**Market Corrections
Profile**

**July 2008**

To objectively assess whether market corrections have natural changeover points...percent changes carrying implications for emerging trends...we compile a profile of all corrections and their aftermath.

**Tabulate
the Data.** The
panel below tabulates *completed* S&P 500 corrections from 1972 to July 2008.
A correction is defined to begin with *any* decline from a historic high,
and is "completed" when the index has recovered the entire loss and
posts a new high. The top data line of the panel, for example, says there have been 125
declines of less than 1% in the study period, comprising 50.2% of all pull-backs of any
magnitude. The second data line says there were 42 declines of 1% or more
but less than 2%, and (in the final column) that drops of any size up to 2%
comprised 67.1% of all pull-backs. Thus, 32.9% of corrections were 2% or
greater.

**Downside
Continuations. ** The
"Prob Continue" column shows the mix of all declines greater than the
given (row) drop as a proportion of all equal or greater declines. For
example, in the third row, of all declines that reached the -2% to -3% range,
69.5% continued at least somewhat lower than -3% (so 30.5% stopped at no more
than -3% and recovered to new highs).

The
4th column ("Prob Continue") shows that once a correction
greater than 1% is underway, the
probability is greater than 50% for downside continuation of *some*
additional decline in every case (aside from the final, 50% drop of course).

**Strange
Attractors.** Also in the 4th column, there are two "cusps" of abnormally high
frequency of downside continuation;

Declines
of -3% to -4% show 84.2% continuation

Declines of -8% to -9% show 93.3% continuation

These two probability cusps suggest a possible "attraction to round numbers" effect as the S&P approaches declines of 5% and 10%. Such effect might result from open orders (or intentions) to buy only when the market drop has reached such points.

**Bear
Threshold.** Also notable is the "gap" of no occurrences from
-20% to -27%. This gap (and the sparseness below it) represents an
unexplained, but nevertheless apparent "bear market" effect; when an
S&P decline passes 20%, more is in store. Either it's a symptom of a
larger problem unfolding, or perhaps the accompanying media attention shifts
investor intentions. In either case, additional market setback is likely
until there is evidence to the contrary.

**Click
Threshold for analysis of S&P
decline extent after falling 20%.**