Friday, May 18, 2012

The Epic Momentum Bubble

Out of curiosity, I thought up a simplistic momentum system today and opted to backtest it.


Click to enlarge.

Here are the rules.

1. If the S&P 500 was up yesterday then it will be up today too. Go long!
2. If the S&P 500 was down yesterday then it will be down today too. Short it!
3. If the S&P 500 was unchanged, do nothing.

This system starts in 1957 with $1. The chart above shows how the system would have worked. Amazingly well, wouldn't you say? Well, sort of. I must admit that the last decade was a disaster. And to be fair, it really didn't work all that well from 1980 on (compared to buy and hold).

Keep in mind that this system shouldn't technically care if we are in a bull market or a bear market. It bets both ways, like a hedge fund.

Unfortunately, this system fell apart starting in April of 2000. Remember the wisdom of Suze Orman that very month? Reminds me a bit of a "sure thing" system I once tried. It worked great until it didn't.

Check out one of the more amusing losing streaks as momentum day traders apparently went head to head with high frequency trading algorithms.

2002-10-18 884.39
2002-10-21 899.72
2002-10-22 890.16 Long! 1.1% loss!
2002-10-23 896.14 Short! 0.7% loss!
2002-10-24 882.50 Long! 1.5% loss!
2002-10-25 897.65 Short! 1.7% loss!
2002-10-28 890.23 Long! 0.8% loss!
2002-10-29 882.15 Short! 0.9% gain!
2002-10-30 890.71 Short! 1.0% loss!
2002-10-31 885.76 Long! 0.6% loss!
2002-11-01 900.96 Short! 1.7% loss!

For example, the system goes long after seeing 884.39 become 899.72 on October 21st. On October 22nd, it was still long at 890.16, lost, and after seeing a day of losses switches to shorting it. And so on.

That isn't the only way investors lost though. As I sift through the data I see lots of small gains with a big loss every now and then, almost like little carrots are being handed out to get investors near a whack-a-mole hammer. That's a subjective painting the tape opinion at best.

Also note the behavior change in the chart over the past few years. See that flat line? Let me zoom in a bit.


Click to enlarge.

Hello arbitrage.

That's probably indicative of high frequency trading algorithms going head to head with competing high frequency trading algorithms. May the best computer win! Perhaps Jamie Dimon should think about upgrading. The hunter has apparently become the hunted.

And one wonders why I am not content to play in the casino? Seriously? The less I trade the better. I have absolutely no desire to compete with that. The lowest-lying fruit is gone, and probably half the tree with it!

Source Data:
St. Louis Fed: S&P 500

1 comment:

Stagflationary Mark said...

Troy,

We're not playing by the rules here at the farm. Are we? Are we? Sonny? - James Clayton, The Recruit (2003)