System Design  |  JUL 1996

A/D Volume, New-High, New-Low System by Dennis Meyers, Ph.D.

V.14:7 (304-310): A/D Volume, New-High, New-Low System by Dennis Meyers, Ph.D. Here, classic stock market indicators have been combined into a trading system for the market. Previously, I developed a stock market system using the weekly advance-decline line and the weekly new highs and new lows. While a system using weekly data is valuable for filtering out daily price fluctuation noise, it does, however, leave the trader exposed to large adverse price movements during the trading days between the weekly buy/sell signal calculations. This one-week gap of inactivity could result in serious losses, or at least the loss of opportunity to get in at bargain prices. This time, I will build on the ADL-NH-NL system and develop a new market system using the daily New York Stock Exchange (NYSE) advance-decline volume and the daily NYSE 52-week new highs and new lows. Volume has long been considered not only an indicator of the strength of a price move but also a precursor of one. Beginning in 1965, daily volume on the NYSE was tabulated separately for advancing and declining issues and called advance/decline volume. Many useful indicators were developed using the advance/decline volume, with the Arms indexY´ (also known as the trading index, or TRIN) probably the best known. The NYSE daily 52-week new highs (NH) and new lows (NL) have often been cited as valuable indicators of future market direction. Since January 1978, the NHs and NLs have been calculated on a 52-week basis. Before January 1978, however, the NHs and NLs were calculated differently: Starting on March 15, the NHs and NLs were calculated from closing prices on December 31, two and a half months before. As the year progressed, the NHs and NLs were calculated from that same December 31. On March 14 of the following year, the NHs and NLs were still calculated from that previous December 31, 14 1 /2 months before. Then the next day, on March 15, the calculation date changed and the NHs and NHs were calculated from their closest December 31, two and a half months before. It is clear that the NH and NL series from January 1978 forward are a different series from the earlier NHs and NLs. As a consequence, systems developed using the post-1978 NHs and NLs probably will not work on pre-1978 NH and NL data. As previously, I will treat NHs and NLs as representing two distinct classes of investors with different time horizons and different market expectations. Therefore, indicators on the NHs and NLs will be developed independently of each other. This will be discussed in greater detail later.

by Dennis Meyers, Ph.D.