Wednesday, February 28, 2007

High Momentum Stock Market Declines: What Comes Next

My recent post took a look at what typically follows a single day of large price declines. Here is a very different take on near-term prospects following a high momentum decline. Recall that one of my core trading principles is that a trending move with strong momentum tends to persist in the short run. Let's see if that might apply to the current market.

My best measure of stock momentum is the proprietary Demand/Supply index that I track daily on the Trading Psychology Weblog. Demand, you might recall, is an index of the number of stocks closing above the volatility envelopes surrounding their short- and intermediate-term moving averages. Supply is a similar index, tracking the number of issues closing below the volatility envelopes surrounding both their moving averages.

Demand and Supply follow every operating company trading on the NYSE, NASDAQ, and ASE. They tell you the relative number of issues closing either very strong or very weak in the equity markets. I have collected the Supply and Demand data since September, 2002. Tuesday's market hit the highest level of Supply recorded over that period: 420. We have only had seven other days with Supply levels above 200. The next day in SPY, five of the seven days finished lower in price, by an average of -.38%. Five days later, six of the seven occasions finished lower, by an average of -1.06%. These are much weaker results than the average market performance over that period.

Of the seven sessions in which we had very strong downside momentum, six traded below their day's lows during the following session. Indeed, four of the seven broke those lows by more than ten full ES points. Weakness, then, tends to carry over to the next day following a high momentum decline.

Interestingly, when we had days in which downside momentum was strong but not extreme (Supply greater than 1500, but less than 2000; N = 24), the next five days in SPY averaged a healthy gain of .67% (15 up, 9 down). By contrast, the average five-day gain over the entire period was .24%.

In sum, when we have large declines on downside momentum that is extremely broad--as we had Tuesday--we tend to see followthrough to the downside over the short term. When large declines are not so extreme in the breadth of their negative momentum, we tend to see reversals over the short term. It is for this reason, in spite of the small sample size from my data, that I will be open to selling strength early on Wednesday in anticipation of breaking Tuesday's lows. Note, however, that as I'm writing, we have bounced significantly higher in the overnight stock index futures. A recent comment on my previous post from Trading Nerd took the data back to 1930 (!) and found evidence of positive returns after a 3% decline, but that most of those occurred the next trading day. Such a pattern--sharp bounce, then some follow through decline--would help to connect the results from my previous post and this one.


Johan said...

Both this post and Trader Nerds post indicate that there is good to sell todays close and hold for 4-9 days. Wouldn't that be the ideal trade with the info we have instead of fading todays strenght? But I guess that's up to the trader which time horizon one seems more comfortable with.

Brett Steenbarger, Ph.D. said...

Hi Johan,

I think the general strategy would be to wait for strength to come into the market as buyers take advantage of low prices and nervous shorts cover, see how much volume and strength are associated with the bounce, and then--if the bounce is unconvincing and the buying wanes--fade it for a possible test of the Tuesday lows. If subsequent selling cannot take us to those Tuesday lows, I'd expect further upside from the bulls. Tracking the markets in Europe should provide leads.


b hong said...

And so, it appears, that the concluding statement is:
"strength begets strength and weakness begets weakness." But, ultimately, it is not a mechanical trade. "Past results does not guarantee future performance." The individual trader must still think... AND REACT.

Brett Steenbarger, Ph.D. said...

Yes, indeed, Bruce; great point. Historical tendencies are only that, not guarantees. What matters most is how large traders are behaving *now*. You'll note that, in my morning comments, my view of the market shifted as we saw selling dry up (the anticipated weakness following weakness) and buying enter (bounce following a large decline). Thanks for your comments--


Pete said...

Brett: Very nice research. From a phychological stand point, damage has been done. Traders who were milking the last few points out of their long positions will be much more anxious to take profits. Yesterday's deline reminds me of 1998. Then, earnings were great, the internet was starting to boom and the valuations were resonable (SPY was 45% below its 2000 high). Then out of no where, there was an arb trade that went sour. Initially everyone thought... so what, it's a hedge fund. The stock I liked yesterday is the same today, except it is 3% cheaper. That sell off was created by two of the founders of modern day option pricing models (Robert Merton and Myron Scholes). They started a hedge fund called Long-Term Capital Management. They used arbitrage models and the fund collapsed in 1998 when Russia defaulted on their debt. Here is a link for more info. That was only one part of their model, but it created a domino affect. The sell off in 1998 was huge and the SPY fell 25% in a month. It was nasty and the market was up big going into the collapse. The Yen-Carry Trade is in trouble if: the Yen rises, Japanese interest rates continue to move up or traders using the strategy get nervous. There is still room in the spread, but as it narrows, the pressure will increase and traders will start to unwind it. Borrow cheap, leverage big. There are more hedge funds than ever and someone is going to fold. Retail margin debits as a percentage of the account balance are at 2000 levels. I sense a shakeout.

Brett Steenbarger, Ph.D. said...

Hi Pete,

Great thoughts; thanks. Your scenario is certainly one I'm keeping in mind, although I would have chosen 1997's October scary drop and not 1998 as my analogue. That having been said, I suspect we'd have quite a few nervous and panicky longs should we take out Tuesday's lows in the large cap indices.


Brandon Wilhite said...

My thoughts on the carry trade are this:

Even if we bottom out in JPY and CHF and then move higher...well now we're at a point where the risk is greater than it was pre-tuesday. So even if we gain some more over there, the next one will most likely be much worse because I believe traders got spooked yesterday. Therefore, it's very likely the beginning of the end for the carry trade. How will this affect other markets? Probably not in a good way for Average Joe investor. On reflection they are the ones who got stomped on yesterday.


Brett Steenbarger, Ph.D. said...

Hi Brandon,

Very interesting thoughts; thanks. Related to your idea, I wonder if we've put in a long term bottom in VIX. Going forward, the options may price in more risk than they had previously--