Tuesday, June 06, 2006

S&P Momentum Effects Are Not Dead

Let's say you have a very large basket of operating company stocks. You calculate a short-term moving average for each issue in the basket and a 2 standard deviation envelope around that moving average. We will call the number of stocks closing above their upper band Demand. The number closing below their lower band will be called Supply. When Demand is high, it means that many stocks are displaying very strong short-term price momentum. When Supply is high, it signifies the reverse. Observe that this is not your usual momentum measure: Demand and Supply are measuring breadth as well as extent of momentum and are setting relative momentum criteria for each stock based upon its volatility.

(Note: A version of Demand/Supply is posted daily to the Trading Psychology Weblog.)

Since March, 2003 (N = 816 trading days), when the Demand reading is above 600 (N = 58), the next three days in SPY average a gain of .61% (45 up, 13 down). That is much stronger than the average three-day SPY gain of .16% (473 up, 343 down) for the entire sample. Broad, strong upside momentum thus tends to follow through with strength over the near term.

During that same time, when the Supply reading has been above 700 (N = 37), the next two days in SPY have averaged a loss of -.01% (18 up, 19 down). That is weaker than the average two-day gain of .11% (442 up, 374 down) for the entire sample. As we saw in the previous post, broad downside momentum tends to follow through with weakness over the near term.

When the difference between Demand and Supply is greater than 500 on a single day (N = 40), the next three days in SPY average a gain of .33% (30 up, 10 down), much stronger than average.

When the difference between Demand and Supply is less than -500 (N = 53), the next two days in SPY average a loss of -.04% (23 up, 30 down)--much weaker than average.

The bottom line is that momentum effects in the market are not dead. They are simply mediated by the breadth--not just the extent--of the momentum.

P.S. - Here's the link if you want your own digital church sign.


baifriend said...

You say 500 means 500 stocks in the whole stock markets? I remember there are more than 8,000 stocks in the whole market.

Greg Lepiaf said...

Which time scale do you use in the Demand/Supply evaluation ? And do you find useful to use two or three different time scales (eg 1 hour - 1 day - 1 week) with some appropriate weighting ?

Once again thank you for all these very stimulating posts.

Brett Steenbarger, Ph.D. said...


Thanks for the question. Notice in the first sentence of my post, I am using "a very large basket of operating company stocks". I am not collecting data on every listed issue, because many issues are preferred stocks, closed ended funds, etc. that don't reflect the broad market. 500 was my cutoff for strong and weak days.


Brett Steenbarger, Ph.D. said...

Hi Greg,

Your idea of different time scales is very interesting. At this juncture I am using end-of-day data only. I do suspect an intraday version of this indicator would work quite well.