The Cost of Contango—It’s Not the Daily Roll


Updated: Sep 27th, 2016 | Vance Harwood

It’s well known that long volatility Exchange Traded Products (ETPs) like VXX, UVXY, and TVIX often experience devastating losses during market quiet spells—even when the value of the VIX is staying relatively stable.   These heavy losses occur when the VIX futures that underlie these funds are in a price/time arrangement called contango.   The chart below shows an example of VIX futures in a contango configuration.

ts-16sep16

The blue dots show the prices of various futures and the horizontal scale indicates the month of expiration.  The horizontal green line shows the current VIX price— also known as the “spot” price.  You can’t tell it from the chart, but in this example the leftmost future has 4 days until expiration.  At expiration, a VIX future’s value will be very close to the VIX spot price.

When futures are in contango the longer the future has until expiration the higher its price.

If you were to take a time lapse video of this chart over time with a stable VIX you would see the blue dots moving down the blue line, eventually intersecting with the green VIX line at expiration.

Volatility Exchange Traded Funds Roll Futures

In contrast to VIX futures contracts, volatility ETPs don’t expire; instead, they hold a mix of VIX futures (the two leftmost ones for VXX, UVXY, and TVIX) that’s adjusted every day in order to keep the average time to expiration of the portfolio at 30 days.  In this daily transaction some of the leftmost, next to expire futures (I’ll call them M1), are sold and longer dated futures (M2) are purchased.   In futures parlance, this is called a “roll.”  When the M1 futures expire all the M2 futures become M1 and the roll process starts selling the futures with that expiration date rather than accumulating them.

When the futures are in contango it’s bad news for the price of the ETP’s holding them.  They are buying futures when they are higher on the curve, at a more expensive price, and selling them when they have decayed to a lower price.  Buying high and selling low—not a profitable practice.

The Daily Roll Does Not Reduce the ETP’s Value

There’s a widespread misconception that the roll process itself is the cause of the ETP price erosion—after all, you are buying high and selling low.  But the roll is not the reason the ETP price erodes.

Saying that the roll is the cause is equivalent to saying that selling your old, depreciated car to the dealer and buying a new one is the reason your car dropped in value.   Of course, everyone knows that your car dropped in value as it aged, and this is also the case for VIX futures in contango.  The transactions just shift the depreciation process to a different vehicle.

You might think this is a harmless misunderstanding, and in the grand scheme of things it is trivial, but that doesn’t mean it can’t leave a trail of confusion in its path.

For example, what happens if M1 and M2 stay stable at 17.6 and 18.45 for a couple of days?  If the roll causes decay then we would expect the price of VXX to drop as it sells M1 contracts and buys more expensive M2 contracts—but it doesn’t, as the detailed example later in this post shows.

I used to accept the daily roll explanation for the decay of the long volatility ETPs, but once I did the math it didn’t add up.  The reality is that when the VIX futures are in contango both M1 and M2 tend to drop down towards the VIX price over time.  Given that, it’s not surprising that the long volatility funds that hold these futures tend to drop in price.   The next section shows analytically that the daily roll has nothing to do with the decline—it’s a value-neutral operation.

A Detailed Analysis of VXX’s Cost of Contango

A few years ago I developed a consolidated spreadsheet to organize historic VIX futures data from the CBOE into a single spreadsheet.  Using this spreadsheet I calculated the short (SPVXSTR) and medium term (SPVXMTR) rolling indexes that underlie the various volatility Exchange Traded Products (ETP) like VXX, UVXY, XIV, and ZIV.  The image below shows a small sample comparing my calculations (M1-M2 Short Term Rolling Index) with the official value of the short term index.

M1-M2-Sample

The percentage differences between my index and the official index (e.g., -0.000256%) aren’t cumulative and are probably due to rounding.

Once I verified my index calculations I wanted to look at the nemesis of the long volatility funds like VXX and UVXY—yield losses.  These losses, which can be 5% to 10% per month, occur when the CBOE S&P 500 Volatility Index (VIX) Futures that underlie these ETPs are more expensive than the market, or “spot” price.

I wanted to quantify this loss with my spreadsheet without the noise of everyday volatility moves, so I left the term structure in contango but held the futures prices constant in my spreadsheet from day to day as an experiment. The results were surprising.

M1-M2-OnlyC

My calculations showed no daily roll costs in the index. The 0.01 upticks in the index are due to Treasury bill interest.  Of course having the futures fixed like this is not a realistic situation—as futures near expiration they usually move closer to the spot price.  But the example proves that the specific roll mechanism specified by the index and followed by ETF issuers does not decrease the value of the index on a day to day basis.

The usual explanation for roll costs, discredited in the sample above, asserts that losses are incurred when funds sell cheaper shorter term contracts and buy more expensive longer term contracts every day as specified by the indexes they follow—a sell low, buy high situation.   A closer look illustrates the flaw in this explanation.   Suppose you have a $10 million position after market close on 19-Oct-12— after the contract rolls. Fractional contracts aren’t supported, so the unused money goes in the cash bucket.

day1-value

The next table shows the positions at the end of the next day, neglecting transaction costs and interest, and assuming no change in the futures values.

day2-contracts

The number of contracts changes, but the total value doesn’t.

There’s no doubt that these indexes lose money when the VIX Futures term structure is in contango—so where do the losses come from?

The chart below shows the real root cause:

Nov / Dec VIX Futures vs VIX
Plotting the actual 2012 November and December VIX futures values during a period of contango vs the CBOE’s VIX®, we see that both contracts decline in value over time, eventually converging with the “spot” VIX price at their expiration. With both sets of contracts held by the long volatility funds generally declining it’s not surprising the overall value drops. Since we are typically looking at term structures over the span of multiple months, the small daily “slide” down the curve isn’t noticeable—especially when obscured by the up and down moves in volatility driven by daily stock market fluctuations.

This analysis also explains why the mid-term volatility index SPVXMTR which holds contracts that are 4 to 7 months out also declines in contango situations, even though 2/3rds of the contracts (M5 & M6) aren’t rolled on a day to day basis.

Volatility Futures aren’t always in contango.  If the markets are panicky enough the futures contracts get less expense than the VIX index.   The chart below from VIX Central shows the June 10, 2010 situation when the VIX index closed at 36.57.

TS-7Jun10

In this configuration, called backwardation, the long volatility funds have the wind at their backs, every day the futures they hold are sliding up the curve, getting closer to the spot price.

Knowing the real reason for term structure based losses / gains hasn’t changed my volatility investment strategy, but it has removed one source of confusion in understanding the daily moves.

Related Posts



Tuesday, September 27th, 2016 | Vance Harwood
  • Nice! I’ve seen the light, it seems so obvious now. I’m seeing the descending futures in my head… 😉

    Now, to the important part. What’s your volatility strategy?

  • Ardbeg Lagavulin

    Thanks Vance.
    Am I correct that in your last example (backwardation) in order for your logic behind the backwind to work there should be spot VIX higher than M1 and M2?

  • When the market is choppy I see the vol ETPs as just highly leveraged S&P proxies. If volatility has spiked, either due to a correction or an event (e.g., Brexit, FOMC) then I like to take advantage of the volatility collapse afterward.

  • Hi Ardbeg, As you suggest, normally when the VIX futures are in backwardation VIX will be higher than the futures, and the long ETP will be boosted. However, I have seen situations where the VIX has dropped after a market scare and the VIX futures are higher than the VIX and still in backwardation. In this configuration there’s a high risk of the futures dropping rapidly in value.

  • Ardbeg Lagavulin

    Hi Vance,
    Thanks for the explanation.
    Now just to summarize my understanding of the article’s concept: regardless of M1/M2 relation (backwardation or contango) only the value of VIX spot relative to values of M1 and M2 matters. When the spot is higher than the futures then (generally) it’s a backwind for long ETPs and if it’s lower then it is a tailwind. Do I see it correctly? If so, then this bizarrely negates the basic concept of backwardation/contango importance in VIX products’ trading… Not that I am complaining if this works). Thanks again.

  • I wouldn’t say it negates the concepts of backwardation / contango, I just think the convergence to spot is a more fundamental root cause. The term structure generally is a very good predictor of what to expect.

  • Robert Durand

    Hmmm… it looks like you have not understood that holding futures is not like changing cars. When you hold a future contract, you hold a bet that pays off a certain amount per volatility point. By rolling forward you need to pay more to get the same exposure to one volatility point.

  • “daily roll” doesn’t mean the physical act of rolling a tiny bit of your position from M1 to M2, it means the the cost of roll down, or the roll yield of your entire position, as the contangoed futures converge to spot.

    So yes, you are correct, but I hope that most people trading these knew that already.

  • Do you figure daily contango decay per day by dividing it by calendar days (30) or trading days (20). So if contango is 20% it is either 1%/day for 20 days or less for 30 days?

  • No real standards on this, and the decay rate isn’t all that linear, it really accelerates in the last few dates before expiration. VIXcentral.com uses the M1 / M2 slope as the contango rate, which is reasonable, but you could also make a case that the M1 to VIX gap is a better measure. It certainly matters whether you are talking about an individual future or an ETP that is a mix of futures. In that case I have proposed a methodology for calculating monthly and yearly decay rates. https://sixfigureinvesting.com/2016/06/monthly-yearly-decay-rates-contago-losses-vxx-uvxy-vxx-vixy/

    I prefer trading days myself over calendar days.

  • Hi Robert, I certainly get that when in contango the further out future is going to cost more. The point of the article is the roll itself is a value neutral event.

  • Thanks Vance, I was referring to UVXY which is a combo of M1/M2 weighted of course and I also think the M1/spot gap is critical to decay rate (not sure how to figure that into decay) but just wanted your input on a daily estimate of “tailwind” I am getting by shorting UVXY in contango.

  • BTW, Thank’s a million, I have read everything in sixfigure and advise everyone study it before touching these ETN’s/F’s or they will be in for a surprise. Just need a good way to figure daily decay using contango (as listed by vixcentral) and/or M!/Spot gap. If there is a way of combining the two in a spreadsheet or an algorithm type calculator, that would be cool!

  • I’ll take a look to see whether the difference between the appropriately weighted M1/M2 and the spot VIX divided by 30 gives a reasonable tailwind metric for funds like VXX, UVXY, & TVIX.

  • hatee goog

    Hi Vance, did I understand you correctly?
    9/13 vix was lower than the previous day but UVXY was higher than previous day
    8/24 crash 2015, vix on 9/1 was lower than 8/24, but uvxy 9/1 was higher than 8/24 crash,

  • Robert Durand

    It is not neutral “per volatility point”. To get the same exposure to volatility you need to put more money on the table. There is a combined effect. Convergence to VIX and rate of convergence. A plain investment in the front month contract will not return as much as a continuously rolled position from the front to the second contract will.

  • If VXX’s movement is solely tied to mo 1 and mo 2, why is VXX only down .82% and the Dec futures down 1.65% and Jan futures down .80%?. Conversely I have seen days where VXX is down and both M1 and M2 are up.

  • Hi Greenlander, Most/all of the on-line brokers use the wrong closing time for their VXX quotes. VXX”s final quote is driven by the VIX futures settlement time which is is around 4:15PM. For more see https://sixfigureinvesting.com/2013/03/is-etf-etn-broken/

    — Vance