Rolling Into Trouble

November 11, 2009

For reference, the ETF Securities Crude Oil ETC (LSE:CRUD) was up 5.2% over the period from 31 December to 31 October, while the ETF Securities Brent Oil ETC (LSE:OILB) was up 33.1%. Both ETCs track the front end of the oil market futures curve, so why the difference?

As it turns out, there are two primary reasons.

First, Brent oil (which OILB tracks) has significantly outperformed WTI (tracked by CRUD) this year, as the chart below shows.

Oil_Index_Performance

Daily oil price data are taken from the US Energy Information Administration website and show that over the period from 31 December to 3 November, Brent was up 111% while WTI Cushing rose 78%.

If you look at the relative performance of the two benchmarks in the early weeks of the year you’ll notice some weird price spikes – up for Brent and down for WTI – and these are better illustrated in the second chart, which shows the differential between the two oil prices during 2009.

Oil_Discount_2

Brent moved from close to a US$9 discount to WTI on the last day of 2008 to a record US$13.2 premium on February 12. Since then it has gradually declined to a discount, which is where it has historically traded.

The relative price performance of Brent and WTI accounts for most of the difference between the returns of OILB and CRUD – around 21% on a relative return basis.

The second key influence on the oil trackers’ returns has been the effect of their rolling policy.

CRUD, which tracks the DJ-UBS Crude Oil Subindex, rolls its positions every two months and from the near (expiring) NYMEX WTI contract into the third month contract. The index roll takes place from the fifth to ninth business days of the month.

OILB rolls its market exposure monthly from the first to the second month ICE Futures Brent oil contract, with the roll occurring from the first to fifth business days of each month.

While contango (having to sell out of an expiring contract at a lower price and reinvest in longer-dated contracts at a higher price) has reduced returns for both trackers to well below the performance of spot crude in 2009, there was a particularly expensive roll for CRUD in February.

CRUD’s underlying benchmark, the DJ-UBS Crude Oil Subindex, rolled its futures exposure from the March to May 2009 NYMEX WTI contract over the five business days from 6 to 12 February, inclusive. (For further information on the index methodology and roll dates, click here).

Here are the NAVs of CRUD and the daily WTI Cushing spot prices for the five roll dates plus 5 and 13 February (the days before and after the roll period).

CRUD NAV (US$)

WTI Spot Price (US$)

Thu-05-Feb-2009

22.23

41.15

Fri-06-Feb-2009

21.69

40.24

Mon-09-Feb-2009

21.42

39.58

Tue-10-Feb-2009

20.44

37.54

Wed-11-Feb-2009

19.90

35.93

Thu-12-Feb-2009

19.71

34.03

Fri-13-Feb-2009

19.29

37.63

Period Return

-13.22%

-8.55%

While WTI crude declined from US$41.15 to US$37.63 over the period from 5 to 13 February (i.e. by 8.55%), the spot price seemed particularly depressed on the days when the index roll took place – 6,9,10,11 and 12 February – falling to US$34.03 on the last day of the roll before rebounding the next day.

CRUD, which had already traded out of the spot contract by the end of the roll period on 12 February, was therefore unable to benefit from the snapback in the spot price on 13 February (as you can see from the change in CRUD’s NAV from the twelfth to the thirteenth).

Overall, this single roll period cost CRUD over 5% in performance when measured against spot crude price changes.

Did oil traders “front run” CRUD and similar index-tracking products, or at least position themselves to benefit from the roll? That’s difficult to prove, though there seems to be a prima facie case to answer.

But this single example illustrates how careful one has to be as an investor when committing to a preannounced trading algorithm and raises much broader questions about the suitability of futures-based products for passive investors.

 

 

Find your next ETF

CLEAR FILTER