Trend Following versus Gap Downs

For the last few weeks, I've shown some nice trends that have developed and were captured by my moving average system.  Like any system, however, my system has weaknesses.  The achilles heal of my moving average system is "V" bottoms or tops, as the following chart shows:


After a moving average crossover in December, EWJ proceeded to trend impressively until about 3 weeks ago.  The ETF since then has been crashing down, erasing most of the year's gains, and what's worse is that the moving averages still have not crossed yet.

I would now like to introduce a rule that I have developed that will reduce the likelihood of this problem occurring.  The rule is:  if a stock gaps down significantly, then liquidate the position the next day, regardless of what the moving averages indicate.   I define "significant" as a a move greater than 1 ATR for the stock.

Here is an example:




Despite ending up with over a 14% profit,  AMGN's uptrend ended badly with a big gap down on earnings day.

The way I see it, when a stock gaps down significantly, it tends to cripple the stock.   After being crippled, the stock either can "heal" over a period of several weeks and eventually make a new high, or, as is often the case, the stock doesn't heal and is devoured by the bears, so to speak.

I sold AMGN shortly after the gap down.





Another stock that I traded this year was GV.  As the chart above shows, the gap down fatally wounded the stock and, with the benefit of hindsight, was a signal to exit.

Another chartist who I have been following for several years, Olivier Tischendorf, has also noticed this same pattern and commented on it here.


Lastly,  here is a chart of LNKD:


In what I believe was a case of good process, bad outcome,  LNKD ended up being my worst trade so far for 2013.

I sold the stock shortly after the huge gap down.   It's not easy to do.   Your ego definitely wants you to wait for the stock to get back to break even, so that you can be "right", but that's a mistake in my opinion.   Based on thousands of hours of chart observation, I can say that more often than not, these type of gap moves are an exit signal and if don't get out, you can end up getting crushed: