Sell-off's Low Volume Not Necessarily A Positive

08/26/08 9:15 AM EST

Good Tuesday morning...We begin the day with flat pre-market futures.  The Dollar is strong, Oil is fairly flat and foreign equities were mixed to slightly negative.  We have a bevy of economic data due this week, including this morning at 10:00am which could help sway this fickle, light-volume tape.

A couple of weeks ago, we went over the probability that volume this week and last would be about the lowest of the year, or very close to it other than immediately surrounding the major holidays like Thanksgiving, Christmas and New Years.

The market is not disappointing us in that regard, as yesterday's session sported the lowest volume of the year according to my data sources (using exchange volume only, not necessarily composite volume).

So while the pathetic volume is no surprise, it is a bit unusual to see such a paucity of trading activity on a day the indices lost so much of their value.  I went back over the past 43 years to try to find any other time the S&P 500 sold off nearly 2% or more and NYSE volume was the lightest in six months - and I couldn't find one.  Yesterday was the lightest-volume big down day in four decades.

If we relax the parameters a bit and look for 1% or greater losses on the lowest volume in six months, then there were four occurrences.  The last one was in August 1999, and curiously the S&P sold off the following day all four times by at least 1% each time.  Even by three trading days later, the S&P was still negative each time and the average loss climbed to -2.4%.

That's disturbing, but we can't read too much into four occurrences, so let's relax the parameters again and check for a 1% or greater loss on the lightest volume in three months.  That's a little better, since we get 11 occurrences, but the results weren't better at all.  The next day, the S&P was positive only 1 time and the overall average return was a hefty -1.2%.  The most the S&P was able to climb at any point during the next session averaged a miniscule +0.3% while the most it fell averaged -1.9%.  Three days later, the S&P was positive only 3 times and the average return fell to -1.3%.

That's not good, especially when combined with the other negatives we've gone over since late last week -  a spike in positions in VIX futures, potential trouble for the US Dollar (which has had a high correlation to the S&P), broken uptrend lines in the S&P and DJIA, an inability to rally well from short-term oversold conditions, and a market that has already met its typical bear market rally objectives from the panic readings on July 15th.

Mitigating the negatives somewhat is the simple fact that we just sold off 2% in a day.  When that occurs in the days leading up to an exchange holiday, the market has typically held up well heading into the day immediately before the holiday as we discussed yesterday.  Another positive is that several of our shorter-term guides are now oversold, though as I mentioned last week I'm no longer all that enthused about trying to buy into oversold conditions given the negatives outlined in the paragraph above.

So quite frankly, I'm not sure what to expect here.  The low-volume sell-off stat from above has been consistently negative, but the oversold conditions and pre-holiday trading tendencies are quite positive, and with such low volume it's not going to take much to move us either way.  With these conflicting signals, I'm keying off the 1260ish area on the S&P 500 (last week's lows), giving the bullish tendencies the benefit of the doubt above that area, and the bearish ones below.  We can't rule out a false breakdown either, where we dip below 1260 enough to flush a flurry of sell stop orders resting just underneath, then recover back above.  Bottom line, it looks like anything goes here and I'm not trading it.