Tiger  Software

PO Box 9491
San Diego, CA 92169
(858)-273-5900

How to Order

 

An Introduction to Computerized
Technical Analysis



Forty Four Years of Computerized Buy and Sells Signals on The U.S. Stock Markets

by William Schmidt, Ph.D. (Columbia University)

The stock market CAN be predicted much more than most ever imagine. This article will introduce you to concepts used by my Peerless Stock Market Timing Software. These concepts have proven their predictive value over and over. They let us pinpoint the market tops in October 1987, October 1989 and July 1990.

Real-time warnings were then issued to subscribers in our newsletter, PEERLESS FORECASTS, and in our Nightly Hotline. Equally important, the PEERLESS system called the beginnings of major bull markets in August 1982, July 1984, January 1988 and January 1991.

Since its first publication in 1981, thousands of professional traders have read my PEERLESS STOCK MARKET TIMING book and run our Peerless Stock Market Timing Software and our Tiger Software's  Power-Stock-Ranker Software. This introduction should give you a basic understanding of many of the key concepts we use to predict stock market’s swings. These are the insights distilled from nearly 30 years of personal trading experience. Each has been verified by computer analysis of as much as 30 years of back data. Their real-time use by hundreds of professional traders has further refined them.

The first thing I want you to do is to greatly reduce the weight you might otherwise attach to the day’s national news - economic, political or financial. We are only interested in the market’s reaction to such news, not the news itself. In that sense we are contrarians. When the market refuses to go down on bad news, it will probably soon make an advance. When it refuses to go up on good news, it will probably next swing downwards.

We are market technicians. We believe the market can best be predicted by computerized analysis of its own statistics. These statistics include the daily highs, lows and closes of popular indexes or stocks, as well as their trading volume. We also break down volume by whether or not the stock rose or fell for the day, and with the New York Stock Exchange we closely examine statistics derived from the number of shares trading in advancing stocks and declining stocks. Using our computers we study market history to find patterns in the data which tend regularly to precede a market advance or decline. The patterns we test must make intuitive sense. Most often, they stem from the standard literature of technical analysis.

In the final analysis, we are historians and odds-makers. Absolute certainty is not possible. Rather, we speak in terms of a particular pattern’s past track record. For each statistical pattern or PEERLESS automatic signal, we offer data showing the number of such cases and the resulting gain or loss in, for example, the Dow Jones Industrial Average (DJIA) or the Standard & Poor’s 100 Index (OEX) over a specified number of days or at the time of the next reversing automatic PEERLESS signal. This then permits us to speak in terms of a signal’s average gain and its probability of success.

As technicians, we must first decide the general market’s trend. This has more influence than anything else on how a stock is apt to perform. Next we must also determine whether the current stock market is in a stable trading range or if it is strongly trending up or down. In a trading range, profits are generally made by buying on weakness at the “support” created by earlier price lows and selling on strength at the “resistance” created by earlier price highs. However, in a strongly uptrending market, we do best by buying and letting profits run and by adding to positions when a stock successfully gets past its earlier price highs. The trick here is to determine the trend, if any, as soon as possible.

The sooner we see that a market is changing its character, shifting, say, from a trading range to an uptrending market, the quicker we can profitably shift away from trading tactics and employ more a buy-and-hold approach. To recognize the general market trend and to ascertain the strength of the trend the PEERLESS automatic trading system has developed a number of time-tested major signals which denote a strong uptrend.

The major PEERLESS buy signals consistently get very sizable general market gains at the time of the next major reversing sell signal. Our PEERLESS system’s 5 major automatic buy signals show the following track record since 1969 using the Dow Jones Industrial Average (DJIA).

PEERLESS MAJOR BUY SIGNALS
TRACK RECORD: 1969-1995

No.of Cases   Major Buy Signal     Average DJI Gain at Time
                                                      of Next Major Sell Signal
-------------  ------------------     --------------------------

25--------------- B4 ------------------ 14.8%
113--------------B9 ------------------- 9.6%
14 -------------- B10 ----------------- 14.2%
14 -------------- B12 ----------------- 24.6%
18 -------------- B15 ------------------ 7.1%

I am writing this when the most recent major signal was a BUY "B15". This occurred on March 27, 1995 with the DJI at 4157. The signal historically averages a gain of 7.1%. Adding this to 4157, we get an upside target of 4452 - assuming this B15 is typical. (In actual fact, the DJI moved up to 4872 before a major sell signal warned of a coming correction.).

The B15 signal occurs because the DJI has shown significant upward momentum. Specifically, it occurs because the DJI’s 21-day moving average has been rising for 40 straight days. The 21-day moving average is calculated by taking the sum of the last 21 days of closing price levels and dividing it by 21. As a new day’s data is added, the data for the day 21 days earlier is subtracted. We rely heavily on the 21-day period because that is about the number of trading days in a calendar month and it has been demonstrated that there is a monthly cycle in the market. Thus, a 21-day moving average is a simple, but effective, way of smoothing the data to show the true intermediate-term trend.

Click this icon to see the 1994 PEERLESS chart
of the DJI


The chart above shows the DJI for much of 1995. It also shows the PEERLESS automatic buy and sell signals, you will see that the rising 21-day moving average has not been violated for a very long time. In addition, the DJI is showing special momentum. Its uptrend can neatly be fitted by a parabolic or accelerating uptrend-line. And, the DJI and the other indices are at all-time highs, where the “path of least resistance” is up. The market is being driven by human emotions. No one knows where to sell and there lots of on-lookers who are eagerly waiting for any pull-back to buy. When the parabolic DJI’s uptrend-line and the 21-day moving average are finally violated, we would typically expect a severe bout of profit-taking. The market would then probably enter either a downtrend or a trading range depending on the PEERLESS automatic signals given.

Your personal computer CAN help you predict the stock market much more than you probably ever imagined. Your computer is a great equalizer when it comes to the stock market. You can easily do the same type of sophisticated analysis of market statistics and trading volume that used to be the reserve of big brokerage houses and mutual funds. Now you can see what the “insiders”` and “big boys” are doing, rather than what they are saying, and gain a new understanding of how mass emotions and market swings are regularly predicted by market professionals.

Let's continue our exploration of the most useful tenets of computerized technical analysis. To begin: when a market is not strongly uptrending or downtrending, the market tends to move in regular swings up and down. Technicians believe that such a market is sometimes “overbought” - ripe for profit-taking - and sometimes it is “oversold” - ripe for bargain hunting. In this they are observing the way the market moves in cycles of greed and fear. We have found that reliable overbought conditions may be defined with fixed-interval bands built around a 21-day moving average. Back in 1981 I introduced this concept to technicians world-wide. It quickly became popular and is now a standard feature in most technical analysis software packages.

Try this idea. Construct a 21-day moving average of the Dow Jones Industrial Average’s (DJIA) closing prices. Next have your computer build a band 3.5% above that moving average and then one 3.5% below it. We have found that for the period 1965-1995, the DJIA tends to stay within a 3.5% upper and lower band about 90% of the time, usually bouncing back and forth between the bands.

You will also discover that big market advances many times begin with an initial thrust up past the normal bands. Much more often, however, the DJIA or a typical individual stock or market index, stays within the normal bands. Is there a way to tell when the bands will stop a rally or act successfully reverse a decline? The answer lies in how accurately we can measure the market’s, or a stock’s, “internal strength”. Let me now introduce the theory of “internal strength” and then show you how we measure it. This will often provide a strong clue as to whether the upper band will successfully block a rally and the subsequent decline will be steep and disastrous or whether the upper band will be knifed right through and much higher prices will follow.

Internal strength has many aspects. Here are the most important, in my view. Technicians often consider a stock to be strong when it rises on high volume and declines on light volume. In this case, the stock is attracting big, deep-pocketed, new buyers as it advances while it tends to go down on just light volume, the result of normal profit-taking. By contrast, a stock which goes up on light volume may not be able to sustain its advance. It may be rising because the buyers are acting especially emotionally. Greed has overcome care and caution. At the same time, short-sellers may be hitting the panic button. Under duress, these players are not very careful in their placement of orders as they bid up prices. If the rally is heavily due to the covering of their positions by short-sellers, once the rally ends, the stock or the market may be even more vulnerable on the next decline.

Low volume rallies are suspect for another reason. Sometimes the rally is less durably taking place, as on a trading day right before a holiday, when a number of market players are absent and not actively participating.

In the general market, technicians often want a rally by the Dow Jones Industrial Average to be confirmed by increasing trading volume and “expanding breadth”. They consider an advance by the DJIA to new highs to be confirmed if the other popular averages, like the Standard & Poor’s 500 or the NASDAQ Over-The-Counter Index make new highs, too. They also want the number of stocks making new highs to be expanding. And, very importantly, they want the NYSE Advance-Decline Line to be making new highs that correspond to the DJI’s.

If a rally is mainly in high profile “blue chip” stocks, like those that make up the Dow Jones Industrial Average, the market breadth is considered “narrow” and untrustworthy. The NYSE A/D Line is built by computing the difference between the number of advancing NYSE stocks and declining stocks and then adding that value to the on-going sum computed from earlier Advances-Declines data. When the DJI makes a new high which is not followed quickly by the other averages, or the number of new highs, or the NYSE A/D Line also making a new high, technicians usually consider bearish “non-confirmations” to have occurred. At the upper band, these signs of weakness make the market or a stock particularly bearish. Closely related concepts are the basis for our major “SELL S9” which you can see in the chart of 1990 just below.

Oppositely when only the “blue chip” stocks are falling, the decline is narrow and false. The appearance of weakness is being maintained, but under the surface the market strength is building and a good rally is near. If the DJI is reaching its lower band, where a turn upwards is typical anyway, the market may be ready for an explosive advance. This is the basis for our major “BUY B9”. (See the 1982-3 chart attached below.)

Click this icon to see chart of DJI early in
the 1982-1983 Bull Market.

The most important PEERLESS sell signal is our famous “S9”. This occurs when the DJIA reaches an “overbought” position at the daily upper band with our internal strength indicators negative. We build proprietary internal strength indicators from advance/decline, up volume/down volume, total volume and momentum (today’s price divided by the price 25 days ago) data. These internal strength oscillators swing back and forth, from positive to negative, around a base line of 0.

 

 

In the chart of 1987 you can see how this juxtaposition of technical conditions prevailed in early October 1987 just before the DJI suffered its biggest plunge since 1929. From the pinpointed SELL S9 signal on October 5th, at 2640, the DJI plunged 30% in three weeks to 1740. You can see that this was the fourth “S9” signal of 1987. A fourth “S9” signals in a year’s time is particularly important in telling us that a major bear market is about to start. They are as powerful as they are rare. Earlier instances were in January 1973, July 1974, March 1981 and October 1987. These were the market tops right before the biggest general market declines since 1970. Is what I’m saying just perfect hindsight?

Don’t believe for a second that biggest stock market plunges cannot be predicted.

Look at what our subscribers heard in early October 1987, just before the biggest market crash since 1929. On October 8, 1987 with the DJI at 2517, but about to plunge to 1740 in two weeks, our bi-weekly PEERLESS FORECASTS (VOL. 4, NO. 18) warned: “DID YOU CATCH THE SELL S9 WAVE? (Lead Headline) With the DJI at 2640 our Hotline told subscribers this past Sunday and Monday nights (October 4th and 5th) to be on guard because a top was probably forming and that a decline to the lower band was about to commence. We pointed out that the DJI had tagged the upper band with the PEERLESS ... oscillator negative, for a SELL S9 PEERLESS signal...On Monday night I said that ... if 2500 did not hold, there would probably be an avalanche of sell orders, as institutional profits were taken more quickly than carefully.”

The rules that produce S9 and all the other automatic signals are embodied in our WINDOWS or DOS PEERLESS software. The HELP routines let you easily see the track record and get a full listing of each of the 1220+ major and minor signals since 1969. Their average gain on the DJI is 5.2%.

Hard work led to my discovery of how to recognize major stock market tops and bottoms as they occur. Early in 1981 I hand-entered 25,000 data elements into an APPLE II+. This was data for the 1970’s. There was no guarantee when I started that this enormously tedious effort would have any pay-off. I simply wanted to avoid personally being caught in another sharp down-turn like that of October 1978, October 1979 or February 1980. What I discovered and then published in 1981 went far beyond my wildest hopes, although at the time I certainly doubted if history would keep repeating itself. Only in the course of the 1980’s, when the system was applied on a real-time basis, and history did repeat itself over and over, did I realize I had, indeed, found a gold mine. It highlighted the timeliness of the historic SELL “S9” signals between 1972 and 1980. In 1988 I rewrote my book, emphasizing the major signals, the BUY “B4”, “B9”, “B10” and “B12”, as well as the fearsome SELL “S9”.

Unless we learn from market history, we are doomed to repeat the same mistakes over and over. PEERLESS shows you how the major tops in the 1970’s and then in 1983, 1987, 1989 and 1990 were all formed the same way. In each case the DJI reached an overbought condition at or near the upper band with at least one of our key internal strength oscillators still in negative territory. This is the bearish divergence I have labeled an “S9”. It is automatically signaled on graphs of the DJI by our software. No interpretation is needed.

Over the years our most skilled professional users have asked us to make PEERLESS as simple and automatic as possible. They do not want any doubt about the system’s position on the market. They know that their own emotions will invariably intrude if there is anything less than objectivity.

Our users also want to be able to back-test and validate these signals. Only in that way can they place confidence in them. To that end we routinely provide the back data to 1987 with a purchase of the software. The back-data to 1965 is also available.

Your job - and it’s not as easy as it sounds - is to believe the automatic sell signals from PEERLESS even if well-known market pundits are wildly bullish.

List of All Automatic Reversing Major Buy and Sell Signals
Since 1969


List of All Automatic PEERLESS Buy and Sell Signals Since
1969

 

Hit Counter
3-20-98