How To Filter Investor Sentiment
One of the reasons equity market participants should
be aware of fund flow data is that it often confirms or denies
investor sentiment readings. It is particularly important now as the widely
followed Investors Intelligence Sentiment Index, which measures the ratio of
bulls to bears, is currently reporting that out of the 100 advisors surveyed in
the index, 56.5% are bullish and 18.5% are bearish about the stock market.
Historically, whenever the ratio of bulls to bears exceeds 2 to1, the S&P 500
usually suffers a significant correction in price. However, fund flow
data–including that of retail and institutional flows–suggests that investor
commitment towards equities is not as euphoric as the index would have us
believe. Let me explain.
For the week ended July 9, AMG data is reporting that equity funds–including
those that invest abroad–witnessed net cash inflows of $2.4 billion, which,
admittedly, is not chump change by weekly flow standards. However, when one
considers that for the first half of 2002, equity funds took in $70 billion and
this year have only taken in $47 billion, investors are far from
over-committed. Â
More importantly, investors continue to pile into taxable bond funds–last
week to the tune of $1 billion. And, so far this year, investors have committedÂ
a record $93 billion to fixed income funds, which again, hardly suggests that
investors are overly bullish about stocks.

The lack of equity exuberance is also evident on the institutional side.
In its most recent fund flow data report, the US Federal Reserve reported
that US equities make up 39.8% of the $3.6 trillion worth of financial assets
held by private pension funds, as evidenced in the chart below. To give readers
an idea of how this measures up on a historical basis–for the past 30
years–the average equity holding as a percentage of total financial assets has
been about 43.1%. So by this measure, pension funds, like retail investors, are
underweight equities–by about $117 billion, if one uses the historical average
as a measure.
 