130/30 Indices: True indices or like playing chess against a computer?

130/30 07 Jan 2008

During the final months of 2007, Credit Suisse and S&P both launched what they called “130/30 indexes”.  CS didn’t say much about their methodology at the time, but S&P published the entire index construction approach online.  After reading that approach on November 20, we wrote the following:

“…here’s the thing we don’t quite get: why would you want to benchmark yourself to another active manager? There is no common risk factor underlying these funds that can serve as a benchmark. There is no 130/30 beta. In fact, all a 1X0/X0 program aims to accomplish is to lever pre-existing alpha for greater returns if alpha is already positive or greater losses if alpha is negative. As IPE reports, a speaker at a recent conference referred to 130/30 as just a prescriptive technique. How do you index a ‘prescriptive technique’?”

Clearly anticipating such a line of questions, the developer of the Credit Suisse index, AllAboutAlpha Hall-of-Famer Andrew Lo of MIT, addressed this question head-on in his December 11 paper on 130/30 indexation (“130/30: The New Long-Only”).  The paper has generated quite a bit of chatter recently (the subject of this Pensions & Investments story today and a column in the Economist this week.)

Say Lo and co-author Pankaj Patel:

“…our proposal to put forward an algorithm or dynamic portfolio as an index is a significant departure from the norm. Existing indexes such as the S&P 500 are defined as baskets of securities that change only occasionally, not dynamic trading strategies requiring monthly rebalancing. Indeed, the very idea of monthly rebalancing seems at odds with the passive buy-and-hold ethos of indexation.” (emphasis added)

He goes on to admit such a view of indexation as “rather unorthodox” and a “break from tradition” and at times even sounds like he was responding directly to our November 20 comments when he acknowledges:

“A common reaction to the use of any strategy as an index is to cry foul. How can an active portfolio be used as a benchmark for other active portfolios, particularly if the very purpose of a benchmark is to gauge the value-added of active management?”

Okay.  So what gives?

Lo argues that as long as the index is mechanical and easily implemented, it can be called “passive”.

“However…as trading technology becomes more powerful and increasingly automated, the need for more dynamic indexes – indexes capable of capturing time-varying characteristics – will arise…In this paper, we argue that a dynamic strategy can also be passive if the rebalancing algorithm is succinctly mechanical and easily implementable.”

This makes sense.  As he points out, a lot of strategies that were once considered to be “active”, have long since become “passive” since they can now be executed using a rules-based, computerized approach.

He invokes two examples to make the point: fundamental indexation and life-cycle funds. He says there is little argument that life-cycle funds (that also dynamically reallocate according to a set of rules) are passive, even though they are not traditional “buy-and-hold” passive strategies.

We agree.  But we still wonder if this makes any one lifecycle fund a representative “index” of the strategy?

In contrast, we feel that Fundamental Indexation might be a more direct comparison.  Here is a strategy that many argue is just an active value-bias strategy, not an “index” at all.  Still, it seems to have a somewhat legitimate claim to the “index” moniker since its factors are not only transparent, but highly intuitive.  A typical active value fund has a far more idiosyncratic feel – even if it were highly rules-based.

So it seems quite plausible that a dynamic strategy can certainly be considered “passive” as Lo points out.  But we still don’t see the value of an “index” that does not seem to capture any risk factor or be representative of any underlying strategy.  As Lo himself says in the paper:

“Of course, it is a well-known fact that a long-only portfolio with no alpha will not benefit from the exibility of leverage and short-sales…For a 130/30 portfolio to yield positive excess return above and beyond its long-only counterpart, the factors used to construct expected-return forecasts must add value.”

(We’d probably find more comfort in a 130/30 index that measured the relative merits of a generic short-extension strategy (vs. the same investment strategy executed as a long-only fund) given the particular market dynamics at a point in time.  Such an index could integrate exogenous factors proposed by Citibank’s Manolis Liodakis in a guest posting here last summer and/or the factors proposed by Roger Clarke, Harindra de Silva, Steve Sapra and Steven Thorley in a paper last year called “Long/Short Extension: How Much is Enough” (see related posting).  Such an index might not tell us much as a traditional index, but at least it would bifurcate the value created by short-extending from the value inherent in the underlying investment strategy.)

In the end, it appears that the CS 130/30 index and the S&P 130/30 index may really just be dueling quant models.  Or maybe they’re something in between quant models and truly representative indices?  Perhaps they should be interpreted as a mechanical bogey or hurdle. (Then again, remember when chess grandmaster Garry Kasparov played IBM’s “Big Blue”?  Big Blue also used “passive” rules-based processes.  But it could hardly have been described as an “index” of all chess players.)

Lo concludes by acknowledging the increased mental “effort” required to make sense of this new breed of index.  Still, he predicts a wave of new dynamic indices in the future:

“Although the interpretation and implementation of such dynamic portfolios will require more effort than the standard buy-and-hold indexes, this is the price of innovation as institutional investors become more engaged in alternative investments. And as trading technology becomes more sophisticated, we anticipate the creation of many more benchmarks from dynamic trading strategies, and we hope that the 130/30 index will pave the way for that future.”

Ed: We will all have to deal with fundamental indexing’s Rob Arnott – he owns the patent to “non-capitalization-weighted indexing” (no kidding – see related posting)

Be Sociable, Share!

2 Comments

Leave A Reply

← 130/30 Implementation Challenges What's behind Trusco's partial acquisition of 130/30 manager Alpha Equity? →