



|
 |
by Howard Simons, NQLX's Special Academic Advisor.
Alpha can be defined as the net risk-adjusted premium
return from a position or a portfolio. Its calculated by subtracting
the portfolios return from the assets return, and can be conceived
of as the average non-systematic deviation of a stock relative to an index.
As an example, we can regress the daily returns for IBM against those
of the S&P 500 over the December 1997 December 2001 period
and find an alpha of .07% per day for IBM. The beta, or relative volatility
of IBM over this period was 1.0495. In the graph below, alpha is where
the red trend line intercepts the Y-axis, and beta is the slope of the
red line.
Since both of these measures are based on statistical
relationships, they may be unstable over time. For example, if we shrank
the estimation period for IBM to a December 1998 start, alpha would change
to .06% and beta to 1.0855. These time-dependent coefficients may represent
one of the biggest adjustments futures traders will have to make when
they start trading single stock futures (SSFs). Since commodities do not
by definition change over time, intermarket spreads such as wheat-corn
tend to be far more stable than the matched-pair spreads that will be
common in the SSF world. Of course, this situation does nothing other
than create a trading opportunity.
Money managers ultimately earn their pay relative
to the S&P 500 or another benchmark by adding or subtracting alpha.
Heres where SSFs could be used. Lets say you want to delete
a particular stock from an index, one that has a negative alpha. To sell
this stock short in the cash market is cumbersome and expensive. Its
far easier and cheaper to delete it from your portfolio by selling the
SSF.
You can even do this in a market neutral manner if you
so desire, even though youll start to deal in institutional sizes.
Lets say IBMs weight in the S&P 500 is 2.039%, and that
the dollar value of the S&P Depository Receipt (SPDR)
is $11,430 per round lot. This gives you an IBM exposure of .02039 * $11,430,
or $233. If IBM is trading at $121.34, each IBM SSF will be worth $12,134
before basis adjustment. A fund manager could then delete IBM by selling
one SSF against every 5,200 shares of SPDRs and have no exposure to the
markets overall direction.
|
 |
|