Friday, October 24, 2014

When Stop Losses Stop You From Winning

More than a few traders were stopped out at inopportune moments during recent volatile trading.  There are few things as frustrating as getting stopped out of what, ultimately, turns out to be a good trade.

I generally find that traders are much less detailed in their planning for exits than their entries.  They will select support/resistance levels as potential stops, but as often as not these represent pain thresholds and not sober assessments of a market's likely movement over a holding period.  Not a few traders lately were stopped out simply because their position sizing was not matched to the market's higher volatility, causing pain thresholds to become exceeded prematurely.

Suppose a trader exits a long position after several days of losses.  That seems prudent.  When we go back to 2006 in SPY, we find that five days of weakness (fewer than 50% of SPX shares trading above their five-day moving averages) has brought a next five-day return of +.43%.  Five days of strength have brought a next five-day return of -.08%.  In other words, if you bought on strength (prudent waiting for price confirmation) and bailed out on weakness (prudent stopping out), you were the very embodiment of negative timing alpha.  Stopping out in such a manner stops profitability.

So what is a trader to do?  One possibility arose recently when I was building a short-term trading system for the ES futures.  The system was modestly profitable after a one day holding period; better after two days; and considerably better after three days.  When I examined next day returns, there were many points where, had I been trading intraday, I would have stopped out.  Those trades went on to become nicely profitable over the next two trading sessions.

After tweaking possible stops and no doubt overfitting some of them, I said to myself, "This isn't working!"  So I eliminated price-based stops altogether and set a catastrophic stop level that I could live with, but that had never been hit during my in- or out-of-sample periods.  So once the signal occurred, I was to hold for three trading sessions and stick with the trade as long as the catastrophic stop did not trigger.  I sized in a manner that allowed me to weather expectable drawdowns without undue trauma to the portfolio and sat back to let trades breathe.

I stopped price-based stops and let time and position size define my risk/reward.  Early days, but so far it's been easier on the psyche and better for the bottom line.

Further Reading:  Why Traders Keep Losing Money