Guys and women,
See article below this. I read these two papers in '09 and they make sense.
I would like to build a simple futures trading monthly rebalance system that ranks a universe of 40 futures contracts on ‘value’ (degree further out contracts are more than rolling forward one month contracts) and momentum and then buys an equally weighted, diverse basket of 10 with stop losses and ‘sector allocation weights’.
This is based on a lot of solid academic research. I would love for it to autotrade. Anyone know of a really good platform (tradestation or fidelity wealth lab?) to do this on? Anyone good enough at programming to want to help?
Thanks,
Tom
Article:
If you’ve studied the literature, you should be very excited about the forthcoming SummerHaven commodity ETF from U.S. Commodity Funds.
The news, if you missed it, was reported here. U.S. Commodity Funds, the company best known for creating the U.S. Oil ETF (USO) and the U.S. Natural Gas ETF (UNG), has licensed a new actively managed commodity index developed by a company called SummerHaven, which was itself launched by Yale University Professor K. Geert Rouwenhorst.
That’s interesting, because Rouwenhorst is one-half of the two-person academic team that wrote the most important academic papers establishing commodities as a legitimate asset class. The other half, Yale Professor Gary Gorton, is a senior adviser to SummerHaven. Rouwenhorst/Gorton are the Fama/French of the commodity world.
The paper that broke things open for commodities was Facts and Fantasies about Commodity Futures, published in 2005. In the paper, Gorton and Rouwenhorst created an equally weighted index of commodity futures and studied its performance from July 1959 through December 2004. They found that commodity futures delivered similar returns to equities but were negatively correlated with both stock and bond returns. Here’s the killer line from their conclusion:
“In addition to offering high returns, the historical risk of an investment in commodity futures has been relatively low – especially if evaluated in terms of its contribution to a portfolio of stocks and bonds. A diversified investment in commodity futures has slightly lower risk than stocks – as measured by standard deviation. And because the distribution of commodity returns is positively skewed relative to equity returns, commodity futures have less downside risk.”
For institutional and retail investors alike, this was manna from heaven: equity-like returns that are non-correlated to other markets. The data was robust, it was convincing, and billions and billions of dollars flowed into broad-based, passively managed commodity index products.
But while investors fixated on Facts and Fantasies, Gorton and Rouwenhorst didn’t stop.
In 2007, they published a paper called “The Fundamentals of Commodity Futures Returns.” (You can download the full paper from SSRN here, or read an excellent summary of the piece here.) In this paper, Gorton and Rouwenhorst, along with Fumio Hayashi of Hitotsubashi University, found that, historically, it’s been relatively easy to improve on a simple naïve index return by focusing instead only on commodities with low levels of inventory.
Common Sense
This intuitively makes sense: Commodity pricing is driven by actual supply and demand, and actual commodity users will pay a premium to get supply now when materials are scarce.
The professors determined that commodity inventory levels are reflected by the degree of backwardation or contango associated with a given commodity: When inventories are low, commodities tend to trade in backwardation (tomorrow’s corn costs more than next year’s); when inventories are high, commodities tend to trade in contango (cheap corn today, pricey corn next year).
They then compared the returns of an equal-weighted portfolio of 33 commodities futures from 1969 to 2006 with an equally weighted portfolio composed only of commodities trading in backwardation. They found that the backwardation portfolio outperformed the equally weighted index by an average of 5.4 percent per year; by comparison, an equally weighted portfolio of commodities trading in contango under-performed by 4.8 percent per year.
They also showed that using momentum as a screen for selecting commodities generated high returns. They suggested that commodities may show strong momentum returns because it takes a long time for commodity inventories to recover from shocks.
Put this way, the insights seem fairly basic, and it has surprised me over the years that investors have stuck with classic, broadly diversified commodity indexes—particularly those that simply buy front-month futures—in the face of this data.
I’m a fan of indexes, but simply because broad-based indexing is the best strategy for stocks does not mean it is the best strategy for commodities, or that all indexes are created equal. The commodity markets are different from the equity markets. They are driven by different factors. The fact that commodities are consumed and produced on a daily basis should be enough to tell you that the rules of stock investing may not apply.
It’s been almost four years since the first commodity index ETF hit the market, and there are billions of dollars invested in the funds. It will be exciting for this index fan to see a new active wrinkle hit the market.
Original post
More on choosing the right ETF