|
REPORTING DELAYS:
One week
EXPLANATION:
Hedge funds, particularly managed
futures (CTA) funds are aggressive investors and traders. They
go long or short, and they cross asset classes readily.
Most important, they are consistent
trend-followers. When trend-following goes too far - gets
overdone - the contrarian principles come into play. Research
shows that when CTA funds go to extreme exposures in aggregate,
a trend reversal is getting near.
Carpenter Analytix tracks CTA fund
exposures via analysis of their returns streams. Like any
portfolio, hedge fund returns "behave" statistically like the
behavior of their component parts. In tracking the evolving
returns behavior, we reverse-engineer the stats to see what
kinds of components are driving the returns.
GUIDELINES:
We treat this data as we do most
sentiment indicators, which is in a contrary fashion.
While hedge funds are supposedly the "smart money", and
undoubtedly many of them are on an individual level, when
considered in aggregate they work like any other large group.
Because of that tendency, we want
to look for extremes in the exposure data. When fund
managers are extremely net long, we want to be on the lookout
for a market correction, since hedge funds would be one less
source of potential buying power if they're already heavily
exposed.
When fund managers are not very
invested on the long side, or particularly if they are net
short, then we should be on the lookout for a market rally.
Funds are an excellent source of "short covering" stampedes when
they're exposed to the short side and the market starts to
rally, so when they are in that position, we should be careful
about being too aggressive on the short side ourselves.
While the bear-market history of
this data is limited, we should be aware that hedge funds were
net short for all of the 2000 - 2002 time period. While
that seems "smart" on their part, and it was, we should look at
the data on a relative basis. Funds were their least net
short near market peaks during that time, and their most net
short near intermediate-term lows. That's not very smart,
and why we still consider this to be a contrary indicator.
You may notice that the indicator
goes "off the chart" to the downside in September 2001.
Due to the difficulty of obtaining reliable pricing data around
the 9/11 tragedy, the indicator values around that time may be
somewhat compromised. Because of that, we truncated the
chart around that time so that it does not meaningfully skew the
other data points.
ADDITIONAL RESOURCES:
Carpenter Analytix (www.CarpenterAnalytix.com) |