Every so often someone shows me a long-VIX product on their screen and asks why it keeps going down when the market is basically flat. Nothing dramatic is happening, spot VIX is sitting where it always sits, and yet the position is quietly leaking value day after day. The answer is not a bug and it is not a fee problem. It is baked into what these products have to do to exist. A long-volatility exchange-traded product does not hold the VIX index, because you cannot hold an index that is just a calculated number. It holds VIX futures, and the way it holds them forces it to lose money in exactly the market conditions people spend most of their lives in.
The roll is the whole story
Start with the fact that the VIX index itself is not tradable. What trades are VIX futures, contracts that settle to where the index is expected to be on some future date. A standard short-term long-vol product targets a constant maturity, usually something like a rolling 30-day exposure, by holding a blend of the front two monthly futures. Every single day it sells a slice of the near contract and buys a slice of the further-out contract to keep that 30-day point pinned. That daily buy-high, sell-low mechanic is the roll, and it is the source of the bleed.
Here is why it bleeds. In calm markets the VIX futures curve is in contango, which just means further-dated futures trade at a premium to nearer ones. The market is pricing more uncertainty three months out than one month out, which is usually rational. So the product is continuously selling the cheaper near contract and buying the more expensive far contract. As each day passes, the far contract it bought drifts down toward the lower spot level as it ages, and the product eats that convergence loss. Nobody is stealing the money. The curve is just charging rent on the position, every day, whether or not volatility moves at all.
How to quantify the monthly bleed
You do not need a model to get a usable number here. The rough monthly drag is close to the annualized roll cost of the curve, and you can estimate the roll cost directly from two futures prices. Take the front-month VIX future and the second-month VIX future. The percentage gap between them, over the number of days between their settlements, is your daily roll yield. When the second month trades above the first, that yield is negative for a long holder, and that is your bleed.
A practical way to eyeball it:
- Pull the front two VIX futures prices.
- Compute the percentage difference between them, second minus first, divided by the first.
- Divide that by the roughly 21 trading days that typically separate the two contracts to get a daily figure.
- Multiply back up by about 21 to sanity-check the monthly cost, and by roughly 252 if you want the annualized version.
When contango is steep, the gap between the front two contracts can sit in the mid-single-digit percentages, which quietly compounds into a monthly drag that is far larger than most people expect. Historically, holding a front-month long-vol product across a normal year has been one of the most reliable ways to lose a large share of your capital, precisely because that daily rent almost never stops. The point is not the exact figure. The point is that you can read the cost off the curve before you buy, so the decay is never a surprise.
The narrow window where long vol actually pays
Long-vol products are not broken, they are just specialized instruments that only make sense in a small set of conditions. They pay when two things happen close together. First, spot volatility spikes hard and fast, so the futures you hold jump in price. Second, the curve flips from contango into backwardation, where near-dated futures trade above far-dated ones because the market is pricing acute stress right now. In backwardation the roll actually works in your favor, you are selling the expensive near contract and buying the cheaper far one, so the rent reverses into a tailwind for as long as the panic lasts.
The catch is that backwardation is rare and short-lived. It shows up during genuine dislocations and then collapses back to contango within days or weeks as the market calms down. So the profitable regime for a long-vol product is a thin slice of calendar time, and the losing regime is almost everything else. If you hold through the calm waiting for the spike, the roll drag usually eats more than the eventual spike gives back, unless your timing is unusually good. That asymmetry is why buy-and-hold long vol is a losing structure and why these products are tools for a moment, not positions for a portfolio.
Sizing a tail hedge so the decay does not eat it
If you actually want protection against a crash, the honest framing is that you are buying insurance, and insurance has a premium. The mistake is treating a long-vol product like a hedge you can set and forget, then watching the premium compound until the protection is a rounding error by the time you need it. A few rules of thumb that keep the decay contained:
- Budget the bleed as a line item. Decide in advance what annual percentage of the portfolio you are willing to spend on tail protection, then size the position so its expected roll drag fits inside that budget. If the math says the hedge costs more than you are willing to lose in a calm year, it is too big.
- Prefer convex, defined-cost structures over open-ended long-vol holds. Buying out-of-the-money index puts, or put spreads, gives you a known maximum cost and a payoff that explodes in a crash, without a daily roll grinding against you the whole time.
- Rebalance the hedge against the thing it protects. When markets are calm and the hedge is bleeding, the underlying portfolio is usually gaining, so trim the hedge on strength rather than topping it up. When stress hits and the hedge spikes, that is the moment to monetize part of it, not to hold for the last dollar.
- Watch the curve, not just the spot level. If the front two VIX futures move into backwardation, the character of the position has changed, and that is your signal that the roll has flipped and the hedge is doing its job.
The single most useful habit is to check the shape of the futures curve before you put on any volatility position, because the curve tells you what you will be paying to hold it. On Blockcircle I keep the term structure and the scorecard side by side for exactly this reason, so the roll cost is visible up front rather than something I reconstruct after the fact. None of this makes long vol a good buy-and-hold. It just means that when you do reach for it, you know what the decay is, you have sized around it, and you are holding it for the window where it earns its keep instead of hoping the spike arrives before the rent does.