Streetwise Professor

March 5, 2013

Riding His Anti-Hedge Fund, Anti-Speculation Hobby Horse

Filed under: Commodities,Derivatives,Economics,Energy,Exchanges,Financial crisis,Regulation — The Professor @ 11:37 am

John Kemp of Reuters-accompanied by cheerleading by Izabella Kaminska at FT Alphavilleis going on about the “disconnect” between nearby and deferred Brent.  He blames-wait for it-hedge funds.  Natch.

The Brent market is currently in a steep backwardation.  I demonstrated empirically almost 20 years ago that backwardation is associated with low correlations between spot and futures prices for oil and a variety of other commodities.  Indeed, my 1994 criticism of the Metallgesellschaft 1992-1993 “hedging” strategy-which led to several confrontations with Merton Miller-was based on the fact that MG’s “hedge” of long the nearby against distant deferred short positions was in fact risk increasing due to the fact that MG implemented this strategy during a backwardation, and this reduced substantially correlations thereby making the MG position very risky.  My 2011 book provides a robust model that predicts exactly this result.

The intuition is quite straightforward-as I’ve been teaching for about 20 years too.  Inventory is what connects spot and futures prices.  When inventories are large, the market is in contango, and spot and futures prices move together: cash-and-carry arbitrage connects these prices.  In contrast, when inventories are low, the market is in backwardation, cash-and-carry arbitrage doesn’t link the spot and the futures, and the correlation between these prices can go very low.  Hence the association between contango and high correlations.

Note: hedge funds, speculation, yadda yadda yadda have nothing to do with this.

Kemp does note that there is physical tightness that does explain the backwardation:

Overlaying all these broader factors are continuing problems with production of the four North Sea crude streams (Brent, Forties, Oseberg and Ekofisk) that physically underpin the Brent futures prices. BFOE crudes remain in short supply, keeping the market in a steep backwardation, with futures prices tending to rise sharply in the run up to contract expiry.

But then he discards this fundamental fact, and mounts his favorite hobby horse of bashing hedge funds and speculation.

Note even in this paragraph there is a telling piece of information that contradicts his view: “with futures prices tending to rise sharply in the run up to contract expiry.”  Uhm, that’s exactly when hedgies and other speculators are liquidating-selling-their nearby contracts and rolling them into the deferred months.  This should put downward pressure on nearby prices, if the speculators were really  in command. Completely inconsistent with his assertion that hedge funds are driving the disconnect between spot and futures.

Kemp also makes a comparison to spot price and curve movements in 2008 to more recent movements.  But as I also show in my book, to explain commodity price dynamics you need multiple shocks of differing persistence.  Curve shape is driven mainly by transitory shocks: that’s what inventory is used to smooth out.  The level of the curve is largely driven by persistent, business-cycle type shocks.  This means that conditioning on price levels alone is insufficient to make an apples-to-apples comparison.  The 2007-2008 boom was driven more by long run, secular factors-namely the Asian/Chinese growth boom.  This resulted in a rise in the price level and only a modest increase in backwardation.  That is completely different from current conditions-hence the different behavior.

Similarly, the differences between high correlations pre-2010 and low correlations now are readily explicable.  The market was much more abundantly supplied in 2009-2010 due to the severe economic contraction following the financial crisis, and as a result, the market was in contango most of that time-at times in a “supercontango”.  Again, one would expect this to be associated with high spot-futures correlations.

Moreover, there is a big difference in the composition of shocks, and the magnitude of these shocks that also explains the observed declines in correlations.  The 2009-2011 period was dominated by macro uncertainty driven by the financial crisis and then the Eurozone crisis.  These shocks tend to be persistent, affecting both current and expected future economic conditions in the same way, thereby contributing to high correlations between points on the curve; moreover, they tend to affect all commodities and asset classes similarly, leading to high correlations across commodities and asset classes.

At that time, macro volatility-as measured by the VIX-was high.   In early ’09, VIX was in the 30-50 percent range, and remained above 20 percent for most of 2010-2011, with spikes up to 35-40 percent.  Now VIX is tame, with levels in the low-teens, just like during the period of the “Great Moderation.”  High macro volatility tends to lead to high correlations along the curve and across commodities and between commodities and equities: common shocks dominate, especially when they are big.  The decline in macro volatility means that commodity-specific, relatively transient shocks tend to dominate: this tends to depress correlations along curves and across commodities, and between commodities and equities.  These are exactly the patterns observed in the past months.

That said, there are reasons to suspect the backwardation and the decline in correlations might be overdone, but not because of the malign influence of hedge funds.  Specifically, given the declining Brent supply base, it is reasonable to ask whether the backwardation is excessive.  At present, Forties is cheapest-t0-deliver, and this represents only about 350kbd of production.  Overall BFOE production is only about 1mmbpd.  Given the immense open interest in Brent futures and OTC derivatives, it is more that possible that large players are exercising market power by taking delivery of too much physical oil, thereby exacerbating the backwardation (i.e., creating an artificial scarcity).

If you want to identify who might be doing that, look at who is taking the physical cargoes.  Those parties would be the squeezers.  If hedge funds are the culprits, they should be taking a lot of physical supply.  Kemp certainly doesn’t provide any evidence of this, and his piece suggests these players are just playing with paper barrels, not wet ones: that’s my understanding too.  Historically, the Brent market has been the scene of many squeezes, but the squeezers have tended to be trading companies (e.g., Arcadia) or perhaps supermajors.

Brent was a squeezers paradise in the 1990s due to declining physical volumes and growing paper volumes.  Platts’s addition of Forties, Ekofisk and Oseberg to the delivery slate increased supply sufficiently to mitigate the manipulation problems.  But the decline in production has continued inexorably, and Brent is increasingly vulnerable to squeezes.  Which is why Platts and Shell have competing plans to tweak the pricing mechanism, namely by providing premiums for OE (Platts) or BOE (Shell), thereby reducing the delivery pressure on the cheapest-to-deliver Forties.

Squeeze-driven backwardation also tends to reduce spot-forward correlations.  The spot price is driven by squeeze-related technicals, the forward price by longer run fundamentals.  Squeezes also distort the inventory holding decision, and cause a breakdown in the cash-and-carry arb mechanism.  As I show in another book (and many articles).

So Brent may indeed be broken, but hedge funds haven’t broken it.  It is a classic problem in derivatives markets: a burgeoning derivatives market balancing on top of a declining deliverable supply.  I often analogize this to an inverted pyramid: and in Brent, the base of the pyramid (the paper market) is growing while its point (the physical market) is getting sharper.

Fixing Brent requires enhancing deliverable supply.  Not an easy thing to do.

Again contrast Brent and WTI.  Brent boosters have constantly bashed the WTI disconnect.  But this can be fixed by investments in infrastructure, which are being made, though not as fast as expected or as needed.  But building infrastructure is a helluva lot easier than building  a new Buzzard. Trust me on this.

So in the medium term, I expect Brent will get broker, and WTI will get better, leading to a shift of much futures and index trading activity (including hedge fund trades) back to WTI, which will no doubt lead John Kemp to saddle up his hobby horse and ride west into the Oklahoma sunset.

Print Friendly, PDF & Email


  1. […] education on backwardization and contango. […]

    Pingback by Breakfast Links - Points and Figures | Points and Figures — March 6, 2013 @ 12:11 am

  2. In early 2009, if you could take physical delivery of the oil, you could have simultaneously bought Feb oil, sold Aug oil and locked in a 12% profit. Trick was being able to take delivery and store it.

    Comment by j — March 6, 2013 @ 12:14 am

  3. The Reuters article’s all about funds speculating in Brent and the graphs are all about…WTI? WTI is 80% CME by volume so I think at minimum he needs to change the title of his article.

    BFOE production is about 1.8mmbd by the way, not 1mmbd.

    In general building infrastructure *should* be easier, yes. But this is the USA we’re talking about, home of Congressmen who lobby both against drilling for hydrocarbons in the USA, and also against reliance on imports ((Dianne Feinstein was it not? Essentially a woman vehemently opposed to the outcome of her own policies), meaning they want more drilling in the USA.

    Against NIMBY / BANANA issues on this scale I suspect that including Urals or Bonny Light into the North Sea assessment on a CIF basis is a comparatively simple task.

    Comment by Green as Grass — March 6, 2013 @ 3:45 am

  4. Arguments about the increase in managed money positions tend to focus on the outright level of positions, and ignore that they haven’t really changed as a share of the market since 2009. Latest CFTC/Ice COT data puts them at 21pc of all open positions on the major oil futures contracts (Nymex and Ice WTI, Nymex heating oil and Rbob, Ice Brent and gasoil). It’s at the top of the range, but still within the 16-21pc market share they’ve had over the last four years (Ice Brent and gasoil data only come into the calculation in mid-2011, but there’s not much discontinuity when they do). And in Ice Brent, they’re less than 17pc of open interest.

    Managed money has been increasing its share into the top of the range since mid-2012, but it’s largely because swaps dealers are reducing their open positions. This more or less corresponds with commercial oil firms’ general move away from hedging on swaps markets now that OTC derivatives have to be cleared. Swaps dealers have less exposure to hedge on old-fashioned futures markets.

    By the way, Green as Grass, BFOE production was 800,000 b/d last year, 900,000 b/d in 2011 and 1.1mn b/d in 2010. And this US Nimbyism point is overdone. Pipelines are being built and drilling is being done. The only real hold-up is Keystone XL, which may well be approved. And even if it isn’t, there will be more rail capacity moving bitumen from Alberta.

    Comment by Down With This Sort Of Thing — March 6, 2013 @ 6:43 am

  5. Derivative markets can alleviate the squeeze with more explicit “exchange for physical” pricing. The paper oil market can exist independently upon many wet markets, and not be beholden to declining Brent supply or WTI filling every bathtub in the inland US. The bond market has a well-developed “cheapest to deliver” (CTD) formula for delivering physical bonds based on a “nominal” bond. There could easily be support for an oil contract that is based on a nominal barrel of oil, and physical delivery could be satisfied with premium or discount formula for sulfur content and delivery cost to major port. Not rocket science, just simple calculations to transform fungible financial contracts into physical goods. The CME and ICE are too afraid to cannibalize liquidity on their favorite contracts, but they risk losing a major opportunity. The Brent/WTI spread is ridiculously wide for an oil producer with good mid-stream capability, leaving them with no suitable derivative contract.

    ICE understands the benefits of electronic trading, but they still seem clueless about integrating physical and financial products. For example, TTF is just another gas contract to them, not a profound way to virtually and financially manage gas throughout Europe, A virtual oil contract is a big opportunity.

    Comment by scott — March 7, 2013 @ 3:54 pm

  6. @scott. There has to be a tie between physical and paper markets at expiration. A truly virtual contract is not realistic, and given the vagaries of determining cash prices, cash settlement is a non-starter. A viable oil futures contract has to allow for delivery. But there are definite advantages for permitting delivery of multiple grades at multiple locations.

    Such delivery options are a way of reducing vulnerability to squeezes. The Treasury mechanism usualll works well, but sometimes (like in June ’05) there can be substantial differences between the price of the CTD and the price of the 2d CTD which gives a squeezer the ability to extract a premium. This means that the premiums/discounts must be adjusted periodically to reflect market conditions.

    What happened to the 10 year in 2005 is similar to the situation now in Brent, where there a substantial discount for Forties vs. BOE. That discount means that squeezers have some ability to jack up the price before they have to worry about anybody delivering BOE. This is why Platts and Shell are proposing premiums to make the higher-priced grades competitive for delivery.

    I could design an oil delivery mechanism with a wide variety of deliverables: multiple grades and/or locations with premiums and discounts to reflect different values.

    I published a paper in J. Futures Markets in the early-90s discussing such a system for corn and soybeans: my book Grain Futures Markets: An Economic Appraisal also includes such an analysis.

    I show that if the premiums/discounts are chosen appropriately, not only will the mechanism be less vulnerable to squeezes, but it will provide better hedging performance for many market participants. The delivery option essentially makes the futures price a weighted average of the prices of the various deliverables. This helps enhance hedging performance for a broad array of market participants by making the futures price like the price of a portfolio. The portfolio effect reduces the impact of idiosyncratic shocks to any single price. Put differently, such a futures price is like an average price of oil, where the averaging takes place across grades and locations. This tends to enhance average hedging performance. It’s not a “perfect” hedge for anybody, but it’s a better hedge for most than any single-deliverable system.

    Speaking from rather intense experience working to redesign the corn and soybean delivery mechanisms on the CBT not just once but twice that although such a mechanism has many attractive features, commercial participants tend to hate it. Hence, it is a major political battle to redesign contracts along these lines. The stories I could tell about the fights-in one instance almost coming to physical blows-that I had with CBT members over my proposal to move to such a system for SY and CN. I had a similar experience with the Winnipeg canola contract.

    The big players like to be able to exploit the technical features of inefficient delivery mechanisms.

    My experience with trying to improve the efficiency of grain/oilseed futures delivery mechanisms is what educated me about the political nature of exchanges. To amend Bismarck: you never want to observe the making of laws, sausages, or futures contract delivery specs.

    The ProfessorComment by The Professor — March 7, 2013 @ 9:18 pm

  7. I have not lived in the WW1 trenches of designing an ideal tie between a physical and paper markets like you, but I am well-versed in the incumbent’s desire to exploit the inefficient delivery mechanism.
    The major players “have” exploited the political nature of futures exchanges, and generally benefited in the short-term, only occasionally being on the wrong side of squeeze sometimes. But the CTD mechanism is not doomed from the 2005 episode, or your under-implemented ideas for SY and CN. Those ideas are just latent, waiting for a market platform to develop them.

    1) Commodities are fungible into individual molecules, not a limited set of bonds, and 2) formal exchanges are unlikely to survive a decade in their current form – with large blockbuster contracts such as Brent, WTI, HenryHub. I am naively hopeful that financial innovation will bring CME and ICE to their knees, as upstart trading hubs develop flexible platforms for trading oil algorithmically making any adjustments for grade,location and quantity. The CME spends all their money on lawyers rather than technology, and they will go the way of the NYSE in equities. A hallowed name without any trading on their floor.

    Designing a virtual oil contract with a continuum of grades and hundreds of locations is straightforward, it just has to be automatically calculated on an electronic platform and deliverable into a physical product. Driving flow to a new electronic platform requires a catalyst, either a compelling economic advantage, or a regulatory threat. Franken-Dodd may qualify? or the platform is designed to arb the current exchange contracts and a hedge-fund starts eating the energy incumbents lunch.

    Your academic efforts maybe the foundation for this transformation. I can’t image a Brent/WTI spread being relevant decades into the future. Change will come.

    Comment by scott — March 11, 2013 @ 10:39 am

RSS feed for comments on this post. TrackBack URI

Leave a comment

Powered by WordPress