At the root of most trend following systems is the idea that you want to be long markets that are moving higher or short markets that are moving lower. One of the simplest and most popular ways to identify markets that are trending up or down is to look for ones that are making new 52-week highs or lows. This simple signal can be used as the root of a very simple and easy to follow trend following system.
About The System
In his book, Unholy Grails, author and trader Nick Radge discusses a number of different systems providing set rules and backtesting results. One of the first systems he describes is based on the concept that stocks making new yearly highs are likely to continue higher, and stocks making new yearly lows are likely to continue trending lower. This is one of the core concepts behind most systematic trend following strategies.
Radge points out that a system that is based on new yearly highs and lows is very user-friendly because most financial resources publish data for stocks making new 52-week highs and lows. This means that we could literally trade this system manually without any trading software.
One way that Radge simplifies the system is by assuming that there are approximately 250 trading days in a year. Therefore, any stock making a new 250-day high or low will also be making a new yearly high or low. This means we can treat this system as a very simple 250-day breakout system.
This system establishes a long position on the open the day after a stock makes a new 250-day high and then exits that position on the open of the day after the stock makes a new 52-week low. These 250-day highs and lows are easy to monitor using a price channels overlay, which is available in most charting packages. The solid yellow lines on the chart below represent the 250-day high and low prices.
Enter Long When:
- Price Closes At New 250-Day High
Exit Long When:
- Price Closes At New 250-Day Low
In order to backtest this system, Radge used the stocks in the All Ordinaries Index, which is the most popular stock market index in Australia. He compares his results to the All Ordinaries Accumulation Index, which is a benchmark that represents a buy and hold approach on the Australian equities markets.
His testing period ran from January 1, 1997 through Jun 30, 2011. The account started with $100,000 and sized positions at 5% of the account. Commissions and dividends were also taken into account.
Trading the New Yearly High System over those 14 years would have produced a total of 170 trades. The Compound Annual Growth Rate (CAGR) would have been an impressive 18.21%, more than doubling the 8.78% CAGR of the benchmark buy and hold account. The system was profitable on 53.3% of its trades and posted a Sharpe Ratio of 0.395.
The glaring negative of the backtesting results would be that the New Yearly High System posted a maximum drawdown of over 50%.
One of the most common things you will hear about system development is that traders place far too much emphasis on entry and exit signals. Conversely, there is generally not enough emphasis placed on risk management and position sizing. This system is a great example for those philosophies because its entry and exit signals appear to work just fine, but its 50% drawdown indicates that it doesn’t do a good enough job protecting its profits.
The New Yearly Highs System has a significant strength in that it works. Its returns actually beat a buy and hold approach by more than double. However, its weakness needs to be addressed, because it would be an absolutely nightmare to attempt to stick to the system while sitting through a 50% drawdown.
Improving The System
The first thing I would do to improve this system is introduce a trend filter using the 200 day simple moving average (SMA). This filter would simply require that the general market be in an uptrend in order for the system to enter a long position in any single stock. This would ensure that the system was not taking long positions in stocks that were trying to trend higher against a downtrending general market. We should have more overall success if we consistently trade with the trend of the overall market.
Another condition I would add to the system would be an ATR-based trailing stop that would ensure that we locked in profit from any profitable trades. It would also minimize our losses on losing trades. Adding this stop might cut down the overall returns a bit by cutting short trades that would have eventually turned profitable, but the downside protection would likely be worth that sacrifice.
Changing the Universe
One of the things that Radge suggests is trading the system exclusively on the market index as opposed to individual stocks. He suggests that this would cut down on volatility and then proves that concept with backtesting results. This idea lowered the CAGR to 13.92%, but also cut the maximum drawdown down to 36.03%.
Keeping with the line of thinking that trading the system on indexes would reduce volatility, I would be curious to test the system trading across multiple indexes or ETFs. It would be interesting to see how the system performed on the ETFs traded by the Ivy Ten System or the markets recommended by Andreas Clenow in Following The Trend.