Testing the SOTM: The Sleepy Trend-Follower

14Jan10

Unfamiliar with the State of the Market report? The SOTM is a free daily snapshot of what some of the simple strategies we’ve talked about on this blog are saying about the market right now.

SOTM users know that we’ve stopped reporting a single next-day prediction. We now break the prediction down into three timeframes: the short, intermediate, and long-term, because there really is no one-size-fits-all answer. Different investors would find very different uses for the same report.

To illustrate, in the next two posts I’m going to test how two diametrically-opposed investors would have fared over the last couple of years using the SOTM to trade the S&P 500: the sleepy trend-follower vs the aggressive swing trader.

Sleepy Trend-Follower


[logarithmically-scaled, growth of $10,000]

In this first test I’ve assumed a trader only followed the SOTM’s long-term prediction, allocating the same % of his portfolio to the S&P 500 as indicated in the report. He’s completely ignoring the short and intermediate-term predictions.

Because our friend is particularly sleepy, he waited an entire day to execute each day’s signal, and instead executed at the following day’s close (i.e. there’s a full one day lag in execution). And he never used leverage, lest he be up all night tossing and turning.

I’ve ignored transaction costs and slippage because these results could have easily been reproduced with actively-traded mutual funds (our weapon of choice).

And for the number lovers…

Like any good trend-follower should, our sleepy friend sidestepped the entire 2008-09 bear, and captured a healthy portion of 2009 gains. He didn’t get rich along the way, but he also didn’t work that hard either – he spent a couple of minutes a day placing his trades at any point after the close.

Note: this is a very short test for such a long-term strategy. The point of this post isn’t to say “wow, look how great that backtest is”. The point is to show how the SOTM might mean different things to different investors. I’d encourage readers to dig deeper into the expanded tests we perform when we introduced each individual strategy (links available from the State of the Market report).

In our next post, I’ll look at how an aggressive swing-trader would have fared using the exact same report, but incorporating the short and intermediate-term predictions.

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.



6 Responses to “Testing the SOTM: The Sleepy Trend-Follower”

  1. 1 mike

    Regarding long term signals as referred to in the post; >50% on the long term signal would be a long signal, and <50% would be a cash/sell signal, am I correct?

    I wonder if any scaling strategy could be logical also (thinking aloud)

    Thanks for your blog, your work is always interesting and helpful to read.

    Mike

    • 2 MarketSci

      RE to Mike: thanks for the kind words sir.

      Take a look at this part of the post above:

      “In this first test I’ve assumed a trader only followed the SOTM’s long-term prediction, allocating the same % of his portfolio to the S&P 500 as indicated in the report.”

      I’m assuming a trader was long the SAME % as the report’s long-term prediction (that answers your “thinking aloud” as well because it is scaling in and out of the long position).

      Hope that helps…michael

  2. 3 Shane

    Does the long-term signal benefit from use of the abnormal market filter, or do you feel the filter is most significant at shorter intervals?

    Thanks!
    Shane

    • 4 MarketSci

      RE to Shane: very good question.

      My thinking is that longer-term strategies (like the one above) are by their nature built to trade infrequently and ignore smaller cycles in the market, so I don’t think the AMF is so appropriate.

      I think it’s much more useful w/ shorter-term strategies.

      Michael

  3. 5 Kevin

    Michael – I there someplace where the past daily SOTM reports are archived?

    • 6 MarketSci

      RE to Kevin: unfortunately no. We keep them locally, but we don’t maintain a public archive. michael


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