This post was prompted by Michael Covel’s interview Traders’ Magazine in which he claims that trend followers don’t try to make predictions. This idea that trend followers do not forecast returns is widely and frequently repeated. It is also complete nonsense.
Every trading strategy makes forecasts1. Whether these forecasts are explicit or hidden behind entry/exit rules is irrelevant. All the standard trend following systems can trivially be converted into a forecasting model that predicts returns, because they are fundamentally equivalent.
The specific formulation of the trend following system doesn’t matter, so I’ll keep it simple. A typical trend following indicator is the Donchian channel, which is simply the n-bar highest high and lowest low. Consider a system that goes long when price closes above the 100-day Donchian channel and exits when price closes below the 50-day Donchian channel.
This is the equity curve of the system applied to crude oil futures:
This system can trivially be converted to a forecasting model of the form
the dependent variable y is returns, and x will be a dummy variable that takes the value 1 if we are in a trend, and the value 0 if we are not in a trend. How do we define “in a trend”? Using the exact same conditions we use for entries and exits, of course.
We estimate the parameters and find that α ≃ 0, and β = 0.099% (with p-value 0.013). So, using this trend following forecasting model, the expected return when in a trend is approximately 10bp per day, and the expected return when not in a trend is zero. Look ma, I’m forecasting!
Even without explicitly modeling this relationship, trend followers implicitly predict that trends persist beyond their entry point; otherwise trend following wouldn’t work. The model can easily be extended with more complicated entry/exit rules, short selling, the effects of volatility-based position sizing, etc.
Footnotes- Unless they use random entries/exits.[↩]
severus says:
Good succinct post. The self appointed poster boy of Trend Following did author a couple of readable books but does go overboard at times with claims that would make a serious quant trader wince.
The Whole Street’s Daily Wrap for 6/30/2014 | The Whole Street says:
[…] …………. […]
David Varadi says:
your observation is exactly correct- i always found it funny that people never seem to see it that way. i like your blog a lot -feel free to drop me a line if you ever want to discuss various topics. ([email protected]) best david (css analytics)
mat kavian says:
http://arxiv.org/abs/physics/0508104 this is an awesome paper that i think explains the mirage caused by the distrubution of returns of a trend follwing strategy
John says:
There’s a reasonable amount of peer reviewed research suggesting trend following has merit so the paper you cite isn’t awesome but merely interesting.
Mat Kvn says:
It’s a very good paper, very few people have actually explicitly mentioned the distrubution of returns of MA strategies can have a mirage effect due to kurtosis of returns distrubution. I think it’s an awesome paper especially derivation of some very interested closed form solutions. I would say it’s an awesome paper- That doesn’t infer that trend following doesn’t have merit. So what is the use of your comment? This paper is awesome to me DOES NOT suggest trend following doesn’t work. If I said trend following didn’t work then your comment may have a point- but I didn’t day that did I?
Mat Kvn says:
Say that*
scott hodson says:
I think this is a matter of semantics. Covel often harps on analysts that publish price forecasts, or predict if something is going to go up or down because of some fundamental reason. However, TF traders don’t predict, they just jump on a wave that they hope will carry them a long long ways, without knowing at what point (price) they will be thrown off of the wave. I guess your main point here is that if you characterize something that is in trend, then you are predicting that the trend will continue, which is more often the case than not. I guess that is a sort of prediction, but not in the sense Covel refers to. Tom-ay-to, tom-ah-to. Feel free to correct me if I’m whack.
qusma says:
I think you’re right, it is a semantics issue. But I think it’s an important one. A prediction is a prediction, I don’t think the trend following type of prediction is any nobler than other types. Also, entering into a position without knowing when you’ll be “thrown off” is not exactly unique to trend following.
shahrzad says:
This post is very helpful. Thank. سایپا–نمایندگی سایپا
Trend Followers Make Forecasts | Price Action Lab Blog says:
[…] best answer to this question that I know off was given in the QUSMA.com blog. The author, who is a quant, called the claim that trend-following does not make predictions […]
maxwellsmart says:
Anyone who places capital at risk (with the intention of making a profit), is making a prediction of sorts. A rules based strategy accepts losses and manages risk. It’s more reactive than predictive.
All Actions In The Markets Amount To Forecasts | Price Action Lab Blog says:
[…] was proven already in QUSMA blog in response to well-known trend-follower who claimed that he just follows markets and not […]