Identification of the secular bear market in advance
In previous articles I have introduced the concept of secular market trends and relative P/R, a valuation concept closely coupled with secular trends. A thorough development of this theory is found in my book Stock Cycles, but for the purposes of this article the summary given in the two referenced articles should give the reader sufficient background to follow the arguments presented here.
The secular trends concept holds that the long-term rise in stock prices displays a "stair-step" pattern. An interesting demonstration of this idea can be seen in Figure 1, in which the total return of a hypothetical S&P500 index fund relative to that of a hypothetical money market fund [1] is shown. The stair-step pattern is evident. The "risers" are the secular bull markets, periods like 1896-1906, 1921-1929, 1942-1968 and 1982-2000, when stocks handily out-performed money markets. The "steps" are the secular bear markets, periods like 1881-1896, 1906-1921, 1929-1942, and 1968-1982 when stocks didn't outperform money markets for 13-15 years. In Stock Cycles: why stocks won't beat money markets for the next 20 years, I forecasted a 20 year "step" that would likely begin in 2000.
Figure 1. S&P500 cumulative return relative to that of money market (log-scale) 1871-2002
The prediction was based on a valuation measure called price to resources (P/R). This measure gives similar results as the price to average earnings (P/E) measure used by Robert Shiller in his book Irrational Exuberance. Shiller divides the current index price by the 10 year trailing average of the index earnings to obtain his averaged P/E. Figure 2 shows plots of Shillers P/E and P/R for the last secular bull market that shows how both valuation methods rose more or less in lockstep with each other. The use of these measures is simple. One concludes that the secular bull market is nearing an end when the valuation measure gets "high" relative to its historical levels. For the spectacular bull market that has just recently ended, "high" meant reaching an all-time high. The previous all-time high for Shiller's P/E was 33 in 1929. This level was reached in January 1997 in the last bull market, and those who followed the Shiller method for market timing would have exitted stocks for money markets early in 1997. Dr. Shiller presented this analysis to the Federal Reserve in late 1996, which led to Chairman Greenspan's famous "Irrational Exuberance" speech in December 1996 [2]. We can take this warning as sort of a call for a top based on the averaged P/E methodology employed by Shiller.
Figure 2. Shiller's price to earnings (P/E) compared to P/R for the recent secular bull market
Similarly, P/R broke its all-time record level (set in 1901) in early 1999. I too issued a "call for a top" in fall 1999 that claimed that July 16, 1999 was "the end of the secular bull market that started in August 1982". I was early. I had already begun moved from stocks to money markets at the beginning of 1999 and as of September 3 of that year I was completely in money markets. Someone following Shiller's methods would likely have done the same and move funds from stocks to money markets throughout 1997. In Figures 3 and 4, I show the cumulative return seen in an S&P500 index fund as compared to a money market since January 1997 and January 1999.
Figure 3. Total return from stocks (S&P500 index fund) versus money markets after Jan 1997
Stocks have outperformed money markets for the entire period since January 1997 (as of June 2002). Thus, exitting the market in early 1997 in accordance with Shiller's P/E-based warning would not (so far) have given a prospective market timer an edge. In contrast, money markets have strongly out-performed stocks since Jan 1999. Thus, exitting the market in early 1999 in accordance with P/R has already provided a market-timer employing P/R an edge. From these observations, one might conclude that for the present market, P/R has given a better indication of overvaluation than Shiller's P/E.
Figure 4. Total return from stocks (S&P500 index fund) versus money markets after Jan 1999
As is now obvious, the great bull market that began in 1982 ended in 2000. P/R was better than P/E in forecasting this end in a useful fashion. Those who moved from stocks to money markets in 1999 have seen those assets placed into money markets outperform the stock index. Similarly, those early readers of Stock Cycles who moved from stocks to money markets in late 2000 (Stock Cycles was published in Nov 2000) have also avoided much of the subsequent underperformance of stocks since then. But the question today is, after falling more than 35% from its all-time high, could the S&P500 now be undervalued relative to future performance versus money markets? Characterization of past secular bear markets using P/R. Figure 5 shows P/R since 1999 compared to P/R for the first decade of prior secular bear market eras. Each of the eras shown begin in January of the year when P/R peaked (or when a secular bear market was "called" in the case of 1999). As the figure shows, P/R reached all-time levels early in 1999, showed a broad peak in 1999-2000 and has since fallen around 0.9. This 40% drop is consistent with the early behavior of previous secular bear market eras. The first bear market of each of the secular bear markets in Figure 5 showed drops in P/R of 34%, 33%, 83% and 46% One of the P/R traces in Figure 5 is very different from the others. This is the one from the 1929-49 secular bear market, which showed an 83% drop in P/R over its first bear market. So far the current secular bear market appears to be unfolding more like the other three secular bear markets and not the post-1929 disaster.
Figure 5. P/R for previous secular bear market periods
Although I believe that both the current secular bear market and the 1929-1949 were associated with the Kondratiev fall from plateau, I do not believe that this bear market will play out like the post 1929 one. I believe that since the modern era of economic management was installed during the Great Depression such a collapse is extremely unlikely for reasons I discuss elsewhere. The unusually strong economy and the mild nature of the current bear market (so far) as shown in Figure 5 support this belief.
Thus I believe that the other secular bear markets provide a better indicator of what levels to expect in the coming years. On the other hand, the timing information provided by the Great Depression bear should be valid and we will consider this information along with the others. It is clear that today we are in the midst of the first big bear market of what will be a string of big bear markets interspersed with weaker bull markets that together constitute the secular bear market. I expect this secular bear market to last until near the end of the next decade, so long term (e.g. 10-20 years) prospects in stocks are not expected to be good for a long time. Shorter term prospects will fluctuate, depend on position in the ordinary bull/bear market cycle.
Figure 5 suggests that this first bear typically ends anywhere from just shy of three years to as much as 4½ years into the secular bear market era. On average it ends about 3½ years into the era. This suggests that the current bear could have ended as early as last fall (as many thought) or last as long as until summer 2003, but will mostly likely end this summer or fall. Afterward, a new bull market will begin that should raise P/R some 0.3-0.4 units over roughly a 1-3 year period. This would translate into about a 50% rise in the index from current levels should this be the bottom. We should then get a new bear market that should carry P/R to even lower levels than we saw in the first bear market.
A caveat to this analysis should be noted. Whereas the drop in P/R of 40% is consistent with the three previous non-Depression secular bear markets (average 38% drop) the absolute level of P/R at 0.9 is higher than the value at the bottom of the first bear market for these same secular bear markets (average 0.76). That is, if the relative decline in P/R is what repeats, we could be at the bottom of the first bear market now, whereas if the absolute level is what repeats, then a lower level of P/R (consistent with an S&P500 in the low 800's) would be more consistent with history.
One of the hallmarks of the secular bear market period is a shorter ordinary bull/bear market cycle corresponding to a shorter business cycle. In fact, as was shown in Stock Cycles, is it the shortening of time horizons that a shorter cycle causes which helps produce the decline in market valuations associated with a secular bear market. With a shorter time horizon over which to discount growing earnings streams, the market will assign lower valuations than it did during the secular bull market.
Although a qualitative description is useful, what we would like to know is what are the prospects for investments made today, at today's market levels, relative to reasonably "safe" alternatives. In my first book, Stock Cycles: Why stocks won't beat money markets for 20 years, I examined this question for the early 2000 period, which was when I was writing. I showed that for markets that have a "high" level of P/R, which had been the case since 1998 (and would remain the case for most of 2000) investments made in an S&P500 index fund would be unlikely to beat the "safe" (money market) alternative over the next two decades, hence my provocative subtitle. In the following paragraphs I will apply a similar analysis to the situation now (June 2002). Rather than focus on the very long term (20 years), which is still pretty dismal, we will look at the intermediate term (5 years). As I describe in Stock Cycles, for the short term (~1 year) anything can happen and statistical forecasts such as those I produce have no utility.
Figure 6 shows projected annualized nominal returns over a five year period for three situations. The interpretation of these figures is straightforward. On the vertical axis is projected nominal total return on an annualized basis over a five year period. On the horizontal axis is the probability that such a return or higher will be achieved. The lines represent the different situations. Consider the blue line. This line presents the historical future returns for all months in which P/R was greater than 1.0 and for which P/R had risen at least 0.2 over the previous two years. That is, the blue line reflects the ends of secular bull markets (when P/R has been rising to the very high levels normally only seen then). This is essentially the same set of conditions examined in Stock Cycles. For the last bull market it corresponds to the period between June 1997 and October 2000. The blue line "predicts" that investments made during this sort of a period will show a median return of only 3% over the next five years. Only 20% of such investments will beat a 5% "money market" return [3]. Today, five years have elapsed from the beginning of the most recent such period in June 1997 and so we can look at how this prediction is turning out so far. As of today (June 26 2000) an S&P 500 index fund has produced a small return of only about 2.3% from June 1997, compared to 4.3% for a money market. Thus, the prediction made in Figure 6 has held up quite well with what actually happened (at least so far). Of course, some 20% of investments made over the June 1997-October 2000 period are projected to do better than a 5% annualized return. One can imagine investments made around the October 1998 lows, for example, will turn out to be in this group.
Figure 6. Projected nominal return as a function of probability for three different situations.
The next line to consider is the red one. This line refers to months in which P/R was less than 0.5 and for which P/R had fallen 0.2 units over the previous two years. Investments made during these times show an excellent median return of 15%. The probability of beating the 5% safe return is 96%. One might think, why not wait for such a period to materialize before risking your money during the treacherous secular bear market era? Today's P/R is around 0.9, were the S&P500 to fall below 600 in the next year or so we would enter such an era and it would be about as safe to invest as it can be. Many bears await just such an outcome. However, such particularly fertile periods for investing are quite rare. Furthermore, they tend to occur during the latter portions of secular bear markets, and so are not expected to be seen during this decade. For example, during the last secular bear this period first materialized in mid-1974, some 8½ years after the P/R peak in January 1966. In the early 20th century secular bear, it did not occur until 1917, 16 years after the P/R peak, and it never happened at all in the late 19th century secular bear era. But in the Great Depression bear, it occurred early, in November 1931. It is for this reason that many bears cite the Depression example as reason to wait for lower prices. But as we saw earlier, there are a lot of reasons to expect the current secular bear market to not develop like the post-1929 debacle. And if this turns out to be the case, then the investment nirvana shown by the red line is not going to appear during the current bear market (but rather a later bear market further along in the secular bear market era). Given this, can we use history to say anything about the probability of various returns of the next five years from investments in the S&P500 made today? The answer is shown by the black line. This line corresponds to months for which P/R was above 0.5, but for which P/R had fallen 0.2 units over the previous two years. Late March and April 2001, and the period since August 2001 fall into this category. The median return for investment made during times like this has been 8%. Fully 80% of the time the investment beats the 5% safe return. What does this mean? It can be interpreted in a number of ways. Perhaps the simplest way is to view the 20% chance of below-5% return as reflecting the probability that the current bear market will extend much longer, making current investments underperform money markets for the next five years. On the other hand, it could also reflect the probability of a short or weak bull market even if one does begin soon. There is really no easy way to interpret the story told by the black line. The best way to look at it is to consider investing today to have a 20% risk of failure with an "expected" return of 8%. When we compare today to the June 1997-October 2000 period which showed an 80% chance of failure with an expected return of 3%, it is hard to see why anyone wished to invest back then and why people are so afraid today. It is this irrational fear/greed dynamic that causes the extremes of the stock cycle in the first place and which provides the opportunity for exploitation described by this article, so should we really question it?
References:
1. Before 1987 0.49% was subtracted from the S&P500 index total return. After 1987 the actual returns of the Vanguard index 500 fund (VFINX) were employed. Hypothetical money market returns were estimated by subtracting 0.1% from 90 day t-bills after 1931 and 0.7% from commercial paper rates over the 1871-1931 period.
2. Greenspan Alan, remarks at the Annual Dinner and Francis Boyer Lecture of The American Enterprise Institute for Public Policy Research, Washington D.C., December 5, 1996, www.federalreserve.gov/boarddocs/speeches/1996.
3. A 5% return is assumed for a "safe" return despite the low money market rates of today for several reasons. First, many 401K plans (like mine) have an "income fund" (a guaranteed investment contract or GIC) that works like a money market fund but pays a higher return. My plan's GIC income fund is still paying about 5.5% today. Also, a mixture of bond funds of different types can generally produce a 5% return . Because of the existence of these reasonably "safe" options I have held to a 5% standard for the "riskless return" in this article. When I wrote Stock Cycles regular money markets were paying about the same as my GIC income fund and so I used the term "money markets" to refer to a "safe" return.