Short positions throwing a wrench in the works for traditional (long-only) institutional investors

17 July 2008

The addition of short positions to traditional long-only portfolios adds a whole new dimension of complexity and requisite analytics.  This was true for mutual funds (see related postings) and is even more relevant for traditional (long-only) institutional investors.

We stumbled across this article by BNY Mellon’s fund administration arm that reveals some of this complexity.  The paper argues in favour of a holistic approach to portfolio risk analytics - an approach not always followed by institutional investors according to the firm:

“Many institutional clients, lured by the promise of additional returns at a risk discount, are jumping into these more complex alpha-generating strategies — many for the first time and many without reviewing the exposures and risks. Additionally, many of these short-enabled funds don’t have a long history of incorporating shorts into the overall portfolio strategy and may be introducing unintentional risks. In fact, many funds are outsourcing the short portion, resulting in two asset management teams working separately without understanding the total account makeup.

“…When aggregating and viewing a short-enabled account’s structure and fundamental makeup, we must combine the long and short market values instead of looking at the long and short portions separately.”

In essence, BNY Mellon is following the recommendations of Bruce Jacobs and Kenneth Levy who said in this article and others that you shouldn’t analyze a long/short portfolio as separate long and short portfolios since a combined portfolio will always be more optimal (i.e. have a higher risk/reward ratio).

The authors of this article use a simple example they call the “Russell 2000 Argument” to illustrate the analytics of the same idea.  They compare two rudimentary portfolios - one long the Russell 1000 and one long the Russell 3000 and short the Russell 2000 (stocks 1001-3000 of the R3000). 

While both funds are essentially identical from an exposure standpoint, they differ on several common measures, namely:

• Counts and medians – The R3000L/R1000S and R2000 strategies invest in a different number of securities to achieve the same risk profile. 
• Arithmetic Averages – Straight averages do not differentiate between long and short positions. Weighted averages do differentiate long and short positions and the results between the two investment approaches will equal.
• Liquidity measures – The liquidity figures within the profile report shown below don’t utilize the absolute methodology highlighted within this paper.

While this may seem kind of obvious using this simple example, you can see how these same issues can throw a hedge fund administrator or risk manager into a tizzy when things become more complex.  The paper goes through a fictitious example of a market neutral fund to show how exposure figures can get wildly out of whack when the net exposure of the fund shrinks to zero (as in a market neutral fund).  What was a modest negative exposure to tech and healthcare, in the example below, becomes gargantuan when compared to the funds puny net exposure.

To combat this deficiency in traditional analytical techniques, BNY Mellon essentially recommends that net exposures be calculated on a sector basis and by beta- and market cap-weighting the holdings (collectively falling under the name of “tilt” approaches).

But the deficiencies in traditional fund analytics don’t end there.  The traditional approach to calculating portfolio liquidity (”days trading volume held”) uses a net exposure denominator.  Again, this could be a very small number for a market neutral fund - making the fund look wildly illiquid.  Instead, BNY Mellon recommends that the denominator be gross exposure, not net exposure.

While these issues are second nature for hedge funds themselves, they illustrate some of the challenges being faced by traditional long-only managers as they start to add on short positions in 1X0/X0 funds.

 Email this post to a friend   Print This Post 

2 Responses to “Short positions throwing a wrench in the works for traditional (long-only) institutional investors”

1. Friday links: short-selling squeeze « Abnormal Returns Says:
   July 18th, 2008 at 3:12 pm

   […] What will happen to 130/30 managers when shorting becomes more difficult?  (All About Alpha) […]

2. Anon Says:
   July 21st, 2008 at 5:45 pm

   Also see this article, called Falling Short, from Risk magazine’s June edition.