Skip to content


Spread the love

The Reversal System

rev goldOne of the most important discoveries made at Princeton concerning economic and market behavior is that the market system is a dynamic complex network of nonlinear activity possessing an incredible degree of inherent order. The seemingly random appearance of price activity is merely a mask that hides the true nature of events. The proof of this statement is found most vividly in what we call our Reversal System.

Our Reversal System stands alone in the midst of conflicting economic theories and simplistic one-dimensional trading system. Its very existence is based on the key principles of physics. The formula behind our Reversal System is strictly proprietary. The numbers generated by this system we refer to as Reversal Points.

Our Reversal System fulfills a very simple purpose. In any market or economic statistic, there is some point, if crossed, which marks the beginning of a change in trend. For example, during 1987 there must have been some specific price barrier that was penetrated which forced the long positions to start selling. These specific price levels exist in all forms of a time series and might be thought of as key pressure points.

By viewing economic and market behavior as a complex network of nonlinear activity, we were able to focus on the development of a model that was completely nonjudgmental, thus avoiding any human interpretation. Working on a pure mathematical model that was designed for nonlinear activity, our goal was eventually achieved. Since 1970 the research, observation and implementation of our Reversal System has led to a new level of understanding markets and their behavior.

Reversal points are generated each time a market or economic statistic produces a new isolated high or low, either on an intraday or closing basis. We classify our Reversals as Major, Intermediate or Minor depending upon the importance of the particular high or low. For example, the Reversals generated from the 1980 $875 high in gold would be classified as “Major”, whereas the 1983 high of $514.30 would be referred to as “Intermediate” Reversal points. Any reaction high along the way would be classified as a “Minor” Reversal point. The same is true when dealing with lows with the deepest low being referred to as the “Major” Reversal.

We also differentiate between Reversals generated from highs and lows through the use of our terms “Bullish” and “Bearish.” The Reversals generated from a HIGH are referred to as “Bearish Reversals.” If the market should close below these points, then the uptrend will have “reversed” into a bearish or declining trend. Likewise, Reversals generated from a low are referred to as our “Bullish Reversals.” Should the market close ABOVE these points then the downtrend will “reverse” into an uptrend.

Unlike many of our other models, the election of our Reversal Points provides an actual BUY or SELL signal. The “election” of a reversal is achieved only on a CLOSING BASIS! The Reversals themselves offer key areas of support and resistance within any market. They are the primary pressure points within price activity. Often, minor fluctuations tend to BOUNCE off of these Reversal Points so precisely that this model, above all others, tends to offer proof to our clients that there is indeed a hidden order within chaos.

Our Reversal Points are also generated with a view toward defining trend in the most detailed fashion possible. From each and every high and low, the Reversals that are generated come in sets of four precise points. Each of these four Reversal Points classifies the current trend by Immediate, Short, Intermediate, and Long-Term price activity. Only when ALL FOUR Reversals have been elected do we consider that there has been an important shift in trend. The election of the first reversal warns that a move to the second is likely. An election of the second signifies a move to the third and so on.

It was also discovered that, on occasion, two or more of the Reversals generated from the same high or low would line up precisely or be separated by a single basis point, such as 10 cents on gold. When this occurred, the election of such price levels sparked an immediate and abrupt change in trend. We came to respect these Reversal Points more so than individual Reversals and began to refer to them as “DOUBLE” Reversals. Through years of historical research, we also discovered that on even rarer occasions, three of the four Reversals would line up which we called a “TRIPLE” Reversal. Double Reversals materialize a few times during the course of one year on a daily level, on a weekly level they may only develop once every two or three years. As for the Triple Reversal, we saw this live only once in the US Treasury Bond futures in 1989. Otherwise, this Triple Reversal has not been generated in any US market on any level since 1929!

Our Reversal System was originally discovered on intraday trading on a tick by tick basis. It was later discovered that this model was equally accurate when extending it to daily, weekly, monthly, quarterly and yearly levels of price activity.

The Major Weekly Bearish Reversal in gold generated from the 1980 high of $875 was $280.50. This number was published in our monthly publications between 1980 and 1985 and these publications are copyrighted. Our timing models were pointing to a major low for gold in 1985 in conjunction with our Economic-Confidence Model. When the 1985 low was established, the precise low was exactly $280.50 basis New York spot! For years that number was active on our models and if it had been elected it would have warned of a decline down to $192. Instead, this Reversal provided the precise support from which gold then rallied back to $502 moving into the end of 1987. Once a major reversal is generated, it remains valid until elected, no matter how long it may take. The uniqueness of the Reversal System and its accuracy is a result of the model taking into consideration every possible variable. Its projection is fixed because it has figured out time, price and interaction between relationships. The difficult part is to try and figure out what fundamentals could cause such a projection five years from the date it was generated.

For more than 20 years we have been studying this model in live markets as well as in historical research. The following offers just a few interesting observations we have made since 1970.

  • At major highs and lows in activity, it is normal to generate a “Double Reversal” on a weekly or monthly price level or higher.
  • Each market reacts in a slightly different manner to the Reversals. The metals, currencies and bonds have a high degree of precision. Market movements will bounce precisely on these specific Reversal points. Stock indexes, on the other hand, will find that absolute precision is rare. For example, the Monthly Bearish Reversal generated from the 1987 major high in the S&P 500 nearest futures was 18100. The actual intraday low during the 1987 Crash was 18130. This tends to suggest that since the S&P is introducing a series of 500 variables into the equation, it is possible that absolute precision begins to suffer. Whereas absolute precision is the norm in the cases of gold and many other singular commodity markets.
  • Once a Reversal has been executed, its validity as a Buy or Sell signal is ended. Nevertheless, old Reversals of major or intermediate levels have often reappeared again from completely different highs or lows. This was the case in gold between 1983 and 1984. The Reversal in question was $404.60. This number was generated more than 10 times from separate and distinctly different highs and lows. Initially, $404.60 was a Bearish Reversal. Once elected, it suddenly became a Bullish Reversal. Eventually, this Reversal provided the precise highest daily closing in March 1984 just before the major decline down to $280.50 began.
  • A Reversal is NOT elected should the market close precisely on it. In the case of Bullish Reversals, a successful election requires that the market close ABOVE the Reversal. Likewise, an election of a Bearish Reversal requires that the closing be BELOW the Reversal.
  • Although once a Reversal has been elected and it can NO LONGER provide a buy or sell signal, elected Reversals may still provide resistance or support at times during reaction price movements.
  • At times, a market will elect a Reversal by barely one or two basis points. This frequently results in an immediate move in the expected direction of the Buy or Sell signal.
  • When a Reversal is elected by a close greater than 1.5% away from the actual number, the market will usually retrace back to the Reversal Point and test that price level before moving in the anticipated direction of the signal. For example: a Bearish Reversal in gold at $400 is elected by a closing at $390. Gold might rally the next day reaching $400 intraday and then decline to the next level of support thereafter. This may NOT be the case when dealing with several Reversals elected simultaneously.
  • Reversals are generated on all levels of price activity from daily to yearly. Daily Reversals are elected on a daily closing basis. However, in each case, the next higher level of price activity will override the signal of a lower price level. Therefore, a Weekly Reversal will override that of a Daily when it is the last trading session of that particular week. Monthly Reversals will override both daily and weekly Reversals on the last trading day of the month. A daily or weekly closing below a Monthly Reversal during the course of a month DOES NOT constitute an election of the Monthly Reversal. Elections of Weekly, Monthly, Quarterly, and Yearly Reversals take place ONLY on the last trading session of a given week, month, quarter or year respectively!
  • The election of a Reversal normally indicates that the expected high or low that should unfold could take place in as short a time span as 1 to 3 units of time, be it daily, weekly, monthly or quarterly. Therefore, a low might develop the very next day following the election of a Daily Bearish Reversal or within the next few days. The same is true for all price activity levels.