John Hussman wrote an anti-trend following piece yesterday. My views:
1. Hussman puts forth sensibilities on the economy that I often share. He sees the issues clearly. In fact, I would love to get him on my podcast.
3. Those who have not read my books will accuse me of being lazy by not responding here online. Patience. That is coming, but for now I wanted to acknowledge Hussman’s views.
I posted the above retort late last night and this morning no more comment was needed as John Hussman had responded directly via email:
Saw your quick post this morning relating to my weekly comment. I think I’ve unintentionally drawn fire because my key point may not be clear. My argument is not a criticism of trend-following in a general sense:
“Does all of this suggest that trend-following measures should be ignored? Not at all”
The point of the piece was to discuss the difficulty with simple moving-average crossover systems, which many investors seem to think are wildly effective in and of themselves. We periodically get questions of the form “the S&P 500 is above its 200-day moving average” as if this is a sufficient and obvious basis for hedging decisions. It is that question that the piece addresses.
The second part of the piece, on multiple sensors, discusses factors that I believe are useful in building a trend-following method. In particular, it focuses on using a variety of confirming evidence, rather than a single moving-average. We certainly do use trend-following factors in our own approach, but they are not limited the simple MA crossover variety.
From a broader perspective, valuations, trend-following measures, market internals, economic factors, sentiment, and other factors are all important in our investment analysis. It may seem that we don’t rely on trend-following considerations given our “miss” in 2009-early 2010. But our defensiveness in that period was due to my insistence on stress-testing our methods against out-of-sample Depression era data, which I’ve regularly discussed in our weekly comments. The ensemble methods that satisfy our validation requirements would have indicated a constructive position during much of the 2009-early 2010 period, supported by trend-following considerations, but there are certainly periods – particularly today – when our stance will differ from the stance that a pure trend-following approach would advocate.
We’re probably actually much more in agreement in terms of investment philosophy than your brief post indicates. Again, the piece was not an attack on trend-following itself, but was focused on simple MA crossover strategies that investors imagine to be wildly profitable and applicable without further analysis. If you re-read the opening paragraph of the piece, I think that intent is right up front.
Hope that clarifies things.
John P. Hussman,
Ph.D. Hussman Strategic Advisors
Thanks John. Now that we are connected I have to angle him onto the podcast!