The Drop Better Stop Soon

08/19/08 3:50 PM EST

As of: SPX 1251 HELP  ARCHIVE

This morning we went over the idea that given the large gap down opening, if the S&P went on to make lower intraday lows after the first hour of trading, then we should expect to see even more of a sell-off the remainder of the day, with only a very minor chance of a later-day upside reversal.

That has played out so far, with the indices showing a day just about as bad as yesterday.  The S&P and DJIA have clearly broken their uptrend lines from the July low, which nearly every technical trader is watching.

That could perhaps be a good thing if it ushers in some exhaustive selling pressure, and there is some evidence of that here given the readings in some of our shorter-term guides.

It's also somewhat encouraging that we've gotten these days out of the way - since 1965, there have been 17 instances of the S&P selling off 1% or more on a Monday and Tuesday, accompanied by an Up Issues Ratio below 30% both days (meaning that of all stocks traded on the NYSE, fewer than 30% closed higher each day).  The S&P 500 cash index isn't quite down 1%, but the futures and ETFs based on the index are, due to different closing times.

The following day, the S&P rose 12 of the 17 times, averaging +1.0%.  Since 1982, the S&P went 8 for 8 in next-day rebounds, and also in showing positive returns over the rest of the week (averaging +2.4%).  If we take away the parameter that it was Tuesday, then since '65 the next-day winning percentage drops to 55% and since '82 it drops to a still-respectable 73%.

If we look at SPY, the S&P 500 tracking fund (which only has history back to 1993), then anytime we saw a Monday and Tuesday with a 30% or less Up Issues Ratio, the fund rallied the next day 9 out of 10 times by an average of +1.6%.  Only one of those occurred during the 2000 - 2002 bear market, however.

The pressure is finally starting to get to options traders, who have looked relatively calm over the past couple of weeks.  The Equity-only Put/Call Ratio from the CBOE looks like it will close near 0.90 barring a last minute surge in call volume, which would be the biggest display of (potential) fear since the July low.  I'm not reading too much into it until I see the closing figures, and besides I don't usually read too much into a single day's reading.  Plus, the intraday ratio from the ISE options exchange is not confirming an excessive amount of protective put buying today.

Bottom line, we're oversold enough on a number of different measures after the past two days, which typically means at least a short-term bounce even during bear markets (except the wicked 1973-1974 period).  Technically, indices like the S&P and DJIA don't look all that great at this point, and we could be just at the beginning of a push towards the July lows.  Given what we went over on July 15th and in the subsequent few days, I don't think we're at that point, but the failure of the breakout to hold over the past two days is a definite warning sign.  I'm willing to continue to entertain the thought of pressing long-side bets into oversold conditions as long as we remain over 1250ish on the S&P - a drop and hold below there, and all bets are off.

Mid-Week Reversal About To Be Tested

08/19/08 9:10 AM EST

Good Tuesday morning...We begin the day with a major gap down in the major equity index futures, as an early morning slide turned into a waterfall after worse-than-expected reports on inflation and the economy.  Commodities are also getting hit across the board, and foreign markets were exceptionally weak with 2% losses common.

Yesterday afternoon we went over some scenarios related to the large down day, post-option expirations (and Mondays in general) and low volume.  Basically what we saw was that the market tended to bounce back mid-week from severe Monday declines (whether or not it was following an option expiration), but the low-volume conditions didn't add any extra edge.

I've been asked a few times whether being in a bull or bear market made any difference in the results.  So let's re-run the study, this time looking for 1% or larger declines on the lowest volume in a month, while the S&P 500 is in a bull or bear market.  For these purposes, "bull market' is defined as an upward-sloping 200-day moving average; vice-versa for "bear market".

During bull markets, we've seen this 18 times since 1965, with the S&P bouncing back the next day 50% of the time and 56% of the time by three days later.  Average returns were positive, but just barely.  During bear markets, we've encountered this situation 22 times and the S&P bounced back the following day only 23% of the time averaging -0.8% and three days later 36% of the time averaging -1.0%, so there was a fairly large negative edge.

I always like to check to see if these kinds of tests are consistent over the decades - a lot has changed in the past 20 years in terms of who is trading the market, and how they do it.  Out of the 22 instances occurring during bear markets, 13 of them took place prior to 1990 (the majority in 1973 and 1974) - and every single one led to more of a sell-off the next day, averaging a wicked -1.3%.

Since 1990, though, there have been 9 occurrences and the S&P was up the next day 5 times with an overall (slight) positive average return.  Three days later, the S&P was up 6 times averaging +0.7%, so there was a pretty major difference over the past nearly 20 years compared to that prior.

The last time the S&P sold off as much as it did yesterday (at least 1.25%), then gapped down as much as it is indicated to this morning (at least -0.75%) was March 17th, the day of the spring low.  Over the history of the S&P 500 tracking fund, SPY, there have been 36 such instances and the index ended up closing higher than the open 67% of the time by an average of +0.9%.  By the next day's close, that worked up to 70% of the time and an average of +1.4% before the consistency and average return began to peter out, so there was a tendency to see very short-term relief bounces from extreme short-term selling pressure like we're seeing here.

If we stipulate that it's Tuesday, then we're only left with three occurrences.  If we relax the parameters to a 1% decline yesterday and 0.5% gap down this morning, then we get 7 instances and buying the open on Tuesday and holding through Wednesday's close led to 6 winning trades out of the 7, averaging +2.5%.  It was volatile, with an average maximum loss (i.e. the most the market went down) during the next two sessions of -1.0%, compared to an average maximum reward (i.e. the most the market went up) of +3.6%.  There was still a small sample size, but there was good positive consistency and a largely skewed reward/risk ratio.

I mentioned yesterday that an argument could be made that we were at an "oversold on support" setup as our short-term guides for the S&P had cycled into extreme territory and the index was sitting right on its uptrend line from the mid-July low.  My thought was that allowing for some follow-through weakness this morning, we should be able to rally into this afternoon and tomorrow, or we'll likely have a date with the 1250ish area on the S&P before too long.

This morning's decline should take us to 1265ish, which was a minor low from about a week and a half ago and I'm going to want to see a bounce almost right from the open.  If we go on to set lower intraday lows after the first hour of trading, then our chances of a dramatic upside reversal later in the day drop dramatically, and I'd expect to see 1250 this week.