FIGURE 35.1 Sector Rotation in the Cycle
On the other hand, late-cycle leaders are still in the terminal stages of a bull trend and are helping to push the averages higher. If strength in the earnings-driven sectors outweighs weakness elsewhere, the averages move to new bull market highs. This is the principal reason why breadth divergences are so prevalent at market peaks. In 1973, the commodity boom had the effect of prolonging the bull market, as far as the averages were concerned, but the NYSE A/D line had peaked 9 months earlier. In 2000, it was the unprecedented strength in technology stocks, another lagging sector, that propelled the averages higher, but under the surface, the average stock peaked 2 years earlier in 1998.
What Is a Peak?
The objective of this chapter is to offer a checklist of characteristics present at a typical market top or bottom. In reality, there are no “typical” turning points because no two situations are exactly the same. However, there are enough characteristics to enable us to identify the intricate tapestry of a market top or bottom.
It’s possible to categorize three types of market peaks. The most important are those that develop after a prolonged bull market, extending over several business cycles. These are the secular turning points described in Chapter 23. Around the time of the peak, it is possible to point to some kind of speculative bubble, often concentrated in a few sectors. At such times old rules are thrown out and virtually everyone becomes an overnight investment genius. Such peaks are very hard to pinpoint at the time because they have defied virtually every traditional rule. Eventually, new rules or rationalizations are developed to justify the final ascent to the equity summit. In this way traditional valuations are surpassed by several degrees. Normal extreme technical benchmarks are also exceeded, and divergences are so plentiful that they are ignored because they no longer work. Finally, contrarian positions based on observation of normal crowd behavior prove futile, as the majority march on to new extremes of irrationality. Old rules are thrust aside and ridiculed as new-era thinking predominates. The rallying cry of the bulls is almost always inculcated in the mantra that “this time it is different.” In all cases, the leveling factor that brought investors back to reality was a good dose of rising short-term interest rates. Three such peaks possessing at least some of these characteristics developed in 1929, 1966, and 2000. The 1990 secular peak in Japanese equities also qualifies. In 1929, stocks lost close to 90 percent of their value in 3 years. The bear market that followed the 1966 peak lasted a lot longer and consisted of several mini-bull and mini-bear markets. When stocks are adjusted for inflation, the bottom, as measured by the major averages, did not take place until 16 years later in 1982. The top established in 2000 is likely to be of similar importance to the 1929 and 1966 experiences, since many of the valuation, sentiment, and technical characteristics cited in Chapter 23 were present.
There are really two factors that cause the ensuing bear market to be so devastating or lengthy. The first is a direct result of the long and persistent bull trend and falls under the title “careless investment decisions.” It is the function of the bear to cleanse the system of these poorly thought-out and precariously financed positions. Such is true of every market peak. It is just far more prevalent at these super-cycle turning points. Second, it appears that the pendulum of human emotions, the ultimate arbitrator of prices, moves to an extreme of extremes. If such a level is reached on one side, an extreme of an extreme is likely to develop on the other. It almost appears as if a prerequisite for a huge bear market is an exceptionally large speculative bull market.
The second peak is what we might term a recession-associated top (RAT). It is the most common and best fits the primary trend description described in Chapter 1. In such situations, the unwinding of distortions in the economy brought about by the recovery is sufficient to trigger an actual recession where profits are under attack. In each cycle, the sector responsible for the distortion will be different. In the early 1970s, it tended to be real estate; in 1974, bloated inventories from the commodity boom were the primary culprit. In 1990 and 2007, the financial sector was under pressure, and so forth. Since these RAT-type bear markets are associated with economic contraction and recovery, which take time to accomplish, they usually last 1 to 2 years, are broadly based, and can be quite severe. The most severe of these bear markets develops under the context of a secular bear market.
The third kind of peak precedes so-called growth recessions or double-cycle peaks as described in Chapter 2. They are usually followed by shallower and shorter bear trends. Growth recessions involve the slowing down of the economy but not an actual contraction. In this process, several industrial sectors will experience recessionary conditions, but strength elsewhere will offset this weak activity so that the overall economy escapes the indignity of the “R” word. These weak sectors, therefore, experience a true 1- to 2-year bear market, whereas others merely experience a major trading range. Growth recessions are more likely to develop during a secular bull market.
Examples of double-cycle bears are 1984 and 1994. Other bear trends are associated with a slowing in some of the economic indicators and are preceded by an unhealthy level of speculation. They take the form of severe technical corrections, but are so dramatic that the change in psychology is sufficient to correct much of the speculative juice. The 1962, 1987, and 1998 declines come to mind as good examples.
All of the characteristics listed in the following section do not appear at every market peak, and neither is their intensity of the same magnitude. They are presented solely as a guide for things to watch out for at major highs.
Characteristics of Primary Market Peaks
1. In order for a bull market top to form, it must be preceded by a bull market. Thus, it must be possible to look back and identify a rally of at least 9 months duration.
Monetary and Sector Rotation Factors
2. Almost all market peaks are preceded by a trend in rising short-term interest rates. The lead times vary from a couple of months to years. If interest rates have not begun their ascent, the odds of a top are greatly reduced. For examples, refer to Charts 31.1a and b in Chapter 31.
3. Watch out for hikes in the discount rate. Has the “three step and stumble” rule, discussed in Chapter 31, been triggered? If so, the distribution process may well be under way, and the reward of owning stocks is usually outweighed by the risk.
4. In many instances, market tops are preceded by a peak in the Dow Jones Utility Average. This is because this is a very interest-rate-sensitive market sector.
5. Observation of the long-term relative momentum for a selection of early- and late-cycle leaders often points up a bull market peak. This will be true if the long-term smoothed relative momentum for financials, utilities, and consumer nondurables (staples) peaked out some time ago. By the same token, a characteristic of a market peak is either continued strength in late-cycle leaders, such as basic industry and resource stocks, or an actual peaking of their smoothed momentum.
Technical Factors
6. When interest rates start to rise, this adversely affects financials and preferred shares. Consequently, it is normal for the NYSE A/D line to peak out ahead of the Dow Jones Industrial Average (DJIA) and S&P Composite. The length of the divergence often has a related effect on the size of the ensuing bear market.
7. If the NYSE daily A/D line and or the Value Line Arithmetic Index are below their 200-day moving averages (MAs), this indicates that the broad market is probably in a primary downtrend. Even if the DJIA and S&P Composite are at new highs, this weak breadth position is telling us that the market has become very selective. Under such an environment, it is more difficult to find stocks that are advancing. Cutting back on equity exposure, therefore, makes excellent sense because your odds of success are less. The opposite would be true in a declining market, where the A/D line is advancing.
8. An alternative way of looking at the technical position of market breadth is net new high data. Has the numbe
r of net new highs diverged negatively with the major averages? If it has not, this may not be a top. If it has, then fewer stocks are driving the market higher, and this is a sign of weakness. If the number of new highs has shrunk considerably, it is very difficult to make money because the selection process is becoming progressively more difficult.
9. Momentum typically leads price, especially at market tops. Occasionally, it is possible to identify multiyear tops or trendline violations in several long-term rate-of-change (ROC) indicators. For example, if you can spot trendline breaks in the 12-, 18-, and 24-month ROCs, this indicates that the cycles they reflect are all starting to turn. The more trend breaks from different time spans, the stronger the signal. Sometimes, an individual ROC moves between specific benchmarks, and these have traditionally offered good timing points for major tops or bottoms. Charts 21.5 and 21.6 offer some useful benchmarks for a 9-month ROC of the S&P Composite. Always make sure that any momentum signals are confirmed by some kind of trend-reversal signal by the price itself.
10. Another way in which momentum can signal a maturing bull trend is with the use of a long-term smoothed oscillator such as the KST or by an identifiable peak in the Special K. In the case of the KST, market tops are typically signaled by a reversal to the downside from an overbought condition. This is represented in Chart 35.1 by the vertical lines, where the KST crosses below its 9-month MA.
CHART 35.1 S& Composite, 1952–2012, and a Long-Term KST
The chart shows that this approach works well when a normal cyclical bull market is experienced. However, during a secular bull market, such as the one that developed in the 1990s, this indicator gives premature sell signals. These are flagged with the dashed vertical lines—all the more reason to make sure that such signals are confirmed by a trend break in the price.
Alternatively, a primary-trend reversal may be signaled by short-term oscillators registering mega oversold or extreme swing conditions.
11. If the DJIA and the S&P Composite have just crossed below their 12-month MAs, there is a good chance that a bear market is under way, provided, of course, that other indicators are in agreement. Quite often, a negative 7-month MA crossover will work as well as a 12-month time span. Also, in the trend-reversal department, monitor the current position of the Industrials and Transports as called for in Dow theory.
Psychological Factors
12. If key companies report excellent earnings and the prices of their stocks decline, this adverse response to good news indicates extreme technical vulnerability. Whenever a stock or market index fails to respond to good news, this is a sign of technical weakness. Remember, if good news cannot send a security higher, what will?
13. On the sentiment side, a reading in the Investors Intelligence percentage of bears would be in the 10 to 20 percent range. Another sign might be the market’s failure to respond favorably to a prominent earnings report that was well ahead of expectations. Such signs often develop during the course of a primary trend, say at an intermediate turning point, so it’s their collective rather than their individual warnings that should be heeded. Finally, in the sentiment area, if the level of margin debt has recently crossed below its 12-month exponential moving average (EMA), this tells us two things. First, traders are losing confidence. Second, the market is no longer being supported by an expansion in margin debt, and worse, the implied contraction in margin exposure will put downward pressure on stock prices. Consequently, a 12-month downward EMA cross is usually an excellent long-term sell signal. The problem is that the data are published with a 2-month lag.
14. If the media are full of optimistic news and stories of huge returns are being publicized, this is indicative of a topping environment. Supplementary evidence in this direction might evolve from magazine cover stories on the market itself, especially ones citing a new paradigm, or concerning companies or groups that have been leaders in the bull market. In very strong trends, such as the 1990s tech bubble, such signs can be early. For example, an unprecedented amount of cover stories developed in the late summer of 1999, 6 months or so prior to the March 2000 final top, a period in which prices doubled. The good news is that a correct reading of this information told us that a major peak was at hand. The bad news, of course, was that what would normally have been a really strong signal worked after an unduly difficult delay. It is a stark reminder that some major market peaks are associated with a substantially less intensive peak in bullish sentiment than previous lower peaks. Consequently, it is not possible to conclude that just because a new high is not accompanied by a higher level of optimism that everything is okay. The whole situation regarding negative divergences and the action of other indicators must be considered.
15. When the market is rallying, there is always a substantial amount of doubt. However, as equities approach a peak, the question of whether the security in question is a good investment is not even debated. Instead, for the stock market, for example, the debate revolves around which sectors will do well, how high prices will go, or what the next levels of resistance are. In such an environment, forecasts that would previously have been ridiculed for their excessive optimism are given widespread publicity and credibility.
16. Since brokerage companies thrive on bull markets, they become very prosperous during these times. If you see reports of any of these companies moving into larger and more expensive office accommodations, this is often a sign that the uptrend is in a very mature stage. Alternatively, this could show up in the exchange or the back office of brokers being overwhelmed with backlogged orders.
Chronological and Cycle Factors
17. A quick review of the three markets in relation to their 12-month MAs will offer a quick reference for the current stage of the cycle. A market top should see the 3-month commercial paper yield above its 12-month MA. The “three steps and stumble” rule (see Chapter 31) will probably have been triggered as well. Bond yields in the corporate and governmental sectors should be in a similar position. The S&P Composite should be above its average, and the CRB Spot Raw Industrials should be above its average as well. In cases where stocks have fallen quite some way from their peak, it is possible that the S&P would already be below its 12-month MA. If the CRB Spot RM Index has also recently crossed below its average, then little support from commodity-based equities is likely, and at the tail end of the cycle, this usually means a bear market.
18. The 4-year stock cycle has been extremely reliable in offering a buying opportunity once every 4 years. It follows that if the current year is 2 or 3 years from the previous 4-year low, the odds favor a bull market peak. If some of the other signs discussed here are present, this will obviously increase the probability of a top.
19. Is it possible to observe three identifiable intermediate advances that have already taken place? If this is the top of the third, it may be a primary peak. This is by no means infallible, because some bull markets consist of two advances and some more than three. However, when combined with others, these signs can be quite a useful benchmark.
Primary Market Bottoms
Bear market bottoms develop under exactly the opposite conditions as tops. The news is bad, sentiment is extremely bearish, and the long-term momentum indicators are usually extremely oversold. Perhaps the key difference between a major low and a major high is that in almost all cases, bear markets are shorter than bull markets. We have to say “in almost all cases” because the 1929–1932 bear market lasted for 3 long years, whereas the average primary decline has mercifully been much shorter. Characteristics of a major low are as follows.
Monetary and Sector Rotation Factors
1. Bear market lows are invariably preceded by a peak in short-term interest rates. I say “short-term rates,” but in almost all cases, a peak in long-term rates also precedes an equity bottom. It is just that changes in the level of short-term rates have a much larger influence on equities than yields at the long end because they reflect liquidity conditions deriving from short-term business decisions. Bor
rowing costs at the short-end change on a week-by-week basis. Because long-term bonds lock in rates for a considerable period, their effect on the economy is more gradual. The interest rate lead time varies from cycle to cycle, but generally speaking, the longer the lead, the stronger the ensuing bull market. In 1966, for instance, interest rates and equity prices reversed almost simultaneously, whereas the lead was almost a year in 1920 and 1982. The 1967–1968 bull market was tame in comparison to the bull markets of the 1920s and 1980s. This is not the same thing as saying that every time the lead is a year or more, the market will experience a mega rally—merely that there is a rough tendency for that to be the case. The reason for this phenomenon lies in the fact that the longer the lead, the weaker the economy. That weak economy has its pluses since it encourages firms to lower their break-even levels, so when the economy rebounds, new revenues fall to the bottom line.
2. The industry group structure at market bottoms should reveal an improving trend of relative strength and relative smoothed long-term momentum favoring early-cycle leaders, such as utilities, most financials, and most consumer staples (nondurables). To that list you could add homebuilders, telecom, and brokerage stocks. There should also be a pattern of deteriorating relative strength and relative smooth momentum in some of the lagging groups, such as resources, basic industry, and technology. Quite often, when the long-term smoothed relative momentum of the S&P Financial Index bottoms, this is quite close to a market bottom. The weakness in earnings-driven stocks is used more as a confirmation.
Technical Analysis Explained Page 52