Test of Ablin Trend Following Strategy

18Nov08

This is a quick test of a Jack Ablin strategy shared over at the Big Picture.  To be fair, I can’t find where Ablin originally discussed this strategy, so I don’t know what context he put it in, but the rules appear to be to go long the S&P 500 when it closes more than 5% above its 200-day moving average and hold until it breaks 5% below its 200-day MA.  See graphic from Ablin (click to zoom):

2008111801

For comparison’s sake, I’ve also included a simpler strategy in this test: buy when the S&P 500 crosses over its 200-day MA and sell when it crosses below. 

The results for Ablin’s strategy (red) and the simpler version (green) trading the S&P 500 frictionless and with no return on cash from 1951:

2008111802
[logarithmically-scaled]

And for the number lovers:

2008111803

With the exception of the very narrow window of time from Ablin’s original graphic the super simple strategy outperformed both Ablin’s variation and the broader market. And this is a good example of the problem with drawing conclusions from just part of the story. 

Which one is going to be better tomorrow?  I have no idea.  But I have a sneaking suspicion that the 5% bands were added because without them the strategy would have entered and exited the market prematurely quite a few times over the last 10 years (and that’s never good for impressing people).  But armed with the bigger picture over the last 50+ years, I would guess that some folks would wager that the simpler strategy stands a better chance of prevailing in the future.

Just my $0.02.

Happy Trading,
ms

 

To stay up to date with what’s happening at the MarketSci Blog, we recommend subscribing to our RSS Feed or Email Feed.



8 Responses to “Test of Ablin Trend Following Strategy”

  1. 1 Sam

    I love your blog and read it daily! I know you mentioned running a report on Mebane Faber’s strategy which is similar (using 10 Month MA) but doing it across asset classes ( Bonds, Commodities, US Equities, Intl Equities, REITS.) This seems to equal market performance with significantly less volatility. With 2X leverage you are getting about equal volatility to the market and 1.6 to 1.8 X the return. Interested to hear your thoughts. It did make me think of the original Market-Sci strategy which was very diverse. Might there be some benefits to applying these timing strategies to mutliple asset classes to improve the probability of them working more often than not and providing diversification. (YK Included)

  2. 2 marketsci

    RE to Sam: THANKS for the kind words. A test of Mebane’s strategy has been bumped up a few notches on the to-do list (promise, one day I’ll dig into it). It’s not that I in any way don’t trust Mebane’s numbers mind you, it’s just nice I think to see a strategy from someone else’s perspective.

    RE: trading multiple asset classes. In short, yes, great idea. In long, it’s more difficult than it looks. Individual sectors (including REITs) and international equities all adhere to the “in a crises, correlation moves to 1″ rule I feel and that’s why we’ve pared down our exposure to sectors in the MarketSci portfolio. I would love to add a bond angle to YK (and MarketSci), but unfortunatley, I’ve never found anything good enough in the very short time frames we trade in (trust me, I’ve done a lot of looking).

    Thanks again for the comment Sam and more to follow (some day) on Mebane’s strategy.

    Thanks,
    michael

  3. 3 eber terandst

    Excellent site !.
    I used to click on this link only ocasionally, but now it is the very first thing I do when starting the day (well, after reading the comics).
    EB

  4. 4 ducati998

    ms,

    With regard to Bonds [Notes, Bills]

    What maturity would you want to trade?
    Treasury, Corporate, Mortgage or Municipality?
    How liquid? [I'm guessing very liquid]

    Convertibles?
    Preferreds?
    Swaps?

    jog on
    duc

  5. 5 Sam

    Yes the correlations of everything but bonds has moved very close to 1. I think the multiple asset allocation strategies will work in most of the non-crisis markets. It may be that other than getting lucky, cash and short term bonds may be the closest to an alpha strategy available. What the multiple asset classes bought you in this down term was the ability to ride each class close to its top before getting out.

  6. 6 marketsci

    RE to eber: that’s the best thumbs up I’ve heard all month. Many thanks. michael

  7. 7 marketsci

    RE to duc: I’m assuming you’re referring to my comment to Sam that I would like to add non-equity elements to YK/MarketSci. In my case, I trade leveraged mutual funds (from Rydex, ProFunds, and Direxion), so my choices are limited. I’d most like something good for 30 year UST long/short, US dollar long/short, and possibly the commodity indexes. jog on.

  8. 8 ducati998

    ms,

    Ok, so you are looking for an entry/exit trigger?

    jog on
    duc


Leave a Reply

Please log in using one of these methods to post your comment:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s