Progress of the secular bear market: position as of July 31, 2006

The value for R is 1380 as of July 2006. For S&P500 of about 1280 this gives P/R of 0.91

Stock Cycles: Part III. Comparison of P/R with P/E and Q

So far we have seen that the stock market moves in long multi-decade cycles. The performance of long-term buy-and-hold stock investments is heavily dependent on when such investments are made with respect to this cycle. The position in the cycle is determined using a long-term valuation method called price to resources or P/R, which was described in last month's article.¹ The value of P/R relative to its average value within a cycle was shown to correlate well with historical periods of low and high valuation. Back in 2000 this relative P/R was used to forecast stock returns over a 5-20 year period going forward. It was concluded that stocks bought in 2000 would probably underperform money market funds for the subsequent 20 years.

P/R is not the only valuation tool that can be used to show the stock cycle. Because the stock cycle is the phenomenon responsible for the predicted poor stock returns after 2000 (what is called a secular bear market) other better known valuation tools could have been used to make the same long-term forecast in 2000. In this installment P/R is compared to two of these methods for what they have to say about the recent stock cycle peak and its aftermath. Each method was featured in an important book with a bearish theme that came out in the same year (2000) as did my book Stock Cycles.

The first method is an averaged price to earnings ratio described by Robert Shiller in his book Irrational Exuberance. The S&P500 index value is divided by the average index earnings over the previous 10 years. The price to earnings ratio (P/E) is probably the most widely used stock valuation measure. By averaging earnings over ten years, Shiller smoothes out short-term fluctuations, revealing the long trend trends that define the stock cycle.

The second method examined is the Q ratio originally developed by the late Yale economics professor James Tobin. The Q ratio measures the ratio of the market value of factories and other corporate assets to their replacement cost. When Q is low, as it was in the late 1970's and early 1980's, companies tend to expand by acquiring other companies instead of building plants or buying equipment. When Q is high it makes better sense to build new assets directly. Obviously, if one can buy an asset more cheaply by buying the stock of the asset-owning company than buying the asset directly, the stock is undervalued. This is the basis of the use of Q as a measure of stock valuation. When Q rises much above one, stocks are overvalued.

Smithers and Wright featured Q in their bearish book, Valuing Wall Street. They extended the record of Q back to 1900, covering three previous secular bull market peaks. They found the value of Q at the 1960's peak (1.06) was the lowest of the three, with the highest (1.35) occurring in 1929. Thus, markets can rise significantly above one at major market peaks, but extreme valuation is indicated by Q values approaching the 1.35 value in 1929.

The behavior of different valuation methods in the most recent secular bull market

The purpose of these valuation tools with respect to investment planning is to give an idea of where we are in the Stock Cycle. During a secular bull market, the goal is to remain fully invested until near the end of secular bull market, at which time one shifts to an alternate investment strategy. One does not wish to shift too soon, as the returns expected from alternate investments are likely to be much lower than those from stocks in a secular bull market.

Historical values for Shiller's P/E are available from 1881 on. Values between 17 and 32 have been seen at previous secular bull market peaks. Shiller's P/E rose above its previous all-time high of 32 in January 1997. Indeed, shortly following Shiller's presentation of his P/E model to the Federal Open Markets Committee in December 1996, Chairman Alan Greenspan gave his famous "irrational exuberance" speech, in which he cautioned investors about high stock valuations. Followers of Shiller's P/E valuation would have heeded the Chairman's warning and made preparations for a secular bear market at the end of 1996. Q reached its 1929 level in the third quarter of 1997, indicating extreme overvaluation. Followers of this measure would likely have left the market shortly after the Greenspan's irrational exuberance speech.

In contrast to P/E and Q, P/R did not reach its all-time high until January 1999. Thus, a follower of P/R could have remained fully invested for an additional 1-2 years of bull market. But there was yet another advantage. In late summer 1997, Congress passed a capital gains tax cut. This act was intended to encourage investment -- specifically the kind of investment that produces capital gains. That is, this act was intended to make stocks go up in price. On two prior occasions, (1921 and 1981) when capital gains taxes had been lowered to 20% or lower, a most satisfactory stock bull market had ensued afterward. So it seems that this sort of policy works. Thus, there was reason to believe that with the boost produced by the capital gains tax cut, the bull market should end at all-time high valuation levels or beyond. Hence I only started reducing my stock allocation below 100% in late 1998 and when record P/R values were reached in January 1999, I was still 50% invested in stocks. I did not completely exit stocks in my 401K until September 3, 1999 after P/R had reached a new all-time in a July 1999 market peak, that I had interpreted as the end of the secular bull market.

Of the three valuation tools, P/R reached record levels the latest, two years after Shiller's P/E and a year and a half after Tobin's Q. Significant stock gains occurred over these periods, which followers of P/R would get. But this may be merely a fluke. In order for P/R to be considered an superior valuation measure, it is necessary for it to give a proper signal for re-entry. In the next section, the three measures will be used to provide an assessment of the market's prospects as of its summer 2002 levels.

Figure 1 Shiller's P/E during secular bear markets

The behavior of different valuation methods in past secular bear markets

Figure 1 shows a graph of Shiller's P/E for the present secular bear versus that for previous secular bear markets: 1881-96, 1906-21, 1929-49 and 1966-82. Several things are evident. The value of P/E at the beginning of the current secular bear market was much higher than that at the beginning of the previous bear markets. This reflects the three additional years of bull market advance that occurred after P/E reached its previous all-time high. Figure 1 shows, in terms of P/E, how out-of-line the recent market performance in the late 1990's (as measured by P/E) was compared to all previous periods, including the 1929 peak. Either the market performance in the bull was grossly abnormal or the P/E measure isn't relevant anymore. In the first case, the market would have to come down a lot in order for the market to look "normal" in terms of P/E. In the second case, we should disregard the message apparently being sent by P/E.

Figure 1 shows that Shiller's P/E fell dramatically since the 2000 peak, but as of March 2003, it was still well above the levels it has seen at this stage in previous secular bear markets. The S&P500 would have had to fall into the 400's for P/E to be within the historical range. Thus, if P/E valuation was still valid, the market needed to fall another 40% from the fall 2002 lows.

Figure 2 Tobins's Q during secular bear markets

Figure 2 shows a graph of Tobin's Q for the current secular bear market and three previous ones. As with P/E the level of Q reached in 2000 was way above any of its previous levels, but not quite as extreme as with P/E. This reflects its reaching a value equivalent to its previous all-time high 2.5 years before the actual market peak. We can interpret the 2000 peak as either the largest bubble in history (when viewed through a Q lens), requiring a correspondingly large correction to deflate, or we can question the precision of Q valuation.

Q also fell dramatically after 2000. At the bear market bottom in October 2002, Q was still high relative to its historical values in early stage secular bear markets. The S&P500 would have had to fall to the 600's for Q to fall into its historical range in previous secular bear markets. Thus, advocates of Tobin's Q would have cautioned against re-entry at the fall 2002 lows.

Figure 3 shows a graph of P/R for the current secular bear market and the four previous ones. Like P/E and Q, P/R in 2000 was higher than it had ever been. But the margin of excess was less, since P/R reached its previous all-time high only 14 months before the ultimate peak. We can interpret the high value of P/R in 2000 as a stock bubble extended by the capital gains tax cut or as a flaw in P/R valuation. P/R has fell like the other two measures after 2000. Unlike the other two measures, P/R in July 2002, October 2002 or March 2003 was not particularly high relative it its historical values in previous secular bear markets. By mid-summer 2002, P/R had fallen below the levels reached during the first bear market of three of the four previous secular bear markets. Lower levels than those in 2002-2003 were obtained in the subsequent bear markets, and it is expected that P/R will decline well below these P/R lows in future bear markets during this secular bear period.

What this meant was, according to P/R, by summer 2002, the market had fallen enough to be consistent with past behavior. That is any bubble present in 2000 had been deflated to the normal extent for this early in a secular bear market. Thus, although the index certainly could have continued to fall and P/R would remain within its historical norms, it did not have too. Followers of P/R would not caution against re-entry at the lows in summer and fall 2002 or late winter 2003.

Figure 3. P/R during secular bear markets

This concept is the basis of the charts shown at the front of each Stock Cycles article. The graph displays how "high" current valuations are compared to the corresponding periods in previous secular bear markets.

A statistical assessment of future returns can be made much as was done in Part I. In September 2001 P/R reached 0.97 and by October 2002 it was 0.77. In 2003, I made a future return projection by calculating future five-year capital gain returns and P/R for all months between 1871 and 1997. Those results with P/R between 0.77 and 0.97 were removed and ranked according to return and the returns at the various percentile levels noted. An estimated dividend yield of 1.8% was added to these capital gain data to give total return. These values would correspond to S&P500 purchases made in Sept 2001 or June 2002-June 2003.

The median total return projected was 5.9%. A 70% probability of beating the money market funds was projected. Average return since 2002-2003 (so far) has been about 8% annualized, considerable more than the money market returns over the same period. In October 2002 I wrote the following:²

The current market provides a good test of the three valuation methods. Suppose the July 23 low ends up being THE bottom? (Note: since this post was written the S&P500 set a new low on October 9, 2.5% below the July 23 low.) This would invalidate Shiller's P/E and probably also Tobin's Q as useful valuation tools. Only P/R would have given a valid reading of valuation.

On the other hand, suppose the S&P500 drops to an ultimate bottom below 550 without a deflationary depression. In this case, P/R valuation would be invalidated. Such a low level of P/R this early in a secular bear market has only occurred once, during the early 1930's. It is inconceivable that it happen again without an accompanying depression. Shiller's P/E says such a development is to be expected, no depression is needed. And Tobin's Q does not rule this out, only P/R does.

Suppose a subsequent decline takes the index to an ultimate bottom well below the July 23 close, but above 600? In this case, P/R would not be invalidated, but it would be Tobin's Q that did the best job. Shiller's P/E would still be invalidated. Finally a drop below 400 without a depression would probably invalidate Q as a useful measure and move Shiller's P/E into top place as best valuation model of the three.

So here we have a real-life test of three valuation methods. We should obtain an answer within a year--two at most, I should think.

The results are in. The October 9 2002 low mentioned was the bear market bottom. July 23 was within 3% of the ultimate bottom and an excellent buying opportunity as P/R valuation implied. The S&P500 did not fall well below the July 23 2002 level, and did not venture remotely near 550, eliminating P/E and Q as useful valuation tools. This test establishes why I prefer the P/R tool to the others.

In next month's article, P/R is analyzed in light of the stock cycle to obtain an understanding of why the cycle has come to be so important to stock market returns.

References:
¹ Alexander, Michael A., "Stock Cycles-July 2006" Safehaven, July 15, 2006.
² Alexander, Michael A., "How low can we go?" Safehaven, October 20, 2002.