Declining Trend Trade Length Responsible for Recent Losses?

I much enjoy Michael Harris’ Price Action LabBlog and have listed it in the websites section here at Traders Place. I have been working my way through Michael’s articles on trend following and in particular his article “Is There Any Future in Equity Index Trend-Following?” gave me pause for thought and prompted me to do a little research of my own.

Michael states:

“The problem of trend following is fundamentally simple: choppy markets reduce the effectiveness of trend-following algorithms and shorter trend durations reduce their profitability. There is no need for a more sophisticated analysis. “

“An example of a trend following system that shows how recent conditions have affected the robustness and efficiency of such methods is the 50-200 MA cross system. Basically, this is a long/short symmetric system that establishes long positions when the 50-day MA crosses above the 200-day MA and exits and reverses when the opposite happens. This system has worked very well in SPY since 1994 and even during 2008 with a return close to 34%. But after 2009 this system stopped performing well because, as will shall see via the help of a special indicator I have developed, the market dynamics changed drastically for a period of almost two year. During that period the system was subjected to a large drawdown of about -38%.”

One of the points raised and shown in the chart was as follows:

“The bottom pane shows the downtrend in the duration between two consecutive MA crosses of the system and the downtrend is clear.”

It sounds logical enough to me. If the markets have become “choppy” and have recently moved sideways more often than they have trended, then surely this would show up in shorter trade lengths over an entire portfolio and not just in the S&P?

In my “Trend Efficiency Index” I showed that trends are noisier now – there much retracement on the journey from point A to point B. Again, surely one result of this should be that trades in general have become shorter in length in recent years, not only in the S&P but also across the board?

I took Michael’s work a few stages further. I took back adjusted price series for a portfolio of 107 futures contracts: 14 bonds, 10 currencies, 8 energy contracts, 13 grain contracts, 8 short term interest rates, 13 metals, 4 “other”, 10 softs, 24 stock indices. No attempt therefore to balance the portfolio. I tested with single contracts rather than fixed fractional position sizing and no risk constraints – each trade was taken unless it was a locked limit day in the relevant market (in which case the trade was taken on the next available day). The system was a simple dual moving average identical to that Michael used, taking both long and short trades: a reversal system. I tested in the ratio of 1:4 SMA:LMA. Beginning at “5 and 20”and ending at “50 and 200”. For completeness sake a also tried a couple of other variations: “25 and 200”; “10 and 200”.

I tested for the period 1st January 1970 to date. I made no allowance for the fact that during that time, the number of contracts available increased dramatically – for these purposes it should not matter.

To my surprise, a visual examination of the charts showed little difference in trade length over the years.

But.............the addition of trend lines to the charts told a different and more complex story.

Anthony FJ Garner

 

there was some research article I read a while back that showed the markets still trend as much as they did 40 years ago with the same win % but the magnitude of the trend has decreased substantially.

the effect of this is that a decrease in the magnitude of the trend by 40 % reduces profitability by as much as 80% hence the winners can't pay for the losers.

 

got the PDF

 

Thanks, excellent. I'm doing the same sort of research myself. Also see the Estlander report which said much the same sort of thing re "magnitude".

 

Interesting to look at that report written in 2001 and note how prescient or otherwise the author was.

 

yep....I saw this a few years ago when I was researching on the viability of classical trendfollowing and I got convinced it was becoming less profitable.

 

See Baltas - Kosowski, too
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1968996

 

Thank you - and bang up to date. I look forward to reading it and hope that I can understand it!

 

can jump to summary...

"Our results also imply that there are no statistically significant capacity constraints momentum strategies,

but this leaves a separate important question unanswered, namely, why CTA performance has been lacklustre during the recent period 2009-2011. In principle, there can be three main reasons for poor
performance:

(a) capacity constraints (which our results do not support),
(b) absence of sufficient pricetrends for individual securities,
(c) increased correlation between futures markets, which reduces diversification benefits.

Anecdotal evidence suggests that monetary/fiscal policy uncertainty in the 2009-2011 period has been particularly high, which may have driven market sentiment and could explain why price trends during this period tended to reverse more often than in the past."

I've seen other literature as well, regarding the unusual recent increase in correlation across markets .