The Senate Permanent Subcommittee on Investigations released a report on “Excessive Speculation in the Wheat Market.” The report blames index funds in particular for driving up wheat prices, and particularly for driving a wedge between wheat futures prices and the price of cash wheat, both prior to and at contract expiration.
The methodology of the report should be tiresomely familiar by now. It notes concurrent events–(a) increased volume of trading by long-only index investors, and (b) higher prices and lack of convergence–and asserts (a) caused (b). To be fair, the report does make comparisons between Chicago wheat and Kansas City wheat, which have experienced different degrees of index participation (measured by fraction of open interest) and convergence problems, but this analysis is weak because it does not attempt to determine whether other differences between the contracts might also be responsible for the differences in behavior. (A bit more on this below.)
However, the report largely ignores the rigorous empirical evidence that has examined the relations between index fund trading and price changes, and found no reliable evidence of such a connection. Moreover, there are numerous commodities embedded in the indices, and the following the logic of the hypothesis, increasing index activity should affect price levels and cash-futures relations in all of these markets. The report does not undertake a rigorous cross-sectional or time series of the cross-section analysis of the relation between index participation and the basis or convergence.
There are also serious logical problems with the analysis. The main issue relates to the convergence of deliverable cash and futures prices as a contract moves to expiration. It is very difficult indeed to see how index funds could possibly be implicated for this phenomenon. Index funds roll from the expiring contract to the first-deferred contract before expiration. They therefore cannot have a direct impact on the expiring futures price at expiration. Indeed, during the roll, long-only funds are selling the expiring futures, which per the crude story embedded in the report should cause that futures price to decline relative to the cash price. In brief, I cannot think of any reasonable argument explaining how index funds can prevent convergence when they are out of the market altogether when the absence of convergence has occurred.
Relatedly, the report repeatedly states that the index funds are creating demand for wheat (and other commodities). Uhm, no, actually. They don’t take delivery, and don’t consume the physical wheat or hold it off the market. For every contract they buy, they also sell, so they don’t create any demand. Moreover, although unexpected flows of buy orders can cause price reactions even if the orders are submitted by the uninformed (because other market participants don’t know exactly who is submitting the orders and there is a positive probability that the buy is from an informed trader), index activity is pretty well understood, and unlikely to result in the market being systematically surprised as would be required to support the report’s conclusion.
The analysis also begs the question of why smart money hasn’t piled in to offset the purported effect of the dumb index money. After all, in the scheme of things, the Chicago wheat futures market isn’t that large. If the investment of the index funds was sufficient to drive up prices, not that much investment would be required to have an offsetting effect. During the time that the basis and convergence anomalies were most pronounced, the world was awash in liquidity and capital desperate for returns. Not only was money piling into indices, it was also going into hedge funds and managed futures funds whose whole business is to identify mispricings and jump on them. That is, if this was an era of “excessive speculation” as the report laments, why didn’t other evil speculators trade ruthlessly against the dumb money?
All this said, the absence of convergence during the period studied (most pronounced in wheat, but also present in corn and soybeans) is a puzzle. It would seem to present a fabulous opportunity for operators of delivery facilities (e.g., Cargill, ADM and The Andersons) to make huge money buying cash, selling expiring futures, and delivering, and continuing to do that until convergence was achieved.
In a competitive market, the spread between expiring futures and cash prices should equal the marginal cost of making delivery. This should be true, regardless of whether index funds are inflating the entire price structure or not. What could explain such an extremely high shadow cost of delivery?
One difference between Chicago and Kansas City is that the former’s contracts use a shipping receipt delivery mechanism which is effectively a call option on delivery elevator capacity, whereas the latter’s is a traditional in-store warehouse receipt delivery mechanism. The report doesn’t mention this difference, which in theory could have an effect on futures-cash spreads (as I outlined in posts last summer). However, it is difficult to believe that this option-based story is sufficient to explain the very large futures-cash spreads at expiration.
Where convergence is concerned, the first focus of inquiry should be the behavior of delivery elevators and their operators. It needs to be asked: “Why don’t you deliver more when the expiring future is so far above the cash?” It is an extremely serious failing of the report that it does not ask this question. Interestingly, the names “ADM,” “Cargill,” and “The Andersons” appear only once in the report–and then only to state factually that they operate delivery facilities. It is beyond belief that any serious study of the convergence issue would not attempt to dig into this real mystery of why operators of delivery facilities do not pick up the money that appears to be there in large amounts for the taking. This is not an accusation of any wrongdoing–perhaps there is an explanation, and the apparent large sums of money to be had are chimerical because there is some cost that the delivery facility operators incur that is not obvious to me and other observers of this phenomenon. But it should be the primary focus of any examination of convergence. In the Senate report, this issue doesn’t even rise to the level of afterthought: it isn’t thought about at all.
The flagrant failure to engage the primary puzzle suggests that this was not a serious endeavor to understand an admittedly anomalous regularity in the wheat market. Instead, it suggests that the Subcommittee and its chairman, Levin, had predetermined agenda: Speculation bad! Index speculation bad! The report goes nowhere close to proving that point. Indeed, its watery-thin empirics, shaky logic, and failure to ask the most basic question prove only that one shouldn’t look to legislative reports for a serious discussion of serious issues. They are polemics, not science.