interesting factoid, there is also a helpful message in it; a market that goes unchanged over a span of 4 leap years is not a bull market. We feel rather strongly, as discussed in past analyses (see "Another Lost Decade Ahead?: 2011 Secular Bear Market Update") that we are still in the secular bear market that began in 2000. Therefore, any historical comparisons should be made versus other secular bears of the past and we have done so here.

Note: We have made the best possible attempt at conducting a similar study to Professor Siegel's given the data that we have. However, our results shown below will vary due to the following factors. The market data he uses entail a hypothetical investment in all stocks while we simply use the Dow Jones Industrial Average. His test begins in 1871 while ours goes back to 1900. Lastly, his data is adjusted for inflation and dividends, while ours is not (since we do not have that information). However, it is our opinion that over these shorter 2 and 5-year periods, the data differences do not affect the results enough to meaningfully alter the conclusions reached.

Additionally, while this study is a useful exercise, JLFMI's actual investment decisions are based on our proprietary models. Therefore, the conclusions based on the study in this newsletter may or may not be consistent with JLFMI's actual investment posture at any given time. Additionally, the commentary here should not be taken as a recommendation to invest in any specific securities or according to any specific methodologies.

Dow 15,000 Not a Big Leap

As mentioned, the conclusion of Mr. Siegel's research discussed in the Barron's article was that the worst 5-year periods in the stock market were historically followed by above-average returns over the subsequent 2 years. Siegel defined "poor" as falling in the bottom quartile of all 5-year periods which, according to his research, the 5-year period ending in 2011 did. According to our research, the recent 5-year period ranked in the 26th percentile of all periods, but close enough. This table displays the results of our work, including average 2-year returns following the various quartiles of 5-year periods, from the worst performing to the best performing. It also indicates the future level of the Dow assuming the 2-year returns, based on its 2011 closing price of 12,218.

2-Year Returns Following Varying 5-Year Periods
Dow Jones Industrial Average 1900-2011
5-Year Period Average Subsequent 2-Year Return Projected
Dow Level
Bottom Quartile (worst-performing 5-year periods) 32% 16,134
3rd Quartile 7% 13,067
2nd Quartile 10% 13,381
Top Quartile (best-performing 5-year periods) 11% 13,612

There is clearly a stark out-performance, on average, in 2-Year returns following the worst 5-year periods in the market. In addition, only 3 of the 28 such 2-Year periods were negative. Based on this evidence, Mr. Siegel's bullish forecast that the stock market is likely to put in a strong performance over the next 2 years is certainly well-founded. Curiously though, the professor makes it a point in the article to emphasize the likelihood of the Dow reaching 15,000 in two years. That seems like a rather modest goal given that the average 2-year return following poor 5-year periods would bring the Dow to over 16,000.

That is not to mention the fact that the Dow has already rallied over 6% since the end of the year bringing a 15,000 target that much closer. If the Dow merely repeats its performance of the past 3 months over the next 22 months, it will reach 15,000. Furthermore, the article even provides some wiggle room for the modest target by stating that the market is likely to reach 15,000 over the next two years, not necessarily close at or above that level in two years. Based on the average maximum gain within a 2-year period, it would be newsworthy if the Dow did not reach 15,000 at some point in the next two years.

2-Year Maximum Gains Following Varying 5-Year Periods
Dow Jones Industrial Average 1900-2011
5-Year Period Average Max Gain During Subsequent 2-Year Period Projected
Dow Level
Bottom Quartile (worst-performing 5-year periods) 41% 17,259
3rd Quartile 23% 15,078
2nd Quartile 24% 15,093
Top Quartile (best-performing 5-year periods) 30% 15,903
All Periods 30% 15,847

If the Dow merely put in the average max gain following all periods, it would easily surpass 15,000 at some point in the next two years. Furthermore, the average max gain following even the least ideal of the 5-year quartiles would be enough to reach 15,000. So one can see that it would be more noteworthy if the Dow did not reach 15,000 within the next two years.

So why set the bar seemingly so low? We suspect that Mr. Siegel, with help from his friends at Barron's, is sandbagging a little bit by trumpeting the 15,000 target. By using a high-profile number like 15,000, he is able to generate some publicity buzz, yet also leave himself a relatively easy target at which to shoot. If the market reaches that noteworthy but reasonable level, Siegel will be able to claim victory, specifically for his "bold" call, and more generally for his buy-and-hold philosophy (he is the author of the book Stocks For the Long Run and an advisor to the WisdomTree family of ETFs). Pretty savvy of the professor.

To be fair, despite most of the attention being focused on 15,000, the article and Mr. Siegel do concede that it is not that ambitious of a target. A lesser focus is made on Siegel’s conclusion that the Dow stands a 50-50 chance of reaching 17,000 sometime over the next two years. Since, historically, the average maximum gain after the worst-performing 5-year periods is strong enough to lift the Dow to just over 17,000, that seems like a more compelling drama to monitor. Feeling inspired, we broke down the historical data further in pursuit of a more complete story to this research.

Not all Poor 5-Year Periods are Created Equal

Clearly the stock market’s average historical 2-year performance following the poorest quartile of 5-year periods has been impressively bullish. However, that incorporates fully one quarter of all 5-year periods historically, or 28 periods in our study. Therefore, we sought to identify possible factors that would differentiate some of those 28 periods in order to find a more focused comparison to the current market.

As we mentioned above, the 5-year period ending in 2011 actually came in at the 26th percentile of our study over the past 112 years, just outside the worst-performing quartile of 5-year periods. Therefore, we suspect that since professor Siegel included this period in the bottom quartile of 5-year periods over the past 141 years, it likely came in very near the top of the bottom quartile. That is, the recent 5-year period was probably much closer to the 25% percentile than to the very worst performing 5-year periods. As it turns out, that is not a trivial point given the discrepancy between the average returns following the bottom half of the bottom quartile (i.e., the worst-performing 12.5% of periods) versus the top half of the bottom quartile (i.e., the 2nd worst-performing 12.5%).

Top Half vs. Bottom Half of Bottom 5-Year Quartile
Dow Jones Industrial Average 1900-2011
5-Year Period Average Subsequent 2-Year Return Projected
Dow Level
Bottom Half of Bottom Quartile
(i.e., worst-performing 12.5% of all periods)
43% 17,523
Top Half of Bottom Quartile
(i.e., 2nd worst-performing 12.5% of all periods)
21% 14,745

While the average 2-year return following 5-year periods in the top half of the bottom quartile is still very good at 21%, it is less than half that of the average 2-year return following the very worst 5-year periods. While this may be nitpicking a bit, the inclusion of the 5-year period ending in 2011 in the bottom quartile seems to benefit the current 2-year outlook by riding the coattails of historical returns following the very worst performing 5-year periods. At a minimum, this may at least dampen the rosiest of forecasts for the next 2 years.

Since we are on the topic of dampening forecasts, let's parse the data another way. There are many ways to skin a cat as they say, and so too there are many ways to arrive at a poor 5-year return. The period may be weak in each of the five years or it could be weak in the beginning of the period and strong at the end or vice-versa. For example, during the recent 5-year period ending in 2011, the 2-year period at the beginning was negative and the 3-year period at the end was positive. Taking our cue from that, we've separated the bottom quartile of 5-year periods according to those ending with a negative 3-year return vs. those ending with a positive 3-year return.

Bottom 5-Year Quartile w/ Positive vs. Negative 3-Year Return
Dow Jones Industrial Average 1900-2011
Period Average Subsequent 2-Year Return Projected
Dow Level
Bottom 5-Year Quartile & Negative 3-Year Return
41% 17,183
Bottom 5-Year Quartile & Positive 3-Year Return
19% 14,513

Again, we see that all poorly-performing 5-year periods are not created equal. Having a positive 3-year return seems to counteract some of the positive bounce-back effect stemming from a poor 5-year period. So while the outlook for the next two years, according to these historical averages, is still fairly bright, it is not nearly as bright as it would be if the Dow was coming off of a negative 3-year period.

Stock Market Performance Following Good 3-Year Periods

Sticking with the theme, we researched all 3-year periods back to 1900, regardless of  5-year returns. As it turns out, the 3-year period ending in 2011 was not only positive but it was in the top third of all periods over that time. As one might expect, 2-year returns following these good 3-year periods were not very robust.

2-Year Returns Following Varying 3-Year Periods
Dow Jones Industrial Average 1900-2011
3-Year Period Average Subsequent 2-Year Return Projected
Dow Level
Bottom Third (worst-performing 3-year periods) 20% 14,628
Middle Third 14% 13,919
Top Third (best-performing 3-year periods) 12% 13,661
All Periods 15% 14,066

Again, the market has historically been subject to mean reversion as good-performing 3-year periods have, on average, been followed by below-average 2-year returns. Furthermore, 13 of the 37 "top third" periods showed a negative 2-year return going forward. Based on the building evidence, Dow 15,000 looks like less and less of a sure thing, let alone Dow 17,000.

Stock Performance and the Secular Market Cycle

As we constantly preach, when making historical market comparisons it is crucial to avoid comparing apples to oranges (or "Liters to Gallons"). Doing so can lead to unsound and unreasonable expectations, which of course can lead to loss of capital. Therefore when considering the current (secular bear) market, it is, in our judgement, most appropriate to make comparisons with secular bear markets of the past. For starters, what is the average 2-year return during a secular bear market versus the return during a secular bull market?

2-Year Returns by Secular Cycle
Dow Jones Industrial Average 1900-2011
Secular Cycle Average Subsequent
2-Year Return
Dow Level
Years within Secular Bull Markets
22% 14,939
Years within Secular Bear Markets 9% 13,366

It is not a revelation that the returns are much lower during a secular bear market since it is, by definition, a market that moves sideways to lower over an extended period of time (note: the average subsequent 2-year return of 9% during a secular bear is higher than one might expect because the returns for years at the end of secular bear markets include the beginning of the subsequent secular bull.)

We also looked at years during a secular bear market in relation to 5-year returns. The 5-year period ending in 2011 was in (or near) the bottom quartile of all 5-year periods. However, when measured against just 5-year periods during secular bear markets, the recent period is just under the 50th percentile, or merely average. Therefore, there is nothing further to be gleaned from similar years, historically.

Relative to 3-year returns, however, is another story. The recent 3-year return ranked among the top third of all years historically. Compared to just the periods during a secular bear market, the 3-year return ending in 2011 almost ranks in the top 10%. These are the numbers broken down into quartiles.

2-Year Returns in Secular Bears Following Varying 3-Year Periods
Dow Jones Industrial Average 1900-2011
3-Year Period Average Subsequent 2-Year Return Projected
Dow Level
Bottom Quartile (worst-performing 3-year periods) 21% 14,780
3rd Quartile 18% 14,469
2nd Quartile 6% 12,933
Top Quartile (best-performing 3-year periods) -7% 11,309

The combination of a secular bear market and a good trailing 3-year return has not been a good one for the market. With an average 2-year return of -7% historically, the next two years would be bucking the odds to merely be positive, let alone reach 15,000. Incidentally there were only six years during secular bear markets that had better 3-year returns than the recent period and all six had negative 2-year returns going forward. If the secular bear market persists, which we think it will, this is not good news for the Dow 15,000 crowd.

Putting It All Together

That is probably as far as we want to drill down into historical comparisons lest we be accused of data mining. Since markets repeat similar cycles but are never identical, general lessons are typically the only reasonable takeaway from such research. However, just for curiosity's sake, to which other years historically do the characteristics mentioned above apply as they do to 2011? Specifically, to which years can these following traits also be applied?

  1. 5-year return in the bottom quartile of all 5-year periods
  2. 3-year return in the top third of all 3-year periods
  3. Occurring within a secular bear market

As it turns out, of the past 112 years there is only one year that falls into each of these categories, 1935. Again, it is foolish to make precise forecasts based on historical precedent, especially with just a single data point. However, it does make for an interesting comparison to look at the market's performance after 1935, the most similar year historically to 2011 according to the criteria above.

1935: The Best Comparison to 2011?

 Secular Bear
 5-Year Return Percentile 3-Year Return Percentile  Subsequent 2-Year Return Projected Dow Level
Yes 16th 100th -16% 10,263
Yes 26th 71st - -

Interestingly, not only was the 2-year return following 1935 negative but it was the absolute worst performance after any of the periods in the bottom 5-year quartile. A repeat of that performance through 2013 would leave the Dow at 10,263. Once again, we are not making that forecast but merely pointing out the possibility of a below-average 2-year performance despite the recent "poor" 5-year period. By the way, for those arguing against this likelihood on the basis of the market's strong start this year, consider this: over the first two months of 2012, the Dow has rallied 6%; over the first two months of 1936, the Dow rallied 6%.


Market Analysis

At first glance, Professor Siegel's research in the recent Barron's article makes a compelling argument for above-average returns over the 2012-2013 period. In actuality, Dow 15,000 is a rather unambitious target and was probably more of a product of marketing by Barron's due to the big round number, not to mention setting the bar deceptively low for a potential successful call by Mr. Siegel . We are not accusing the professor and the magazine of anything underhanded, but history tells us that even after the least ideal of 5-year periods, the market rallies enough to lift the current Dow up to 15,000 at some point over the next two years.

The other target laid out in Barron's of Dow 17,000, which Siegel says is a 50-50 proposition over the next two years, would better qualify as a newsworthy goal. But while the evidence -- that the worst-performing quartile of 5-year periods is, on average, followed by above-average 2-year returns -- is impressive, it's not the whole story. Our research finds that the rosy 2-year forecast implied by the recent "poor" 5-year period is mitigated by the following factors: the recent 5-year period falls at the top of the bottom quartile, the 3-year return is among the top third of all 3-year periods and the presence of a secular bear market. By historical tendencies, each of these factors would seem to make Dow 15,000, much less 17,000, less of a certainty in the next two years.

Investment Principle

While the historical comparisons are interesting, we are more in disagreement with Siegel's buy-and-hold philosophy than his research call. Dow 15,000 may materialize at some point in the next two years and as we said, it isn't that ambitious of a goal. For that matter, anything is possible...Dow 10,263...even Dow 17,000. But when it comes down to it, for most investors these numbers are meaningless. Any of the levels in the Dow could be tagged over the next few years but that doesn't matter unless one has an active management strategy to take advantage of it.

This is especially true considering the reality that we remain in a secular bear market, as evidenced by the span of four leap years with zero price appreciation in stocks. Therefore, emphasizing again the importance of making appropriate comparisons, previous secular bear markets make for the best guides for today's expectations. Thus, any significant rallies in the foreseeable future are unlikely to be sustainable, especially coming off of a very good 3-year period. That means that for buy-and-hold investors, Dow 15,000 in the next few years will likely mean as much to them as the October 9, 2007 Dow 14,165 print means to them now, over a thousand points lower.

Investors must adopt an active management process (or preferably a manager with such a process) whereby they can participate during rallies but also lock in those gains before they are erased entirely. Failing to do so can result in their investments spanning a period of four leap years without any equity returns to show for it. That is hardly a compelling case for buying-and-holding, especially considering that four leap years ago, investors were engaged in an orgy of buying tech stocks, many of which have not seen the light of day since. To continue with a buy-and-hold strategy and expect different results would constitute a real "leap" of faith.

Dana Lyons

Vice President

The commentary included in this newsletter is provided for informational purposes only. It does not constitute a recommendation to invest in any specific investment product or service. Proper due diligence should be performed before investing in any investment vehicle. There is a risk of loss involved in all investments.

Copyright (C) 2012 J. Lyons Fund Management, Inc.  All rights reserved.


Current Issue


Sign Up for FREE
Learn more about our active, risk-managed style of investing and why it may be the only chance for investment survival over the next decade. Sign up for our FREE e-newsletter.

Email Address: 
Follow JLyonsFundMgmt on Twitter