Building a system with simple technical tools
Many investors and active traders use technical analysis to help identify trade entry and exit points. New technical indicators are in a constant state of development, including commercial tools as well as those that reside in the public domain. While many of these provide an innovative look at the markets, sometimes a tried and true method proves to be good starting point for traders.
Improved with a few modest tweaks, or used in non-traditional ways, these time-tested tools can become the basis for profitable trading systems.
Here, we start with two common technical indicators—a stochastics oscillator and a moving average—and apply them to daily E-mini S&P (ES) contract to show that small changes in how we use indicators can lead to big improvements in a trading system’s performance.
We start with a stochastic oscillator, which has been used by traders and investors since the 1950s to anticipate areas where the market may change direction. It measures price momentum by comparing current price to a price range specified by the look back period. The resulting indicator is a pair of closely-linked lines that meander back and forth between two values: 0, at the bottom of the indicator pane, and 100, at the top. As the stochastic approaches values closer to 100, it signals an overbought market; conversely, as it heads towards 0, an oversold market is indicated.
Traders commonly set symmetric threshold levels to identify overbought and oversold levels; for example, overbought/oversold levels are often set to 90 and 10, or 80 and 20, respectively. A popular way to use the indicator is to take a long position when the stochastic crosses above the lower threshold (after it has already crossed below), and then enter a short position when the stochastic crosses below the upper threshold (again, after it has crossed above).
To begin our testing, we create a simple stop-and-reverse strategy for the daily ES chart, where each long trade is closed and a new short trade is initiated every time an opposing signal occurs, and vice versa, so we are always in a trade. Rather than taking a traditional approach, we will attempt to enter trends early by going long as soon as the stochastic crosses below the lower threshold, and going short once it rises above the upper threshold.
We use a slow 14-period stochastic with symmetric thresholds, going long when the stochastic crosses below 20, and going short when it crosses above 80. This system produces unremarkable results when tested on five years of historical ES data—a low total net profit and profit factor, and a high maximum drawdown—so we need to make some changes.
Here’s where we can tweak the stochastic a bit so it provides more accurate signals for a specific market. Because markets themselves tend to be asymmetric, we can adjust the stochastic by setting asymmetric overbought and oversold levels, so we abandon the traditional symmetric 80/20 thresholds, and use asymmetric thresholds of 90 and 40.
We now anticipate market turns by selling as soon as the stochastic reaches 90 (overbought) on a daily chart, and buying when it reaches 40 (oversold). Even with asymmetric overbought/oversold levels, we still see a comparable amount of buying and selling. “Stochastic modified” (below) shows decent results for the testing periods; however, a potentially high draw down and lack of stop loss make this more of a “proof of concept” than a tradable strategy at this point, so we’ve still got some work to do.