Barclay's CORALS is a Commodities Index Gem

We're pretty jaded around here when it comes to product launches. 2007 was a year of unprecedented activity in new ETNs and new ETFs. But when someone like Barclays walks in the door with a quant strategy, it's generally worth a look. Barclays - both the Barclays Global Investors flavor and the regular old Barclays Capital - has a bit of a reputation for knowing what they're doing when it comes to tilts and timing strategies.

Last week, Barclays Capital rolled out the Commodities Out-Performance Roll Adjusted Liquid Strategy Index (CORALS). Apart from the obvious excitement (that the guild of acronymists was clearly back from strike), the thing that caught our eye was this line from the press release.

"Extract positive returns in differing market cycles including stagnant or falling markets utilising a long-short strategy."

I'm a sucker for long-short strategies, at least on an intellectual level. Investing long always seems like you're playing with half the cards in the deck. After all, if you've got deep conviction, informed opinion or a really pretty black box that tells you "Buy X," the corollary is almost always "Don't buy Y." I'm not suggesting it's a strategy I've aggressively pursued with great personal success, but it is one that has always intrigued me.

Long/Short investing in commodities isn't entirely new ground. Deutsche Bank officially launched their market-neutral commodity strategy last month as well, but put precious little information out about it. The Barclays strategy seems intriguing though.

Here's the deal: The index allocates assets into 12 S&P GSCI subindexes. Each allocation can be up to 20% of the index, or 10% short.


Underlying Contract

S&P GSCI Aluminum ER Index

Primary Aluminum (LME)

S&P GSCI Copper ER Index

Copper Grade A (LME)

S&P GSCI Silver ER Index

Silver (COMEX)

S&P GSCI Gold ER Index

Gold (COMEX)

S&P GSCI Heating Oil ER Index

Heating Oil (NYMEX)

S&P GSCI Crude Oil ER Index

WTI Crude Oil (NYMEX)

S&P GSCI Natural Gas ER Index

Natural Gas (NYMEX)

S&P GSCI Gas Oil ER Index

Gas Oil (NYSE:ICE)

S&P GSCI Unleaded Gasoline ER Index

Gasoline RBOB (NYMEX)

S&P GSCI Wheat ER Index

Wheat (CBOT)

S&P GSCI Soybean ER Index

Soybean (CBOT)

S&P GSCI Corn ER Index

Corn (CBOT)

Click to enlarge

How the model decides what position to take is interesting. The primary constraint, other than the +20/-10 split, is that no index have a zero position. In other words, cash is not an option, and there is no pretense of market neutrality.

Instead, the model takes a quant view of the world. It builds a forecast for each individual commodity, based on everything from consumer confidence data to MACD (Moving Average Convergence/Divergence, which defines as "a trend-following momentum indicator"). Then the model takes all of those individual commodity forecasts, presses them through the black-box pasta mill and comes up with a long/short portfolio.

With the tremendous caveat that backtested quant models always look good to reporters because companies never roll out a strategy that backtests BADLY, here's what this thing looks like since 2000:

It's easy to be skeptical. Those bars roll up to an annualized return of nearly 25%. But the number to really look at is 2001. In 2001, the Goldman Sachs Commodity Index fell 29%. The Barclays Lucky Charms Index seems to have not only avoided this entirely, but turned it into a 10% winner - a feat very, very few investment managers pulled off in 2001.

So what was the secret sauce? Being short at the right time. Here's the long/short allocation of the strategy during that period.

But here's the thing: You can't buy this strategy. Not only can you not buy this strategy in a convenient ETN/ETF form, it's not exactly easy to mimic, even if you do have an account that lets you easily trade the giant list of individual ETCs from ETF Securities. But perhaps more importantly, you couldn't have invested in this strategy back in 2001.

The problem with quantitative strategies is this: They assume that tomorrow will look like today. While many, many financial relationships are predictable, these assumptions can break down most dramatically when the fundamental nature of the market changes.

So the question becomes, is the commodities market of 2008 substantially different than that of 2001? I would argue that it is - this Web site is proof. 2008 holds many more market participants, more efficient information exchange and a much higher awareness of the very fundamentals - demand from Asia, supply constraints, etc. - that are at the core of the CORALS model.

As with most investing, it comes down to what you believe. The CORALS Commodity Research team says:

"Given that fundamentals are not a fad, they always exist and can be clearly tracked; they offer a solid basis for a long- and short-term investment strategy."

For the moment, the question is moot. Barclays isn't saying when or if they plan on doing anything with this index. but is quietly running institutional money against it. But if, say, an iPath ETN were to follow, it would certainly be something worth considering.

We've long held that, unlike in equities, the long-only indexing may not be the most efficient way to access the commodities market. Equities and commodities are fundamentally different, and factors drive each market. The CORALS are an interesting idea, and we'll be watching to see how they perform in the real world.

Click here for more on ETFs.