During early 2008, Maize Asset Management developed a proprietary phenomenological commodity trading model, which is embedded with statistical modules. These modules help define the direction, strength, volatility and velocity of each commodity’s major and minor trends. This modified momentum and trend following model can be tailored to any market.
The first step is locating the trend. We use long and short exponential moving averages (EMAs) to find the primary trend. Back-testing of results finds that the euro four-hour time frame trend was best monitored by the 10-period and 20-period EMAs. When this trend is bullish (10>20 EMA), the shorter trend is above the longer trend, the model generates a buy signal; when bearish, it generates a sell signal. Our model’s foundation is a simple EMA model, buying or selling when the trends change. We only enter a trade upon a trend change, however we use “girth” as a signal for early exit. In “Getting out with max profits,” the purple line is the longer EMA and the light blue line is the shorter. Therefore the trend is bearish and the model’s basic position is short: 20 EMA > 10 EMA.
Next we determine the velocity, momentum and volatility of the trend. It is necessary to compute these factors, otherwise the trader will not exit in the most profitable position. Reviewing the graph, our model signals exit below $1.2600 and not above $1.2800, where the EMAs cross. A simple trend trade would be set to exit the short and enter a long position on the candlestick after the EMA cross. Using the girth indicator, we exited at $1.2597. We took a long position at $1.2840 after the EMA cross. The girth indicator added 200 pips to our profits.
The mechanics of calculating basic girth are simple. Basic girth is the difference between the longer EMA and the shorter EMA. Girth = (10 EMA - 20 EMA). Implementing trades based on girth requires not only calculating the basic value but integrating this result with momentum, velocity and volatility. Girth coupled with velocity is used to optimize early exits. Girth coupled with momentum and velocity is used to add notational size to exiting positions.
Higher volatility of the underlying price can indicate a higher probability of EMA cross and therefore a weaker trend. High price volatility results in earlier trade exits. Higher velocity and momentum of the underlying indicate a stronger trend and require adding to winning positions. The girth factor indicates both early exit before losing significant profits and increasing the notational size of profitable positions. These indications are consistent with the old trading maxim: Add to winners and cut losers.
The girth number combines these concepts into a single quantity. Girth number = f (volatility, momentum, velocity) per time frame per commodity. Choosing a girth number is an ongoing optimization process.
How did Maize exit a profitable trade early without waiting for the EMA cross? “Girth table” illustrates how Maize used girth to exit early from the trade described above. Girth is calculated for every four-hour period. At the close of each period, the model compares the four-hour girth to the girth number. If the four-hour girth is less than the girth number, then the trader is instructed to close the short position at the open of the next four-hour candlestick. Early exit is indicated at 2 p.m. on Oct. 28, when the absolute value of girth (column 4) is less than the girth number. Girth does not depend on market conditions. Employing girth in our model requires the calculation of the girth number, which incorporates current underlying price volatility, momentum and velocity. Because of recent conditions in the euro market, our girth number is currently 0.01. It was 0.0007 in May, when the model was first employed.
This model is particularly profitable in the EUR/USD because of its strong trends. Tight range-bound markets produce poorer results. By following the model’s indications, we have identified and profited from recent movements in the EUR. The model also applies a discipline to the traders, which helps to minimize the likelihood of human error.
Leslie K. McNew is chief investment officer for Maize Asset Management and clinical professor of finance for the Hanley Group Derivatives Trading Center, University of Dayton School of Business Administration.