Plenty of academic research, fund managers, and individual investors have demonstrated that investing with long duration trends is highly effective when applied on broad equity indices as well as other assets.
Generally, academic research finds support is strongest for investing based on trends that are identified at three months to twelve months in duration. But this is far from exclusive. There are plenty of examples of successful investors who follow much shorter trends; however, careful risk management and trade management becomes exponentially more important as duration shrinks.
I have examined two categories of complex trend systems: short duration (approximately two weeks to three months) and long duration (approximately three months to twelve months). I also looked at long/short applications of the trend system (betting with the upside when the trend was up, betting with the downside when the trend was down) and long/flat applications (betting with the upside, or being partially or completely in cash when the trend is down).
I developed what could be considered complex trend systems. They do not follow a single measure of trend. Instead, they examine trend direction using time-series momentum, price-minus-moving average, moving average crossover, and moving average direction. Combined, these four methods of identifying trends are calculated dozens of times across of range of time periods and carefully weighted to provide an exposure factor between -10 and +10, or between 0 and +10.
A Perfect Trading System is Futile
In the posts I linked above I applied both systems against the S&P 500 index. In doing this, I noticed the S&P 500 has gone through a shift in its nature. Before the 1980s shorter duration trends were very effective. After the 1980s longer duration trends performed substantially better. Markets shift in their own nature over time.
One market is also likely to behave differently from another market. For this reason I think it could be advantageous to diversify across trend durations. A good way to do this is through a complex system that covers a range of market conditions without constant tinkering on the overall system.
Trying to develop a system that is historically the best on a single market and doing that for several markets for adequate diversification is futile and maybe even reckless. A perfect trading system via a backtest is an assumption that the market will remain constant. Markets are not constant and do not behave according to defined parameters (such as a chemical element); instead, they are organic, multi-faceted, and a product of an ever-shifting environment and human behaviour.
Acknowledging this reality of markets, my trend system is an attempt to design an appropriate and diversified method of investing with the trend that works across numerous markets and widely varying market conditions over time. It is adaptable and responds adequately to organic conditions by scaling into and out of trends based on the strength of the trend.
Focusing on the long duration trend system I developed, we will examine the performance of my system on several major international indices: the Nikkei 225, the NASDAQ Composite, and the CAC 40. While limited by good data, I am attempting to choose indices which have unique performance characteristics.
Note: All data is calculated in nominal terms without dividends and is not adjusted for currency exchange rates, inflation, or local interest rates. The objective is to demonstrate the behaviour of my trend system across a range of markets.
The Nikkei 225 is Japan's most popular index and I can obtain daily data going back to 1965. As many investors know, the Nikkei climbed to insane heights over several decades, peaking in 1989. Today, the Nikkei is sitting at barely more than half of its peak value as the Japanese stock market and economy contracted and then stagnated for nearly three decades.
Given the volatile nature of the Nikkei in earlier years and the lacklustre performance over several decades, the Nikkei can provide a great real-life stress test on any investing method. If a well-grounded investing system works on both the Nikkei and the S&P 500, it should work almost anywhere.
3-Year Rolling Returns
Viewing the above three charts, we can see the some important performance characteristics of my multi-model trend system when applied to the Nikkei 225 index. Over the 53 year period the Nikkei price increased at a compounded annual rate of 5 percent. During the same period my trend system would have returned more than 8.25 percent annually.
As should be expected from any good trend system, the drawdowns of my system applied on the Nikkei 225 are much lower than the index drawdowns. While the Nikkei declined approximately 80 percent from peak valuations several times, my trend system had a maximum drawdown of just 30 percent (in 2012). The other major drawdowns of my trend system were held to approximately 15 percent—very tolerable for almost any investor.
The final chart, showing annualized 3-year rolling returns for the Nikkei and my trend system, paints another very compelling picture for trend investing. My trend system is much more stable and had much shorter periods of negative averaged returns. During periods of very positive returns the trend system very closely tracked the index; during periods of negative returns the trend system tracked well above the index.
Overall, the trend system investor would have finished nearly five times wealthier than an index buy-and-hold investor. During the backtest period, they would have experienced much smaller drawdowns and spend more time in positive return territory. The risk-adjusted returns are incomparable in the advantage of the trend investor.
The NASDAQ Composite index is possibly the U.S. stock market's most volatile major index. Filled with growth stocks, the index generated a 16x gain in the 1990s followed by an 80 percent drawdown the next two years. After a short gain, it was followed by another 55 percent drawdown in 2008.
Recently the NASDAQ has made a great gain and is at all-time highs again. The NASDAQ is another market that provides a great example of testing an investing model applied on a more volatile market.
3-Year Rolling Returns
The three charts above showing various metrics of my long duration trend system applied on the NASDAQ Composite index again demonstrates the advantages of trend systems on markets. While the NASDAQ clearly behaves different from the S&P 500 or Nikkei 225, a complex trend system is equally effective at managing risk and generating great returns.
Overall, in my backtest the NASDAQ generated a 9 percent compounded annual return. During the same period the trend application returned approximately 11.4 percent compounded annually. Investing wisely with the trend would have left the investor nearly three times wealthier over four decades.
The drawdowns are also much lower. The maximum drawdown was reduced from nearly 80 percent to just under 40 percent with the trend system. Most of the other larger drawdown periods were also half the severity with just two breaking briefly below a 20 percent decline.
The annualized 3-year rolling returns show improved stability, smaller drawdowns, and just a few brief windows of negative returns over a three year period for the trend system investor. While even the best trend systems will underperform the index in good times, that difference is quickly reversed when the index corrects to the downside.
The CAC 40 is France's most popular index and my data goes back to 1990. The CAC 40 is characteristic of a slower growing index and a mature market that has experienced economic difficulty. Price returns—at 3.3 percent compounded annually over the past 29 years—are not very high. Since 2000 the index price has returned -1.15 percent annually.
The CAC 40 can provide investors with a look at how an investing model can perform on a market with a long range-bound period.
3-Year Rolling Returns
The three charts above show a unique, and still highly beneficial, application of my long-term trend system on a stagnant market. The time period examined in this backtest for the CAC 40 is noticeably shorter than the previous indices.
While lagging the underlying CAC 40 in the beginning of the backtest period (when the CAC 40 was performing very well), the trend system soon overtook the index when trouble started. Performance for my trend system is positive after 2000—growing at 1.65 percent annually. This is compared to a negative annualized return for the index. Since 2007, even the trend system has been quite stagnant, with returns oscillating in a relatively tight range.
As usual, the drawdowns for the trend system are much lower than the underlying index. The CAC 40 itself has declined as much as 60 percent several times in the past twenty years while the trend system held drawdowns to approximately 25 percent. The CAC 40 has been in a drawdown since 2000—nearly 20 years—but the trend system made new highs in 2010.
The annualized 3-year rolling return chart once again demonstrate better characteristics for my trend system. The range of returns is tighter, the system is stable, and periods of negative returns are generally shorter less severe.
My complex trend system, without any changes made from one application to the next, is highly effective across a range of markets. This speaks to its adaptability and mitigation of market risk, timing risk, and specification risk. The benefits are even more pronounced when used in a properly constructed portfolio with several markets that behave differently.
Trend investing may not always provide the highest absolute returns in any single period and certainly should underperform in the upwards half-cycle of any market. But when appropriate designed, a trend investing system can be relied on to provide excellent risk-adjusted returns across full market cycles. We see smaller drawdowns, more stable performance across time periods, and carefully calculated exposure to risk that is continually updated and monitored.
In the next post discussing my multi-model trend system I will demonstrate the use of this exact same system on completely different markets: currency pairs. Currencies are great markets for investing both with the upside and downside of a specific currency pair. They are also demonstrated to provide returns which are uncorrelated with equities, making them a good candidate for diversification in your portfolio.
Comments & Questions
All comments are moderated before being posted for public viewing. Please don't send in multiple comments if yours doesn't appear right away. It can take up to 24 hours before comments are posted.
Comments containing links or "trolling" will not be posted. Comments with profane language or those which reveal personal information will be edited by moderator.