New data leaves 130/30 brouhaha unresolved

130/30 21 Apr 2009

Since 130/30 or “short-extension” funds entered the investment lexicon over 3 years ago, it has generated considerable debate.  While industry commentators derided the strategy in the media, academics remained steadfast in their belief that short extension strategies have merit.

So what gives?  What is it in the academic literature that could support the existence of value in a strategy whose brief history seems to provide empirical evidence to the contrary?  Critics say that short extension strategies face some operational headwinds that long-only funds do not – that shorting is not simply the “reverse of long investing”.

The truth is that 130/30 funds are both academically sound and operationally-challenged.  An article by Morningstar published today provides a relatively dispassionate overview of these factors that stand in stark contrast to the typical critique of the 130/30 concept based solely on performance.  In it, Morningstar’s Nadia Papagiannis highlights several unique challenges faced by 130/30 managers:

  • Negative Carry: Short-selling is not as efficient and long investing.  It requires additional business processes such as locating a borrow and executing the various legs of a short-sale.  As a result, there is a necessary drag that presents itself as a negative spread between the PB’s lending rate and the interest earned on cash balances used as collateral.
  • Trading Costs: While not a cost of shorting per se, Papagiannis points out short positions tend to be more short term in nature, meaning that shorts turnover more than longs.  (We blame that on the catalyst-driven nature of short ideas, though, not the oft-cited risk of “unlimited loss”).

We do not count the rebalancing costs of sticking exactly to 130/30 (and not, say, 128/28) since this is not a necessary element of a short-extension strategy.  We also don’t count the unique skills or experience required to short-sell, since this is a factor relating to the management firm, not to the strategy itself (unless those skills will always cost more).  Further, we’re not counting the cost of paying dividends on shorts (the opposite of receiving them on longs) since the net dividend carry of the short extension (30/30) is a factor that lies outside of the 130/30 strategy itself.

In any case, negative carry and trading costs are unavoidable and real.  In part due to these costs, Papagiannis shows that 130/30 versions of popular US mutual funds has underperformed their long-only sister funds.  (Regular readers will note that this is similar to the back-testing conducted by Gordon Johnson in early 2008 – see post.)

We’ve removed the fund names to protect the innocent, replaced them with the 130/30 fund’s launch date, and created this chart with the data in the Papagiannis’ article.  Returns for both twins are since the 130/30 fund’s inception date.

The median outperformance of the 130/30 versions is about -40 bps.  Since 130/30 funds are assumed to have a 1.0 beta, this suggests that the short extensions have a negative alpha.  But this may not be accurate for two reasons: 1) the number of months is so small that the Beta of 1.0 may not be accurate and 2) The Beta may not even be 1.0.

In fact, Papagiannis shows that the average Beta of a sample of large 130/30 funds now ranges from 0.87 to 1.13.  The funds with a 1.13 beta actually outperformed its long-only sister fund – suggesting that it actually did even better than it first appears.  In the flip side, some under performing funds had betas of less than one underperformed the market, suggesting even worst value-added that it initially appears.

Despite all the criticism of 130/30 fund returns, the removal of the major outlier above (the second fund), bumps the median out performance into positive territory.  So if shorting really does come with new operational costs, then the managers of these funds may be adding enough alpha to at least cover those costs.

But this is splitting hairs.  It’s still far too early to reach any conclusions about 130/30 based on a median out performance that hovers around 0%.  If we are to drawn any conclusion about 130/30 funds, it’s that their alpha, like those of their active management predecessors adds up to zero.

Investors in active long-only funds have ignored this reality and have instead attempted to pick their own alpha-producing managers for years.  Short extension funds are simply an extension of this philosophy.

In conclusion, Papagiannis reconciles the views of proponents and critics of short-extensions:

“While we believe that 130/30 or 120/20 funds are largely marketing gimmicks, the idea of a long portfolio, combined with a short extension to emphasize a manager’s best and worst picks, is a good one. But the key to success is a dynamic short-extension, rather than sticking to a static 130/30 or 120/20 model, which has no economic rationale.”

Be Sociable, Share!

3 Comments

  1. Walt French
    April 22, 2009 at 9:13 am

    …the removal of the major outlier above (the second fund), bumps the median out performance into positive territory.

    Yes, and Bush cut the deficit sharply if you take out just his 2003 tax change.

    Also, the MStar article (the title of which, “130% Gimmick…” hardly seems “dispassionate”) failed to address the key issue in 130/30: the strategy is intended to give you benchmark return plus about two or three times as much of the manager’s active insights as a long-only portfolio. When the manager’s active insights are helpful, it will out-perform by that ratio; when non so insightful, under-perform by that ratio, compared to the long-only strategy. The tested period was one in which many active managers did poorly, so the extension strategies, IF WORKING AS DESIGNED, would also under-perform by more. Yes, there are costs of borrowing (not exactly real high these days) and inefficient borrowing markets, but there are offsetting efficiencies from being able to use more of your research in your portfolio — that small company that’s about to get tiny can make you some real money.

    From talking to portfolio managers who run long-short, the issues raised by the MStar article are quite manageable or downright trivial; the real concern is the quality of the insights.

    “Follow the alpha.”


  2. Rob
    April 22, 2009 at 10:31 am

    Interesting to note that over half of the fund families of which Morningstar is comparing to, is actually comparing against a strategy managed by a different portfolio manager. It seems to me that if you are going to compare a 130/30 to a long-only strategy, it would make sense to do so if both are managed by the same pm. Looking over the list there are some that are comparing quants to fundamentals, as well as completely different subadvisors against each other. While JPMorgan has been raising significant assets in this space, it is worth nothing that Morningstar’s endorsement comes as it is one of the few which is actually being compared against its direct long-only peer.

    Also, for the beta’s they are taking strategies managed to the R1000V and R1000G, and comparing the beta to the S&P 500. The article appears to be rather targeted to achieve the authors objectives, and proves what some researchers have been saying with the difficulty in analyzing these strategies in the traditional fashion.


  3. Peter Zeuli
    April 22, 2009 at 3:34 pm

    Regarding the structure, I really don’t know if the exposure is correct but what I do know is that it isn’t the structure that determines performance it’s the strategy and the effective implementation of the strategy that determines performance. I do think that 130/30 was marketed incorrectly from the beginning because people interpreted it as an absolute strategy and it just isn’t – it is a relative strategy. Also the best performers (to the best of my knowledge) have come from the boutique firms not the big shops and I believe it is for one reason – too many restrictions being placed on managers at big shops resulting in missed opportunities.

    We run a global, all-cap 130/30 on a limited-partnership basis but we are thinking about registering it as a 40 Act fund.


Leave A Reply

← Two studies find active management "put" was AWOL in 2008 Institutional investors driving "the new active management": Report →