We've had quite a selloff today; I will post the indicators in tonight's Weblog and will also post relevant historical patterns right here tomorrow AM before the open.
In the last two posts, I have looked at Institutional Buying (defined as the tendency of buyers to lift offers among actively traded institutional favorite stocks) vs. Institutional Selling (defined as the tendency of sellers to hit bids among those same stocks). The Institutional Composite is a weighted index of these two.
Since July, 2003 (N = 732 trading days), we have seen 36 occasions in which the Composite has hit a level of +500 or greater. The next day, SPY has been down by an average of -.15% (15 up, 21 down)--much weaker than normal.
Conversely, when the Composite has been -500 or lower (N = 46), SPY has been up the next day by an average of .20% (28 up, 18 down)--and by a whopping .56% three days out (34 up, 12 down).
When we take price out of the equation, however, we can see that strength in the Composite on strong up days in SPY leads to more favorable returns three days out than if strength in SPY is accompanied by weak Composite readings. When SPY is up by .60% or more (N = 142) and the Composite is strong (N = 71), the market is up by an average of .17% three days later (48 up, 23 down). When SPY is up by .60% or more and the Composite is weak (N = 71), the market is down by an average of -.04% (36 up, 35 down).
What this means is that when there is net buying pressure (a strong Composite) but price is not strong, the short-term outlook is subnormal. When strong price action is accompanied by strong buying pressure, there is greater likelihood that the rise will continue. The Composite thus must be evaluated relative to the price movement it is associated with, not analyzed in isolation.
When SPY is down sharply on the day *and* we have a Composite reading of -500 or less (N = 27), the market is up three days later by an eye-popping .58% (21 up, 6 down). That is much stronger than the average three day gain in SPY of .17% for the remainder of the sample of weak SPY days.
If we summarize the last three posts, it looks like Institutional Buying and Institutional Selling are worth tracking as separate market variables. Dividing volume into action at the offer vs. bid appears to give very worthwhile information for short-term traders. The balance between the two, as captured by the Composite, appears to be most important at the extremes: when traders are dominantly hitting bids or lifting offers. I will be tracking these extremes going forward in the Weblog.