In the last post (here), I examined what happened after a stock moved down a significant multiple of its previous day’s Average True Range (ATR20 in this case). Stocks tended to have up days on day 1 and days 3-5, with a down day on day 2 as an average. What about bursts upward? Are they the opposite? Would they make a good shorting opportunity?
Turns out, the answer is no!
Big pops upward show a a similar tendency to gain over the next five days. In this case, there are no days that are down days (on average). The above graph shows stocks that popped up at least 3x their previous day’s ATR(20) value.
The results for a >5x pop are similar:
The results are even stronger, although day 3 shows a little weakness compared to the other days. We’ll use that to our advantage when coming up with a hypothetical trading system.
- Stock must be a member of the Russell 3000 index at the time of the trade.
- Stock closing price is greater than $15 and the 20-day average volume is > 100000 shares.
- Look for a stock that pops upward at least 3x its previous day’s ATR(20) value.
- If the third day after the pop day is a down day, buy at the close.
- Sell at the next close, regardless of whether you’ve made a profit or not.
- If you have more choices than you want to trade, pick the ones with the biggest drop on the day you’re going to enter.
An example of a winning trade:
Again, there are some parameters in there that are derived from more than just looking at a bar graph. I optimized over the period of 2000-2009, and looked at the results for the out-of-sample period of 2010 to the present. Here’s an equity graph for the combined IS/OOS periods.
Just like in the past post, I must remind you that there are a lot of trades, and you’ll be racking up a lot of commissions. Your position/commission size would need to be pretty substantial to make this work. The Car/MaxDD and Sharpe ratios are ‘meh’ as well. This is more as a theory or proof of concept than anything else.
Sometimes you can get trading ideas from looking at statistical quirks or anomalies, and then using that as a springboard for developing systems. These are two examples of letting the numbers lead the system design, rather than the other way around. Many traders would have started with the idea “short the stock when it pops upward”, and then tried to design a system around that concept. The bar graphs told me otherwise.