It’s been a very turbulent year for traders in the popular iPath S&P 500 VIX Short-Term Futures ETN (VXX) with shares rallying by over 200% on a year-to-date basis.
Over the past few weeks, I’ve written articles explaining why I both shorted VXX as well as doubled down on the position with more favorable prices. In this piece, I will dig into the VIX’s behavior around historic crashes as well as discuss the long-run implications of VXX’s methodology. I am still very bearish VXX and believe that even if markets continue declining, we are likely going to see VXX fall in the coming months and quarters.
There has been no question that the VIX has been on a tear this year. As seen in the following chart, we have seen the outright level of the volatility index increase by several hundred percentage points.
This certainly is noteworthy and correlates with the record-breaking short-term decline in the S&P 500. However, if you carefully study historic crises, I believe the data indicates that in general, the VIX normally only makes one push into highs in any given crisis before slowing recovering. In other words, if you are trading VXX in hopes of capturing further record swings in the market, the movement has likely already been made and downside is in store.
Let’s start with the infamous financial crisis. I was actively trading during 2008-2009, and I distinctly remember a sense of doom hanging over the financial markets. Headlines were coming in every day that the end truly was near and the potential collapse of our economic system as we know it was in the works. Markets seemed to decline week after week and times truly were scary.
However, throughout this selloff, the “fear index” (aka, the VIX), actually only made one climactic move and then generally trended lower.
As you can see in the chart above, the VIX itself peaked in the October-November timeframe of 2008. It is important to note that this actually did not correlate to the market bottom. Numerically, the market still declined by an astronomical 30% between the peak level of the VIX in October to the actual low in the market. In other words, just because we are in a bear market doesn’t mean the VIX will always go up in conjunction with downswings in the market.
It is important to keep this relationship in mind depending on your forecast for the future movements in the S&P 500. If you are projecting that we are going to enter a prolonged bear market which will stretch several quarters, then history would indicate that the VIX will likely decline from here as seen in 2008. There is only so much “shock” that can be priced into the market before implied volatility levels decline and bearishness becomes the new norm.
However, if you are expecting the market to be nearing a bottom and the recovery to look more “V-shaped”, then this likely means that the VIX has almost certainly bottomed. For a good comparison example, here’s the VIX’s performance during the debt-ceiling crisis of 2011.
This was another time of turbulent financial history in that there was great uncertainty regarding the functionality of the government going forward. During this time period we witnessed the S&P 500 decline by around 20% during the space of a few short months. The VIX rocketed in response, but before the crisis had fully recovered, the rally began to taper off.
This is the key point here: the VIX has recently made a historic move. During previous historic spikes in the VIX, we almost always tend to reach peak VIX levels at the beginning of the decline in the S&P 500 and then it tends to taper off from there. Even if you are expecting a prolonged bear market, the short history that we have of the VIX (less than 30 years of data) indicates that in general, you see peak-VIX levels at the beginning of a crisis and then it tends to taper off through time.
This is why I suggest investors short the VIX. Even if we continue to see a selloff in the market, the odds generally favor a decline from here. I’ve used this chart many times before, but it really is a gold one in that it captures the directional probabilities of the VIX given a specific level.
I’ve limited the data range to be around the typical mean, but the relationship expands out each way. Essentially, the higher the level of the VIX, the greater the odds that it will decline into the future. Since we are at very high levels in the VIX, the odds are fairly slim that we will see the VIX continue to increase from here.
Another way we can look at the data is to glance at the average change in the VIX by the outright level of the index.
Again, the numbers are logical and clear: the higher the VIX, the greater the average drop into the future.
These numbers are robust (IE, they contain 27 years of VIX data and aren’t cherry-picked). This dataset covers both bull and bear markets. And the data is strongly suggesting that we are going to continue to see the VIX collapse from here.
All this said, there’s another force at work beneath the surface which suggests that VXX is going to fall from here: roll yield. In my previous pieces I’ve covered this in depth, so I’ll try and be concise here (but don’t hold me to it). I do feel that any investor in VXX need a firm grasp of this concept to truly understand what is driving returns in this product.
Let’s look at a long-term chart of VXX. What do you see?
I see a chart which declines almost all of the time. That is, over long periods of time, there are brief weeks in which VXX increases in value punctuated by years of crushing returns. If I thought that by purchasing VXX I was gaining exposure to the VIX, then I would have to think again. For example, here’s an actual chart of the VIX during this same time period.
Do you notice the stark difference? During this time window shown, VXX is down about 90% whereas the VIX itself is up probably 300%. In other words, there’s a material difference between the returns of the VIX itself and an ETP which purportedly gives exposure to the VIX. This difference is due to roll yield. And in my experience having interacted with numerous traders and investors in VXX, roll yield is one of the least understood factors driving the long-run returns of VXX.
Put simply, roll yield is what you get when you’re holding exposure along a futures curve and futures prices move towards spot (known as convergence). As you can see from this wonderful chart from VIX Central, VIX futures are almost always in contango (front of the curve is cheaper than the back of the curve).
What this tangibly means is that most of the time, VIX futures are generally in about 5-10% of contango. This tangibly means that futures increase in value along the futures curve (IE, later expiration dates = higher prices and the front month futures contract is above spot VIX).
To see this another way, here’s a graphic which shows the average level of the VIX during a certain day of a trading month compared to the average level of the front two VIX futures contracts over the last 10 years.
If you notice, there’s a few very clear trends at work in this data. First off, the VIX tends to go nowhere. This is to be expected if we’ve studied the VIX before – it really tends to stay around the 15-20 level and movements beyond this tend to revert in fairly short order.
But there’s another trend at work – the futures contracts tend to start a month at a higher price than the VIX and then gradually decline in value as the month progresses. This is precisely why VXX loses money over almost all lengthy time periods. This is the concept of roll yield (futures converging to spot) in one chart created from directly from real market data.
You see, VXX is holding exposure in these two futures contracts. This is because VXX follows the S&P 500 VIX Short-Term Futures Index. I have tried to be gracious in the past in reference to this index, but put simply, it is an index which has destroyed a lot of wealth from investors unaware of its methodology. From a high level standpoint, this index has shaved roughly half of its value off per year over the last decade.
What is particularly poignant here is that at present, according to ETF.com there is over $3 billion in funds allocated to just the top two strategies which either follow this index (VXX) or give a double-leveraged return on this index (TVIX). Unless the majority of these funds are targeting very short-term returns in these indices, roll yield is ultimately going to transfer a lot of this wealth from the hands of these traders into the hands of someone else.
If the forces that be were to appoint me Czar of the Universe for a day, one of the first things I would do would be issue a requirement that all traders of these volatility ETPs sign a waiver indicating that they understand roll yield and its long-run implications. In my numerous dialogs with traders and investors in these notes, I have come to the conclusion that a decent segment of traders are unaware of the either the long-run return of the instrument they are in or the forces which drive the returns. I personally didn’t understand these instruments in that the first time I saw the exponential decrease in the value of VXX, I thought I was just seeing an error in my charting software. Nope, the decline is real. Here is the actual full return of VXX. Do you still want to buy this?
This is the key concept around which I am short VXX: roll yield is a long-term wealth destroyer of holdings. What this tangibly means is that as strategic traders in VXX, we should look to initiate short trades when the VIX rallies, with a target date several months into the future to capture returns. To avoid risk, we should utilize the active and liquid options contracts and in my experience, put spreads enable investors to strip out implied volatility and put on trades with good reward/risk ratios. In previous articles, investors asked for more details regarding my positions, so in the following paragraphs I’ll detail my core plays.
I have two core positions – the first position is in early 2021 and its breakeven is basically at today’s traded price. My idea for this trade is that since the VIX tends to peak early on in the crisis, even if we see a prolonged bear market, VXX is likely to fall somewhat from here. For that reason, this trade is a put spread using an at-the-money, early 2021 put offset by selling a put a few dollars out of the money. My reward per risk on this trade is about 2:1.
And my other core position is in the 2022 timeframe and it is fairly far out of the money and it relies on the long-run forces of roll yield to drag down returns once again. Prior to this crisis, the 10-year annualized rate of return was over a 50% drop per year. If you do the math, this implies that two years from now, the return would be that VXX will drop by around 75%. You can verify through historic price data that this magnitude of decline actually is experienced across two year periods with a good degree of regularity.
Put simply, I am looking to capture the immediate downside in the VIX as peak VIX levels tend to be reached in the first part of the crisis with an at-the-money put spread in 2021. And I’m also looking to capture the long-run roll-yield-induced losses with options out in mid-2022 which are fairly far out of the money.
For investors looking to take the short trade, I recommend either of the above as sound trades built around VIX tendencies – both in the short and long run. In previous articles I have had individuals express a good degree of skepticism regarding this trade, with some calling it reckless. But my rebuttal would be that you should be closely managing risk in all of your positions.
The total risk on these trades is less than 2% of my portfolio and I would suggest that investors be cognizant of risk as well. My personal opinion is that we should never put ourselves in a position in which a single trade can sink us – and even though I’m confident in this trade, I can’t advocate ultimate portfolio-level recklessness. In all situations, I would advise against an outright short on VXX due to the inability to truly manage risk should the VIX gap higher. But all in all, I believe that the VIX has peaked in this crisis and roll yield will pull VXX down from here.
Historic movements indicate that the VIX tends to peak several months before a bear market ends. Roll yield has a history of eroding returns in VXX and this relationship will continue when markets normalize. I am short the VIX in two core positions which capture two distinct themes.
Disclosure: I am/we are short VXX. I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it (other than from Seeking Alpha). I have no business relationship with any company whose stock is mentioned in this article.