back start next

[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [ 45 ] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72]


Duration (Days) 92 107 158

242 342


Probabiiity of Trend Reversal Very Low

Zr)d QuartWe

3rci QuartWe

4th QuartWe

kXI Moderate to Very Hih

Figure 11.1 Frequency distribution for extent and duration in buli market primary swings on the Dow Industrials since t896.

moves failed below 15%. Based on these criteria alone, you would be wise to bet that a primary swing would appreciate somewhere between 15% and 30%. But of course, you would never use these criteria alone.

Think about how life insurance companies go about gathering their information. The first question they ask is age. Then they factor in job hazards, medical history, family history, and so forth. But the mortality tables are the standard of reference, the starting point of evaluating the risk of insuring the customer. In the same way, I use my distributions as a means to establish the base probability that a market movement will reach or go beyond a given extent and duration.! absolutely do not use them to predict the exact levels the current market movement will reach or how long it will last! Market tuming points occur when the tide of market participants judgments changes ... period. This is usually driven by fundamental economic factors, the policies of the Federal Reserve Board, major world events, and so forth, as I discussed in the last chapter.

To use extent and duration profiles to predict in advance exact market tuming points would be like having an insurance company tell you when and how you will die on the day you buy your policy. I dont mean to be morbid, but just as each individual person dies in a different manner, context, and point in time, so markets die. But who are you most likely to see on stage next year, Meryl Streep or George Bums? What these profiles do tell you is the base probability, in the context of history, that the

current stock market trend will continue or fail.

Returning to the insurance analogy, if two men, one 18 years old and one 75 years old (the median age that American men die), both in good health, went into an insurance company to apply for a term life policy, then the younger man would pay a very low premium and the older man would pay a very high premium. The premiums would be set such that the risk, according to the statistics, of the person dying before he pays the entire policy value plus interest is very low. But if the older man had a temperature of 102 and were a heavy smoker, he wouldnt be sold a policy at all.

Consider the examples of the October 1987 crash and the October 1989 minicrash in a similar context. The primary intermediate movement leading to the crash of 1987 began on 5/20/87. By 8/25/87, the Industrials had increased 22.9% in 96 days, while the Transports had increased 21.3% in 108 days. The se were nearly the exact median levels, both in extent and duration, of all bull market intermediate movements in history. In life insurance terms, the market had reached the median life expectancy, meaning that 50% of all intermediate movements in history ended before this one-it was a likely candidate for retirement. From this standpoint alone, caution was warranted, so it was time to examine the medical history.

The market was no Jack LaLanne. There were divergences; the Dow made a new high in August, bu t the advance/decline ratio did not-a bearish indication. PEs were at an average of 21 times eamings, the highest levels since 1962, when they were at 22. The average book value to price ratio was nominally higher than 1929. Govemment, corporate, and consumer debt were at unprecedented levels, and all the rest that Ive discussed before. In October 1987, the market was not

only no Jack LaLanne, it was an alcoholic, with pneumonia, that smoked three packs of unfiltered Camels a day.

Consequently, I was out of the market and looking for an opportunity to short it. The first sign was on October when I read in the Wall Street Joumal: "Fed Chairman Greenspan said interest rates could become "dangerously high if inflation worries "mushroom in financial markets. Greenspan called such worries unwarranted but hinted the discount rate may have to rise to allay them." The next day, stock prices plunged a record 91.55 points for no immediately apparent reason other than Greenspans pronouncement. On October 15, Dow Theory gave me a sell signal, and I went short thinking that the patients heart could fail with even the slightest excitement.

The heart attack occurred when Germany and Japan failed to heed James Bakers request to stimulate their economies (inflate) to protect the value of the dollar. In response. Baker announced to the world on Sunday, October 18, that he "would let the dollar slide." I knew at this point that the financial markets would collapse from the dollar devaluation. When the market gapped down on October 19,1 shorted the opening of the S&P 500 futures and made a substantial profit for my account in that position alone. The bittersweet part of it is that I had independent traders working for me at the time (free to make their own decisions) who were caught long, and some of my profits were offset by their losses.

The point in describing my thinking before the crash of October 1987 is to show how watching extent and duration criteria can clue you in to the potential health of the marketthey act as a caveat telling you to pay close attention to what is happening fundamentally. Sometimes they tell you more than that. For example, I was essentially out of the market from May until October of 1989; partly because I was busy organizing my new money management firm, but also because I feared the downside possibilities. By 10/9/89 the net appreciation primary intermediate up move on the Dow, which began on 3/23/89, was 24.4%. During the same period, the Transports had moved 52%, fue led by takeover stocks. Let me put this into the framework of my statistical base.

The median net appreciation, or yield (the net percentage move if you buy and hold from the bottom of one intermediate correction to the next), for primary intermediate move ments (primary swings) in bull markets is 10%. Further, only 15 of 112 such moves in history yielded more than 24.4% (the gain that was current in the existing primary swing as of October 9). Even more important, only 8 of 174 upward movements in both bull and bear markets had matched or exceeded the Transports 52% gain. In other words, the odds, based on history, that the market would fail ranged from 7.46 to 21.75 ! That told me to close any longs and look for a shorting opportunity. Measured by stati stical criteria alone, to be long in October 1989 was one of the poorest risks on record. And once again, the fundamental and

technical factors were on the side of the statistics.

The new high set by the Industrials on October 9 was unconfirmed by the Tran sportsa bearish indication. As of Friday, October 13, the Japanese and the Germans had raised interest rates, making it more expensive for their companies

and citizens to buy American products on credit. The Fed had reduced the money supply and cred it availability (as measured by free reserves) over the previous two reporting periods, and with the CPI already at an annual rate of approximately 5.5%, they did not have much room to ease in order to stimulate business activity, which was already sluggis h. Northeastern real estate and the regional banks holding the mortgages were already in recession. The bottom had already fallen out of the junk bond market. In short, the patient was once again old and sick and another stroke was imminent-not a good time to write an insurance policy at any premium rate. It was certainly no time to be long. I was fortunate to be short at the right time and in the right instruments, but it was my statistical profiles that clued me in to the downside possibilities.

Anybody who knows me knows that I dont mind playing the short side. But I want to emphasize that this approach to risk assessment is equally valuable to pension fund managers, who must be long to some extent in the stock market. Suppose the fund manager is trying to determine what percentage of capital to allocate to the stock market, with 60% being an aggressive position and 20% being the minimum allowable position according to the investment charter of the fund. If the stock market is in the third primary swing of a bull market, then the statistics say that the median extent and duration for third primary swings are 139 days and 18.8% respectively. If the market had appreciated 12% in 103 days, the manager would know from the profiles that 16 of the 23 third primary movements in bull markets had lasted longer. On that basis alone, the manager should commit no more than 69% of the maximum 60% to stocks. In other words, on the basis of the statistics alone, they should commit a maximum of 41.4% of the portfolio to stocks.

I want to emphasize again that this is only a starting point, a basis from which to evaluate the risk of market involvement. If under these same statistical circumstances, inflation was at 1%, PEs were at nine, interest rates were at 5°k, and eamings were soaring, then the manager would underweight the statistics and invest the full 60% in stocks.

This concept of risk measurement has been absolutely instmmental to my success in the markets, particularly in positions involving either long-term or intermediate-term market tuming points. It allows me to place my primary focus on risk, which is the basis of my approach to speculation. But no matter how you make market calls or stock selections, market life expectancy profiles provide an objective base context within which to gauge the risk of market involvement. They add a unique and powerful new dimension to risk assessment.


As mentioned in Chapter 2, my primary focus is to minimize risk, while simultaneously putting enough capital in the tight place to make consistent profits. What it takes to do this is a pmdent system of money management.

Money management is the art of allocating financial resources and timing entry and exit to and from the marketplace to achieve business goals. A solid approach to money management must consist of three cmcial components: (I) a method of assessing tisVeward, (2) a means with which to determine the probability of success on any given trade (whether short-term, intermediate-term or long-term), and (3) a system of asset allocation. And, since you are dealing with money, these three components must, at least predominantly, be reduced to objective, measurable criteria.

I have already discussed my statistical profiles and how I use them to measure risk. Also, if you recall, I pointed out that my risVreward criterion is at most 1 to 3, and I demonstrated how I determine that ratio technically on the charts. Basically, it amounts to looking at the chart, evaluating where you think the market can probably go-the target point, establishing the point where the market proves that

[start] [1] [2] [3] [4] [5] [6] [7] [8] [9] [10] [11] [12] [13] [14] [15] [16] [17] [18] [19] [20] [21] [22] [23] [24] [25] [26] [27] [28] [29] [30] [31] [32] [33] [34] [35] [36] [37] [38] [39] [40] [41] [42] [43] [44] [ 45 ] [46] [47] [48] [49] [50] [51] [52] [53] [54] [55] [56] [57] [58] [59] [60] [61] [62] [63] [64] [65] [66] [67] [68] [69] [70] [71] [72]