Are Follow Through Days a Reliable Indicator of Downward Trend Reversal?

The concept of the “follow through day” (FTD) was introduced by William O’Neil in his book, “How to Make Money in Stocks” (HTMSS). Its purpose is to increase confidence in when a market correction has run its course. If a FTD occurs shortly after a downward trend reversal, then the likelihood that the upward trend will continue is supposed to be enhanced. Conversely, if a FTD does not occur, then the rally is supposedly more likely to fail.

Since the concept was first introduced, the definition of a FTD has undergone several revisions. The ‘Glossary’ at Investors.com currently says: “a ‘follow-through’ day is identified when the index closes up 1.7% or more for the day on a significant increase in volume from the day before. The first two or three days of a rally are normally disregarded as it has not yet proven it will succeed and ‘follow-through’ with power and conviction. ‘Follow-through’ days therefore generally occur the fourth through seventh day of the attempted rally. They serve as a confirmation that the market has really changed direction and is in a new uptrend”.

There are at least three problems with this definition which make it difficult for the investor to interpret when a FTD has occurred: “the index” is not defined, the first day of the rally is not defined and neither is “significant increase in volume”.

Turning to HTMMS allows us to resolve most of these difficulties. The 3rd Edition, p. 65, refers only to the DOW index but says “A rally attempt begins when a major market average closes higher after a decline, either from earlier in the day or the previous session.” On p. 66 we learn “The market’s volume for the day should be above its average daily volume in addition to always being higher than the prior day’s trading”. (Note that the number of days to include in the average is left undefined. I could not find a definition either in HTMSS or on investors.com).

Based on the above we can now define the rules for a FTD:

  1. Day 1 of a tentative rally begins when a major index closes higher than the day’s low or the previous day’s close.
  2. On the fourth through seventh days of the rally there must be a close at least 1.7% higher than the previous close.
  3. On that day, the volume must exceed the previous day’s volume.
  4. On that day, the volume must be higher than the average daily volume. We will use the commonly accepted 50 day moving average.

Now that we know how to recognize a FTD, is it useful? If so, it would confirm a new uptrend within a few days of a downward reversal allowing entry into the market with some confidence. On the other hand, if a FTD does not occur but the upward trend continues, then an opportunity to enter the market near the bottom will have been missed.

I looked at six reversal points in the current bull market including the market bottom using the DOW (see chart) and applied the rules above for the DOW (DJI), Nasdaq Composite (IXIC) and S&P 500 (SPX).

DOW Bottom Reversals

Although the rules suggest an FTD should be between the 4th and 7th day of the rally, I looked at each day until the 10th day. The table below shows the days in the rally when an FTD occurred.

Day 1 Date Index FTD Day of Rally Comment
3/9/2009 DJI Yes 2 Too early to be reliable
IXIC Yes 2,4,8 Day 4 just right
SPX Yes 2,8 Day 8 just acceptable
7/10/2009 DJI Yes 4 Just right
IXIC Yes 4,10 Just right
SPX Yes 4,10 Just right
2/9/2010 DJI No
IXIC No
SPX No
7/2/2010 DJI No
IXIC No
SPX No
11/30/2010 DJI No
IXIC No
SPX No
3/16/2011 DJI Yes 3 Too early to be reliable
IXIC No
SPX No

The FTD rules were successful in confirming the market bottom of March 9, 2009 for the NASDAQ and S&P 500 but missed the bottom in the DJI which could have caused doubt that the bottom had been reached, particularly after such a long and deep decline. It correctly confirmed the trend reversal on July 7, 2009 in all three indexes. Since then, the FTD rules were not successful in detecting the trend reversals. If you relied on this signal alone to re-enter the market then you would have missed some important trend reversals.

This study has only looked at cases where an obvious downward trend reversal occurred and the ability of the FTD rules to detect that reversal. It has not looked at cases where an FTD occurred but a trend reversal did not  eventuate.

Conclusion

Although the FTD rules did detect the March 9, 2009 market bottom in two out of three major indexes, the rules as construed here do not appear to constitute a reliable system for detecting downward trend reversals. The period of study was limited to only three years of a bull market, however, and a longer period of study may have given different results. The important question of does the system yield false positives was not examined.

Also not considered was the depth of the correction that must occur before the system can reliably detect a downward trend reversal.