Crude Indices and DXO/DTO

So in the last installment of this series, Proshares levered ETFs UCO and SCO were checked and analyzed along with a quick 'n' dirty check of their underlying benchmarks correlation with crude oil. Although one commenter felt that DJAIGCL had little correlation to NYMEX spot, I disagree and recommended he pull the source data and run the closing prices correlation for himself. (Also pay close attention to what I say and more importantly, what I don't say.) Anyhow, for purposes here, I'm going to assume you grant me the earlier statement that DBOLIX actually is reasonably correlated to crude prices. And really, it doesn't matter all that much since this analysis is to see how well the levered ETFs track relative to their respective benchmarks.

While DXO open/close data is available going back to 6/23/08, this data is only available for DBOLIX back to 10/27/08 so the charts here are confined to that date up to 2/10/09. During this period DXO's average track error was -0.71% and the average absolute track error was 3.26%. This compares to UCO's values of -0.16% and 3.62%, respectively. There were 28 positive track errors against 44 negative ones during this period, so one could argue there is a bias for underperforming the +2x intraday percentage expectations. Finally, the closing price correlation between DXO and DBOLIX is 0.987 and the correlation of DXO's intraday percent change with tracking error is 0.01.

As for DTO, the same period applies. During this period DTO's average track error was 1.14% and the average absolute track error was 3.8%. This compares to SCO's values of -0.312% and 3.52%, respectively. There were 41 positive track errors against 31 negative errors during this time. The closing price correlation of DTO and DBOLIX was -0.89 and the correlation of DTO intraday percent change with tracking error is 0.017.

On balance, it seems that DXO/DTO have more difficulties in tracking their benchmark and target performance than UCO/SCO. Part of the explanation for this could lie in the fact that DXO/DTO are ETNs sponsored by Deutsche Bank as opposed to ETFs. The ETN suffers from risk exposure both to the underlying commodity and the sponsoring institution and DB has of late been making news for fairly negative reasons. Here's a choice quote from the prospectus: "The PowerShares DB Crude Oil ETNs are riskier than ordinary unsecured debt securities and have no principal protection." On a more technical point, the ETN is exposed to contango - though some attempt is apparently made to minimize impact - and if you check the NYMEX charts with any regularity, you've seen how large the price spreads have become. An aside: USO also suffers from the same contango problems and as long as this condition exists and if oil stays relatively flat in the front month, there will be steady erosion as the contracts are rolled.

Crude Indices and UCO/SCO

Another installment in the occasional series of posts about leveraged ETFs and oil. As mentioned earlier, I have discovered the bulk of my readers visit here looking for information on oil and their various ETFs, both levered and not. In an earlier long post on the ultras, I tackled the issue of volatility and compounding and their potential effects on performance of ETFs such as DXO, DTO and their ilk.

There is another related issue here and that is the tracking effectiveness of these ETFs with their benchmarks. And taking a step backwards from that issue, how do the various benchmarks perform in tracking NYMEX crude? There are two families of ultra ETFs for oil - ProShares (UCO/SCO) and PowerShares (DXO/DTO) - with the former using Dow Jones AIG Crude Oil Sub-Index (DJAIGCL) as its benchmark and the latter utilizing the rather verbose Deutsche Bank Liquid Commodity Index - Optimum Yield Oil Excess Return (DBLCI-OY_CL or DBOLIX).

I used EIA's data on Cushing WTI spot pricing (see link to the right) and checked the correlation of DJAIGCL and DBLCI-OY_CL closing prices with this. Unsurprisingly, the correlations are tight with DJAIGCL having a bit better correlation over the period 1/2/07 - 2/10/09. DJAIGCL had a 0.99597 and DBLCI a 0.985602. In the chart to the right, I also tossed in USO since it is probably the most widely tracked oil ETF and for most people, a proxy for crude pricing. You will also likely notice the NYMEX front and NYMEX 2-month average. This latter value was thrown in because it provides some contrast in periods of significant contango (such as now) and on days with huge squeezes (typically assignment days.) One thing of note is that the value of DBOLIX charted is the DBOLIX divided by 10 so as to make the scaling a little better.

So, at this point I think it is reasonably established that these two benchmarks are excellent trackers of Cushing WTI spot pricing.

The next question is how well do ultra ETFs and ETNs perform with respect to their stated goal of achieving +2x/-2x daily percentage moves of their benchmark index. I started with UCO and SCO in this analysis. (My apologies to all the fans of DXO/DTO - next post.) Part of the reason for this, is that open/close data is available for DJAIGCL going back much longer than what I could find for DBLCI-OY_CL. The flipside is that UCO and SCO have been in existence a far shorter time than their Power Shares counterparts. It's probably easiest to start off with a chart. The average absolute value track error for UCO since its inception is 3.61%. For example, if DJAIGCL moved 2% intraday, UCO should be expected to move +4% intraday. If UCO instead moved +7%, my terminology is that this is a +3% tracking error. For better or worse, this doesn't appear to be a systematic problem and the average tracking error is -0.16%. A few other statistics seem worth a passing mention: the correlation of UCO with DJAIGCL is 0.979 and the correlation of the intraday percentage change of UCO with tracking error is a negligible -0.11.

Moving on to SCO, here's the same chart on the same scale. Basically, similar behavior and a similar average track error of -0.312% and average absolute track error of 3.52%. And the other correlations of SCO with DJAIGCL and intraday percentage change on track error are: -0.798 and -0.02, respectively.

The lack of any correlation of the tracking error with the direction and magnitude of the intraday percentage change of the index implies there is little that can be done to anticipate them, even within a session. While these two funds do offer leverage, more often than not they fail at their stated goals of 2x or -2x the intraday percentage move of the target benchmark, DJAIGCL.

I’ll likely give DTO and DXO similar treatment in the next week or so.

Revisiting Old Charts

Over the past couple of weeks I've posted up a few charts with some vague intention/goal of updating them now and then.

One of these was the SPX bottom finding chart. When I first posted it up on the 6th, it was somewhat concerning because of the implied risk for a reversal even though the indicator itself wasn't as reliable at finding tops as confirming bottoms. Here's an update of that chart to the right. On the plus side, the market didn't exactly leap upwards since then. On the 6th, SPX opened at 931 and Friday it closed at 831. I guess that's actually not too bad all things considered. This chart still confuses me a little bit when it's viewed over the span from Q1/2007 until now. Strangely, the spikes in risk for a bear reversal on SPX keep getting higher and the bottom spikes continue to be more shallow, forming a channel of sorts. Why speculators should be getting progressively more risk-a-philic puzzles me. Perhaps everyone is just getting more comfortable with the new market realities. The next sign-post I'll look for on this one will be around -0.50 and at that time it will be interesting to see how other SPX analysis is doing. This is the height of earnings season and so anything can happen.

The next thing I've talked about a little more recently is oil and the contango spread. This definitely deserves an updated chart since the spread has collapsed and has been cut roughly in half. Again, chart to the right. The last time I talked about it, the 1-year spread in NYMEX pricing was pushing towards $25/bbl. This has subsequently collapsed to just under $12/bbl and the 6-month has fallen from around $18.50 to just over $7/bbl.

To add a bit more detail, it turns out the spike in the spread was due far more to the plunge in the front-month contract (February until Jan. 20 assignment, March since) than a price run-up in the farther months. You can see this in the next chart that shows the front month price and 1 year out. Again, it should be noted that there was a substantial surge in open-interest in the March NYMEX contracts. My guess du jour is that there were a lot of people scrambling to roll their contracts out of February because of the situation in Cushing. At some point, I expect those contracts will be liquidated but we'll have to wait and see when/if that happens.

Levered ETFs and Real-World Volatility

When I started this blog, I never really expected to get much traffic and that basically has been the case. Still, I'm no different than most and retain some vanity and so installed the obligatory traffic tracker to see where people are coming from. I was surprised to discover that the pieces on DXO, DTO, and oil generally have been far and away the most popular. I guess I expected that after the trouncing that oil took that it would have been a dead area of little interest. But within the oil searches there was a vein of interest in levered ETFs themselves. That part I'll expand on a bit further here.

There are lots of levered ETFs out there but they all function in more or less the same way - they try to provide +/-2x the percentage daily move of whatever its benchmark index is. Just to take one example, this is from the ProShares prospectus (page 9):

Ultra ProShares are designed to correspond to a multiple of the daily performance of an underlying index. Short ProShares are designed to correspond to the inverse of the daily performance or twice (200%) the inverse of the daily performance of an underlying index.
The Funds do not seek to achieve their stated investment objective over a period of time greater than one day.

(emphasis mine)

These are very important caveats to bear in mind and they are in the prospectus to protect ProShares from real-world behaviour vs. uninformed investor expectations. So how can these ETFs be used effectively, given their path dependence? The post on DXO/DTO already showed that there are occasions when the ultra or levered ETFs perform outstandingly well - basically when the underlying benchmark moves in a straight line with little deviation. So what of the flip-side? I skirted this issue with an earlier post called "The Chop!" where it was noted that SPX had been exceedingly volatile on an intraday basis for the previous couple of months. I'll try to give a look at both possibilities.

First, you have to have a basis for modeling or in this case, generating random but similar paths. I took 100 days of SPX data ending on 1/16/09 and noted the open-close price percentage change in the index. Here's that daily price change distribution in a nice histogram. This isn't exactly a normal distribution and for the stats nuts, here is the description: mean=-0.308, std error=0.359, std dev=3.59, kurtosis=0.839, skew=0.178. Unfortunately, I'm somewhat limited by knowledge of JMP and thus confined more or less to what Excel has on offer. So I took a short-cut and decided to pretend this was a normal distribution anyway. I'll accept any help to improve on this assumption so critique away.

To ensure that the normal distribution assumption wasn't wildly off, I decided to randomly generate 5 sets of n=100 based on the mean and std. dev. noted above using Excel. Then these sets were compared to the actual data set. These results are to the left in the JMP graphic. (I apologize for the squashed look but it's a tall graphic.) They aren't dead-on, but statistically speaking they are pretty similar to the original set. (Set #1 in this case is the original SPX data.) Not great, but not terrible.

Moving ahead from this point seemed acceptable and so I decided to model the returns (before expenses) according to the daily goals outlined in the prospectus of 1x, +2x and -2x funds in this environment. I ran this 200 times with some interesting results.

The 1x is the basis for daily calculation of the levered models. But the table on the right is the final return from day 1 to day 100. For the use of these Ultra-ETFs to be ideal, one would hope to see a high probability that the performance would be beyond the 2x daily expectation. Looking just at the mean, one might be tempted to think that these funds are excellent performers but when the median performance is looked at, a different story emerges. Better than 50% (about 55%) of the time, the -2x fund "underperforms" in a choppy market. In this instance, "underperform" simply means less than +/-2x the target benchmark - it says nothing of its performance to prospectus specifications. For reference the actual performances during the period under analysis were SPY= -32.8%, SSO= -60.73%, SDS = 19.14%.

So that's the examination of a relatively choppy market. What about the performance under circumstances like the summer to fall period in crude oil?

You'll have to just grant me the same assumptions as the charts illustrated above because I don't want to further clutter up the space. Briefly, the sets generated are even more similar statistically than in the first case. If you would like to see the numbers from above let me know but I'll summarize with this: the data in question is USO from June 16 until December 24 or 135 sessions. The average intraday change was -0.64% with a std. dev. of 2.65. This is a couple of weeks before the peak, and seems a little more realistic in terms of market timing in that allows some wiggle room for being off.

The results are to the right in the same table. In this environment, the -2x modelled fund performed well even on a comparison of medians and in the simulated runs actually beat a 2x performance of the underlying benchmark 81% of the time.

So what can be drawn from this overly-long post? First, it should be remembered that this simulating is totally ideal - no expense ratio (generally >0.8%) and no slippage/remainder that seems to occur in terms of performance vs. stated goal on a daily basis. My hunch is that on significant volume/volatile days the NAV of these ETFs gets away from share price due to in/out-flows and eventually must be reconciled. But in terms of applicable trading, I suppose if you believe the benchmark that you are targeting is going to enter a period of strong directional movement without much volatility, then a levered ETF might be a decent wager. But beware the chop and churn that, combined with expenses and remainders, can systematically erode returns.

Oil Anomalies

Interesting things in the world of crude are afoot! The spread between the February (front month) and March contracts reached a fairly impressive mark of $8.14 on Thursday before falling back to a still high $6.06/bbl. Pretty impressive. The chart to the right shows what the normal spread between these two contracts has been in the past year and a half. Up until a point in early summer of 2008, crude prices were actually in backwardation or had a flat spread. What we see now is a massively changed market reflecting contango (or super-contango as I have seen it written recently). The 6-month and 1-year spread between contracts closed on Friday at $16.28/bbl and $22.59/bbl respectively. These are very unusual numbers that will be revisited.

The next somewhat anomalous thing is that open interest in the March contracts has blown up in the past week. Since oil crested this past summer, the monthly peak open interest has been falling backwards on a year-over-year basis pretty steadily since. I talked about this a bit in my first oil post. The open interest had seemed to be behaving in a somewhat predictable manner until the last week when I noticed the March contract's open interest rocketing up to last year's peak level for the March contract. Just predicting based on monthly growth from last year a jump in peak open interest to perhaps 325K contracts. Instead, the Feb contract has not yet been assigned and there are already 382K contracts available for trading - roughly equivalent to last year's and a 33% jump from last month's peak.

So what is going on here? First, the NYMEX WTI contract is delivered at Cushing and Cushing is functionally full with negligible storage remaining. The belief is that this is artificially depressing prices. There are numerous commentators pointing out the divergence between WTI and Brent pricing, claiming this as further evidence that WTI is not representative of "real" market prices. I'm not quite as satisfied with this explanation. Here's the chart of spot prices for WTI and Brent since 2005. The entire set from PRI (see link to the right) goes back to 1987 and during that time WTI generally traded at roughly $1.20/bbl premium to Brent. Further, there are clearly long periods where Brent trades at a premium to WTI. Being an amateur, I don't really know what the significance is but this price divergence just doesn't appear all the significant when put in context with the past years. Articles such as this one by Chris Cook (former IPE compliance guy) make me a little skeptical about how much weight the Brent prices should carry but also questions the usefulness of WTI at NYMEX.

Anyhow, there have been lots of articles about the Brent vs. WTI pricing (see links at bottom) and several mentioning the super-contango market currently in effect. There have also been lots of stories about oil speculators leasing out freighters in order to create floating storage. (Also in the links at bottom - FRO is mentioned specifically.) I found this interesting because this is not the first time this type of oil "arbitrage" has been thought up. Prior to the oil bubble being popped, none other than Morgan Stanley had ahem, floated this scheme in May 2007. How that worked out for them, I have no idea. So the resurgence of these reports is interesting to say the least. But the one thing I've not seen mentioned anywhere is the surge in open interest in the March contracts. On the one hand, it would seem to point to a speculative fervor, though it is unclear if that will translate to higher prices. It will be quite interesting to see if the March contracts fall once Febs are assigned, given the yawning gap. I suspect it will also be worth paying attention to when all those extra contracts start to get liquidated. After all, oil is still subject to supply and demand no matter how quirky the market pricing for a period of time. I would welcome any comments or thoughts on what this large spike represents.

Oil News Links:

- FT: It's All About Cushing

- MarketWatch: Crude super contango...

- Climateer Investing: Oil: Record Inventories at Cushing

Oil Updated (DXO & DTO)

Since I wrote my last post on oil, NYMEX crude dropped from 43.60/bbl to 35.35/bbl before recovering yesterday to close at 44.60/bbl. USO (crude oil ETF) dropped from about 36 to a low of 27.73 and finally closing yesterday at 33.10 with an intraday high of 34.87. It would seem that I was a bit premature in my statement of a bottom in crude prices forming. However, despite the drop below the Dec. 16th closing I would still maintain that crude is in the process of forming a level bottom.

Anyone that follows ETFs is aware that some funds attract more volume than others and many should just be taken out back and shot to be put out of their misery. Crude oil has several ETFs available in the ultra and ultrashort forms but DXO and DTO from Powershares got a bit of a headstart on their counterparts UCO/SCO from ProShares. By dumb luck, the launch of DXO and DTO almost perfectly coincides with the summer peak of crude oil prices. Let's look at some charts:

Most people I know don't actually trade crude on NYMEX so the next best thing available is USO - the crude oil ETF. The behaviour of USO is a little different than a 1:1 correlation due to the contango/rolling effects that impacts the fund. Anyway, USO peaked on July 11th at 119.17 before closing that day at 117.39. From that price to the close yesterday represents a -71.8% loss in value. It would have been even worse had oil not rallied so much on... geopolitical instability perhaps? (More on that later.) Let's move on to DXO and DTO.

DXO is a double-long ETN and DTO is its double-short companion. (I'll leave it to the reader to research the differences between ETNs vs. ETFs) Both of these began trading on 6/23/08 when USO closed at 110.92. DXO opened that day at 24.50 and DTO opened 25.29. Interestingly, there was a significant volume disparity with DXO trading a paltry 1900 shares to DTO's 525k. Very interesting considering what happens next...

Take a look at the 13 period MAs in the volume graph. (Red lines) During the run up to oil's peak, average volume moves up for DTO and maintains a fairly steady pace through September when the volume begins to taper off to its current low levels. DXO on the other hand is a mirror image. Volume is rather dead until very late November when the MA(13) on volume begins to register a few blips of a pulse and then picks up real vigor into December as crude prices continued to plunge and DXO's price, from a summer high of 29.65 tumbled -94% to a low of 1.76 on Dec. 26th.

I believe that the mirror image volume behaviour represents the crude high and low reasonably well. Further, DXO - particularly if you were wise/lucky enough to pick it up around $2 -represents a low-risk entry point.

A bit of a digression: I am aware that double-long/short ETF suffer from value destruction during volatile markets due to their daily rebalancing required to meet their goals of double daily percentage moves vs their respective benchmarks. However, when markets move in a somewhat linear fashion - as crude did falling from its summer peak - the compounded gains using the levered ETFs can be very rewarding. As evidence of that, USO fell -71.8% from its peak and its prospectus states that it attempts to replicate WTI NYMEX pricing. DXO fell -94%. DTO on the other hand was a true winner from its inception, with an all-time high closing price of 160.30 representing a gain of 533%. Shorting USO or DXO would not have delivered these gains.

So, assuming the traders in these ETFs know what they are doing, this would be a signal that crude is nearly at a bottom.

Briefly revisiting the geopolitical instability comment above and at risk of sounding like one of the tin-foil hat fashionistas, I will say this: the governments of Iran, Russia, Venezuela and various other oil exporting states had strengthened themselves via oil revenues and developed budgets with assumptions of oil at a far higher price than the current $40/bbl area. I do not believe it is a big leap to connect the dots between Iran, Hamas and a risk-premium that seems to attach to crude whenever some instability pops up in the Middle-East. My guess is that we will see more of this type of stuff going on in the next year as various governments in these oil exporting states try to maneuver under the pressure of reduced oil revenues. Some of these maneuvers will serve as attempts to drive up the price of crude. Others will simply be defaults like Ecuador's Correa. Should be a fun year!

Sources:
USO Chart (Yahoo!)
DXO Chart (Yahoo!)
DTO Chart (Yahoo!)

Oil!

Is there any other commodity that nearly everyone in and out of the market has an opinion on? It’s almost enough that I hesitate to add my small voice into the howl of market opinion. But that’s never really stopped me…

Other websites and news outlets have already reported on vehicle miles traveled. Here are two charts showing vehicle miles traveled (VMT) both on a rolling 12 month basis and with the price of crude since 1986. There is clearly a regular monthly variation and this year is no different in that regard.

This year, the percentage increase from September to October in VMT was 7.27%, as opposed to an average of 5.62% in the 5 years prior, which certainly would indicate that the plunge in gasoline prices has allowed for more driving. The year-over-year percentage change level has not been seen in the data series since the late ‘70s timeframe, which is saying something in this discussion.

But obviously the VMT is not the majority of the story, considering oil continued to accelerate upward for a few years even as the year-over-year mileage driven started to slump. For a second piece of the oil price puzzle, NYMEX has to be checked.

Since last September, I have been tracking open interest of crude contracts on NYMEX since there is no source of this historical data that I was able to locate. Once a couple of months of data were collected, the pattern in open interest was apparent as seen in this chart:

Peak open interest since last May has fallen off considerably on a year-over-year basis reflecting liquidation of a portion of these contracts rather than the continued rolling of them from month to month. Back in July of 2007, I was fortunate to stumble upon Hussman’s article titled, “The Outlook For Inflation and the Likelihood of $60 Oil.” In it, he stated:
“In my view, the problem will emerge a few months from now, as a) economic demand softens further, b) planned production hikes actually emerge, and c) weakening price momentum encourages speculators to close long positions instead of rolling them forward. At that point, I expect that net speculative positions will plunge by 10-15% of open interest and we'll see a sudden glut on the market for spot delivery. It should not be surprising if this speculative unwinding takes the price of crude below $60 a barrel by early next year.”

This statement was made when oil was still above $130/bbl and so really is impressive in my mind. But the interesting thing is that the open interest didn’t take until the fall to unwind – it actually began in July. The chart here shows the peak open interest levels and the year-to-year percentage change. The February contract peak should be occurring in the next week as the January contract nears assignment. I don't expect that there will be significant additional liquidation.

Additionally, there is the subject of contango vs. backwardation. For almost the entirety of the run up in crude oil pricing, the spread between the front month contract and 6 months out was negative (backwardation) meaning that delivery sooner was worth more than delivery later. However, as the peak in oil price approached this 6 month spread started to flatten out and eventually switched over to contango. Last week, the 1 year spread in contract pricing reached $15/bbl meaning that you could buy oil for delivery in the near month, stick the oil in a tank and sell a contract to deliver that oil in one year’s time for $15/bbl more than you paid – a decent return assuming warehousing costs are not too much. At this point though, it looks like this spread is narrowing a bit which would imply some increase in price or at least a bottom.

Another item worth mentioning, the couple of oil stocks that I track in my oil spreadsheet – XOM & RIG- have been moving up in the last weeks, even as crude has fallen. RIG, a bit more than XOM, has generally traded in anticipation of crude price movement. The caveat here is that RIG has been crushed from its high and still apparently has its capacity contracted for 4 years so some of the rebound could simply be reflective of this fundamental economic health. (Peak oil or not, oil is almost indisputably getting more difficult to lift out of the earth.)

Finally, there is the weakening dollar that has recently breached a resistance level. Where this ends up, I’m not sure considering that there aren’t many other regions that are in considerably better fiscal health than the US. During the run up, it seems evident that using commodities as a store of value against dollar depreciation was a typical trade. However, it also seems likely that many of these players forgot that oil, like all commodities, is subject to supply and demand considerations and if demand side collapses so too will the price.

At any rate, this is all a rather long-winded way of making my case for oil having formed at least a mid-term bottom. I hate to be on the same side as GS on this but there it is.

NYMEX Charts

This page will be updated semi-regularly - probably daily.
Last Update: 5/01/09 @ 5:30PM