Backwardation is Not Normal

By Brad Zigler

The Great Commodities Debate ("Part I" and "Part II"), the knock-down, drag-out brawl 'tween Rick ("No to Commodities") Ferri and Larry ("The CCF Champion") Swedroe left some spectators wondering whose colors they should be flying.

"Neil," for example, recalled well-known financial auteur Bill Bernstein's dismissal of commodities in his Efficient Frontier column (www.efficientfrontier.com/ef/0adhoc/stuff.htm).

Bernstein relies upon a Keynesian explanation of "normal backwardation" to posit a demand for risk premiums by speculators in deferred futures contracts.

Ahem ...

This notion that backwardation (that is, when distant deliveries trade at a discount to nearby deliveries, see "The Battle Against Contango") is somehow "normal" for a storable commodity is, to put it charitably, quaint. John Maynard Keynes cooked up the hypothesis in 1923, arguing that large producers, hedging price risk, are forced to offer futures at a discount.

There are a couple of things wrong with this zeitgeist. First of all, it assumes only producers of commodities hedge; commercial users, who'd be buyers in the back months, are conveniently ignored. Most importantly, though, it simply doesn't pan out.

Look at the gold and silver markets. The largest gold futures traders, those classed by the Commodities Futures Trading Commission [CFTC] as "commercial," are net short gold metal in record proportions. Large commercials are nominally short more than a quarter million contracts of COMEX gold. That translates to 52% of total open interest. Yet, the gold market remains firmly entrenched in contango. The negative roll yield annualizes at 3.03%.

And silver?

Large commercial COMEX silver traders are net short more than 75,000 contracts according to the CFTC, representing just over 40% of total open interest.

Any signs of backwardation? Nope. The silver market reflected a 2.74% annualized roll loss at the last settlement.

There's definitely not a silver lining to the normal backwardation story.