An abundance of trading literature suggests trend following is a time-tested way to score "consistent" profits in the market (whether stocks, FX or futures), with evidence going back at least to the great 1637 tulip bubble in Holland. However, when you start testing these systems over long periods of historical data you will likely reach the following conclusion: the fundamental or macroeconomic situation plays a dominant role in the return profile of trend following strategies. This fact is largely overlooked by technical purists that only consider a security's price movement.
The second conclusion you are likely to reach is: the success of trend following is more dependent on money management criteria than the criteria of when to enter a trade and when to exit.
I won't overwhelm you with evidence in one post, but allow me to start by showing you the return profile of a popular approach: buying a 55-day high or shorting a 55-day low. This was a key entry signal disclosed as part of legendary trader Richard Dennis' trend following system (System 2), which were taught to his "turtle" disciples, many of whom used this strategy to make tens of millions of dollars for Dennis' firm. So, we know it should work. …it did work.
Richard Dennis generally had two exit rules for each commodity, either the equivalent of a 2% decline in account equity, which I have no way of testing, or a 20-day low on a long position (20 day high on a short position). To make things more interesting, I will look at the 2-day, 5-day, 10-day and 20-day returns of buying 55-day highs in the settle price of the front month WTI crude oil continuous contract from 1984 to today, as well as like returns of shorting 55-day lows. Contrary to conventional trend following wisdom, results demonstrate fading breakouts has consistently been the more profitable approach, and even more so when adjusting for the macroeconomic situation.
Selling 55-day lows in crude was almost always a losing proposition across all hold periods in the study (from 2 to 20 days). The current environment and the global financial crisis of 2007-2009 are the most notable exceptions where this strategy paid off significantly.
For 55-day highs, buying breakouts has been inconsistent over the past 30 years, but generally performs best as a strategy after a major pullback in oil prices.
Unfortunately, trend following gurus don't often tell you that for any given commodity, the success of trend following strategies is highly variable, so much so that the most common trend following mechanisms that buy or sell breakouts will whipsaw traders for years before providing the situation where a prolonged trend emerges that also has low enough volatility that it doesn't wash out traders using common stop loss techniques (e.g., below an area of perceived support on a chart, or using a multiple of average true range).
In a quest to test various indicators and trend systems I have read about over the years, I built a robust software program that enables "back-testing" across any specified timeframe and any futures contract (continuous, roll-adjusted). I will occasionally share my findings in such Seeking Alpha posts--and feel free to submit a request to me.
Disclosure: I am/we are short WTI CRUDE OIL FUTURES.