Three Indicators To Help Your Strategy Time Trades Better

One of the first suggestions that is usually made when looking to improve a strategy is to add a trend filter of some sort. One of the most common trend filters is a long term simple moving average (SMA), but that isn’t the only option to consider.

There are a wide variety of market timing trend filters, and some may use a completely different type of analysis from your basic strategy. It is up to you to determine, through backtesting, which one is the best fit for your system.

In his article for Seeking Alpha, author Fred Piard compared three different market timing indicators that could all be used as trend filters. What was interesting about his piece was that one of the indicators was based on technical analysis, one was based on fundamental analysis, and the third was based on macro-economic factors. This provided a wide range of options, and all of them performed better in backtesting than a standard buy and hold approach.

Using a simple timing indicator as a trend filter is a great way to increase profits and decrease risk.

Fred backtested each of these three indicators on SPY during the 15 year period from January 2, 1999 through January 17, 2014. During that time, a buy and hold benchmark would have produced an annual return of 4.6% with a maximum drawdown of 55.4%.

The Technical Trend Filter

The technical market timing indicator that Fred suggests is the classic golden cross indicator that makes use of the 50-unit and 200-unit SMAs. The indicator is bullish when the 50-unit SMA is above or equal to the 200-unit SMA, and bullish when the 50-unit SMA is below the 200-unit SMA.

Holding long positions in SPY when this technical indicator was bullish during the backtesting period would have produced an average annual return of 8.1% with a maximum drawdown of 17.0%. That is almost double the annual return of the benchmark with less than one third the drawdown.

Taking a short position when this indicator turned bearish during the backtesting period would have produced another 1.2% average annual return with a maximum drawdown of 47%.

The Fundamental Trend Filter

The fundamental market timing indicator that Fred discusses is based on the S&P 500 current year EPS estimate. When that estimate is above or equal to its value from three months prior, the indicator is bullish. When the estimate is below its value from three months prior, the indicator is bearish.

Holding a long position when this fundamental indicator was bullish over the last 15 years would have produced an annual return of 9.2% with a maximum drawdown of 20.2%. Once again we are going to dramatically outperform the benchmark. Compared to the technical trend filter, we can see that we are giving up some maximum drawdown in order to increase the annual return.

Taking a short position when this indicator was bearish over the backtesting period would have produced an annual average return of 1.9% with a maximum drawdown of 41.4%.

The Economic Trend Filter

The economic based market timing indicator that Fred tests uses the U.S. unemployment rate. This indicator is bullish when the unemployment rate is below or equal to its value from three months prior. It is bearish when it is above where it was three months ago.

Holding a long position during periods when this indicator was bullish over the past 15 years would have produced an average annual return of 8.84% with a maximum drawdown of 27.2%. While this indicator outperforms the benchmarks, it would be more risky than the other two indicators because of its maximum drawdown.

Taking a short position when this indicator was bearish during the backtesting period would have produced an average annual return of 3.2% with a maximum drawdown of 45.5%.