The results were very interesting.
The first rule they tested was a simple crossing of %K above or below %D. They optimized the results and found that every time period was a loser except at 21 months. Clearly, this method should be discarded. The authors then tested a more robust rule of requiring that both the %K and %D had to be rising for a buy signal and dropping for a sell signal. This tends to mean that trades will be in the direction of the longer-term trend. What they found was that every period length was profitable except for the shortest lengths below six. Adding this new filter significantly improved the value of stochastics.
Obviously, many investors will not want to trade in a market using a monthly signal but even these people will want to consider using this as a powerful trend indicator to filter short-term trades. In other words, take short-term trades only in the direction of the long-term trend as identified by the long-term stochastic crossover. Note, too, that this monthly method can be used to time longer-term trades such as forex hedges for corporations.
They then tested the signals on weekly data over two different time periods. In this case, only signals around 50 weeks were profitable. Longer and shorter lengths were unprofitable. I think we can safely conclude that we should ignore this method.
Just because these tests were with stocks doesn’t mean that they aren’t important for us forex traders. But markets are markets. They are all traded by humans. Techniques in one financial arena very often work in other financial arenas.
Colby and Meyers also tested the concept of selling overbought markets and buying oversold markets and found that it was a huge loser. Let’s just stick with following the trend.
WARNINGS Because it is an oscillator, many traders look at the slope of the stochastics for an idea of the momentum of the market. They want to, say, buy when the stochastics are trending down but the momentum of that trend is waning. In effect, they don’t want to wait for the crossover to get long. They want to get a jump on the stochastic crossover.
In other words, they will start to sell a market as the stochastics start to slow their advance. Take a close look at Figure 5.2. Notice that the %K will start to slow down its advance or decline while the %D is still moving at the same rate. Selling or buying when this happens will certainly give you a jump on the market change.
Source: The Profitability of Stochastics You Must Learn
