Henri Caron

# Research Series - Trailing Average

I have been playing around with some new techniques that lean more towards active forms of investing, and I ended up developing a new model which I like to describe as Trailing Average. I mentioned already that it is very hard to time the markets exactly right, and by consequence predict major jumps in stock prices. Instead, I wondered if I would be able to deduct something from stock prices AFTER they went through either a major increase or decrease. Likewise, that would allow me to monitor sudden and hefty price changes, and make a prediction on how that stock price would behave in the near future. The results look extremely promising.

I have been scrutinising historical stock data of the past two years and looking for all stock jumps over 4%, after which I have been looking at the course of that same stock price for the 30 days following the price jump. Finally I have calculated the probability of having either profit or loss for each of the following days. Said otherwise: If a specific stock increases by 4% or more, what is the probability of having either profit or loss for each of the 30 consecutive days? The reason why I think this is a very interesting measure, is because now it allows me to use those significant jumps in stock prices, as a marker, a warning signal to open a position on a stock and know after how many days I ideally close that position in order to benefit a maximum return.

I’ll illustrate with the example below: For the stock $CYBR, I have found 8 days (= observations) during the past 2 years for which the stock closed with a 4% increase compared to its previous day’s closing price. If I would open a __buy position__ on that stock the following day, ideally I would close it after 15 days because a that point I will have the highest probability of having a profit, which is an average return of 5,81% (after 15 days), with a standard deviation of 4,32%, resulting in a probability of 91,10% for profit (assuming a normal distribution).

I have summarised the information for a couple of stocks below. Besides $CYBR, other stocks like $MU, $NFLX and $SHOP are interesting stocks for which to open a buy position the day after an increase of minimum 4%, and ideally close that position after 27, 29 and 28 days respectively. Stocks like $SYNA and $IMPV have a profit probability of less than 50%, indicating there is a higher probability of ending with a loss rather than profit after 1, respectively 13 days.

As I said, those results look very promising. My target is to start including these insights in my investment decisions very soon and, because of the short-term nature of these investment decision, apply a small (2x) __leverage__ on those positions to further increase the potential returns.