Simple mechanical trading systems can generate good profits

Trading a stock or a stock index is very simple at one level. Ignoring the rare third case (no change), it is like a coin toss: the price rises or falls. The trader asks a single, simple question: Up or down?  A slightly less simple way to look at prices is in terms of trend.

Here too, a single question is involved. But answering that question involves processing a little more information because the trader must know the two latest prices: Will the next price change be in the same direction as the last price change? If it is in the same direction, there is a trend.

All trading systems assume that either a trend will continue, or that it will not continue. There are many ways to set rules for trading either situation. Reversion or correction is always likely. But one long trend can be more profitable than six or seven reversion moves.

This brings us to the frequency versus amplitude argument, which is more sophisticated. A trader can look for high frequency of success. Or he can look for low frequency, high-amplitude trades. The concept of risk and reward obviously comes into this decision.

In a high-frequency system, a series of small profits accrue with an occasional large loss. In a low frequency, high-amplitude trade, a series of small losses will be compensated for by occasional large profits.  Any trade must set basic parameters. What timeframes are considered? What sort of magnitude of move is significant? This depends on personal preferences but a basic threshold is set by brokerage. Nobody wants to trade a move so small that brokerage exceeds profits.  

A day-trader who uses index futures can check out a very simple mechanical system. Suppose the index has closed up (or closed down) two sessions in succession. A trend may now be in force. The trader could go with the trend, buying the future if the trend is up (or selling if it's down). He may set a trailing stop loss of one per cent.

The signal of two sessions in succession going the same way has come up 12 times in the 61 sessions since January 1, 2015. But trends have actually lasted three sessions or longer, only eight times. Therefore, the trader loses 1 per cent on four occasions. That is minus four per cent from "system failures".

However, on the other eight occasions, the trend does last three sessions or more. There were four uptrends consisting of a total of 21 winning sessions. The total gains from uptrends would have been about 12 per cent. The exits would have cost one per cent each. Hence, the trader could have gained about eight per cent from the uptrends. There were four downtrends lasting a total of 20 sessions and generating about 10 per cent for the short-seller. Again, after deducting an exit stop loss of one per cent, the trader gains six per cent.

The net gains after deducting for the five failures will be about nine per cent. A return of nine per cent is not bad, given that the total gains amount to 2.5 per cent from January 1 (Nifty 8,284), to March 31 (8,492). The system worked well here because the market moved with a sequence of trends, both up or down. Another period when the market did not throw up so many trending sessions would yield fewer returns for this simple trending system. It's worth noting how such a mechanical, easily implemented trading system can still generate comfortable profits.

The author is an equity and technical analyst