Ways to increase success rates in a trend-following strategy

One of the simplest trading strategies is to back a trend on a breakout. Some short-term traders will enter long (short) on a 20-day high (low). Others will look for a 55-60 day high or low. Some will get in on a 52-week high or low.

The underlying assumption is that the trend will continue and strengthen. This is somewhat against the odds. Computerised trend-following systems have been around for roughly three decades and back-tested data reaches back much further. Breakouts (and breakdowns) fail more often than they succeed. However, when the trend does sustain, it fetches high returns. This happens often enough to make trend-following a profitable strategy despite low strike rates.

What can a trader do to filter out likely trend failures? The more common filters involve volume analysis. Volumes should expand with a breakout. Experienced traders will tinker to set "expansion" indicator levels. For example, some look for a doubling of the last 20-days average volume. Others analyse delivery ratios (delivery : overall volume) in equities, or open interest and put-call ratios in derivatives.

Another filter is peer group trends. If most peer group stocks are moving in the same direction, the trend is more likely to be sustainable. A broader filter is to check if the target stock's movements are correlated to overall index trends. Again, a trend is more likely to be sustainable if it is in line with a broader market-wide movement.

None of these filters guarantee success. You might see trend failure after an apparently strong breakout, backed by volume, and in line with a broader move. But when these filters are applied with discipline, they increase the odds of success for a trend-following strategy. Obviously, given a chance of failure, the trader must also set stop losses. As and when the position turns profitable, the stop is re-adjusted.

Less obviously, most successful trend-following systems rely on pyramiding rather than all-in exposure via a single position. If an experienced trader is prepared to take say, a maximum of 10 lots exposure to a contract, he is likely to take it in stages, starting with say, three to four lots and adding to the position only when already in profit. This reduces the chances of a big loss.

One additional wrinkle occurs to me. If a trend following strategy is ruled out by the additional filters, should the trader opt for a counter-trend strategy? That is, should he assume trend failure and act accordingly? It seems logical.

There isn't a problem with the mechanics since the same indicators are used. But very few traders seem to be prepared to adopt both trend-following and counter-trend strategies.

The author is a technical and equity analyst