Selling Volatility: The Most Seductive Backtest in Finance

The Backtest

Here is a strategy with a thirty-year track record.

Sell one-month at-the-money put options on the S&P 500, collateralised by Treasury bills. Roll monthly. That is the entire strategy.

The CBOE PutWrite Index (PUT) tracks exactly this approach. From June 1986 through December 2018, it returned 9.54% annually - within touching distance of the S&P 500's 9.80%. But here is the part that catches your eye: it did so with an annualised volatility of 9.95%, compared to the S&P 500's 14.93%. The Sharpe ratio was 0.65 versus 0.49. The maximum drawdown was -32.7%, roughly a third less severe than the S&P 500's -50.9%.

Equity-like returns with bond-like volatility. If that does not get your attention, nothing will.

---

Why The Premium Exists

The volatility risk premium (VRP) is the persistent gap between the volatility the market expects (implied, as priced into options) and the volatility that actually materialises (realised). From 1990 onward, the VIX - a measure of 30-day implied volatility on the S&P 500 - has averaged approximately 19.6%. Over the same horizon, realised volatility has averaged approximately 15.5%. That gap of roughly four percentage points is the premium you can earn by selling options.

Why does it exist? For the same reason home insurance premiums exceed expected payouts. Investors are willing to overpay for downside protection because the pain of a large loss is disproportionate to the pleasure of an equivalent gain. This is not a market inefficiency - it is rational risk transfer. Pension funds, endowments, and risk-averse institutions structurally demand portfolio insurance. Someone has to sell it to them.

Carr and Wu formalised this in their 2009 paper Variance Risk Premiums in the Review of Financial Studies, demonstrating that the risk-neutral expectation of variance consistently exceeds the physical expectation, and that this premium is both statistically significant and time-varying - widening sharply during periods of stress.

The practical upshot: implied volatility overshoots realised volatility on roughly three out of every four trading days. That is a powerful tailwind for anyone on the sell side of the options market.

---

The Mechanics

When you sell a put option, you collect a premium upfront. If the underlying stays above the strike at expiry, the option expires worthless and you keep the full premium. If it falls below, you are obligated to buy at the strike price - absorbing the loss.

The risk profile is asymmetric by design: frequent small wins, occasional large losses. The P&L distribution has negative skew and positive excess kurtosis. In plain English, you make money most months, but the bad months can be very bad.

The Greeks frame this precisely. You are long theta - collecting time decay every day. You are short vega - hurt by rising implied volatility. And critically, you are short gamma, meaning as the market falls your position gets increasingly worse. Delta-hedging can mitigate the directional exposure, but it cannot eliminate the gamma and vega risk. Those are the risks you are actually being paid to bear.

---

February 5, 2018: Volmageddon

For years, selling volatility was the closest thing markets had to a guaranteed trade. The VIX spent most of 2017 below 12. Retail investors piled into products like XIV - a Credit Suisse exchange-traded note that was effectively short VIX futures - and watched it climb almost 200% in two years.

Then, on February 5th, 2018, the VIX spiked 116% in a single session, its largest one-day percentage increase on record. The trigger was unremarkable: a routine equity selloff driven by concerns about rising interest rates. The S&P 500 fell about 4%.

What turned a correction into a catastrophe was a mechanical feedback loop. During the final fifty minutes before the 4:15 PM ET futures close, short-volatility products were forced to buy VIX futures to cover their positions, which pushed futures higher, which triggered further forced buying. XIV - which held a net asset value of roughly $2 billion the previous Friday - lost 97%, dropping from approximately $145 to $4.22 per share. SVXY, a similar ProShares product, lost 91%. Approximately $3 billion in value evaporated in minutes.

Credit Suisse invoked XIV's acceleration clause and terminated the product entirely. The product did not survive a single bad day.

---

March 2020: The Second Wave

Two years later, COVID-19 delivered the next stress test. On March 16th, 2020, the VIX closed at 82.69 - an all-time closing high, surpassing the 2008 financial crisis peak of 80.86. The S&P 500 fell 12% in a single session, its worst day since 1987.

Morgan Stanley observed a pattern: each major volatility event eliminated a different class of vol seller. February 2018 wiped out retail VIX ETP traders. December 2018 knocked out iron condor sellers. March 2020 hit active volatility management strategies. Each shock reduced the supply of people willing to sell insurance, which in turn widened the VRP - making the trade more attractive for those with the capital and risk management to survive.

This is the dark irony of selling volatility: the strategy becomes most profitable immediately after it has destroyed the most accounts.

---

What The Strategy Actually Requires

None of this means selling volatility is a bad strategy. The economic rationale is sound, the premium is real, and it has been documented across decades and asset classes. But the backtest alone does not tell you what you need to know. The critical questions are about risk management and position sizing - and getting those wrong is what separates the survivors from the casualties.

Standard Value at Risk models, which assume roughly normal return distributions, systematically underestimate the tail risk of short-volatility positions. A 99% daily VaR number tells you almost nothing about the event that wipes you out. Expected shortfall is a better tool, but it still relies on the tail distribution being well-estimated - which it rarely is for sold options.

The practical requirements include:

• Position sizing that survives the tail. If a March 2020 scenario would cause losses you cannot tolerate, the position is too large. Full stop.
• Understanding of the Greeks. You need to know your gamma exposure and how quickly losses accelerate as the market moves against you.
• Margin awareness. Brokers raise margin requirements precisely when you can least afford it - during volatility spikes. Getting margin-called out of a position that would have recovered is one of the most common ways this trade fails.
• No leverage illusions. The PutWrite Index is fully collateralised by T-bills. Many real-world blow-ups come from running the same strategy with insufficient collateral.

The probability foundations - particularly conditional probability and fat-tailed distributions - are essential for understanding why naive expected value calculations break down here. And the statistics of non-normal distributions explain why the Sharpe ratio, which assumes normality, flatters short-vol strategies by underweighting exactly the risk that matters.

---

The Lesson

The volatility risk premium is one of the most well-documented phenomena in quantitative finance. It is real, it is persistent, and it has a solid economic rationale rooted in risk transfer between those who need insurance and those willing to provide it. The CBOE PutWrite Index demonstrates that a disciplined, fully-collateralised approach to harvesting it can deliver attractive risk-adjusted returns over long horizons.

But those returns are compensation for bearing a specific, dangerous risk: gap moves and volatility spikes that can exceed anything in your historical dataset. The strategy that generates steady monthly premiums is the same one that can lose 30-50% in a week. If your risk framework does not account for this, the premium you collect is just the market paying you to hold a grenade.

Understanding the mathematics of option pricing, the dynamics of volatility, and the practical realities of portfolio risk is not academic box-ticking - it is the difference between harvesting the premium and being harvested by it.

Quantt covers exactly this ground: the probability theory behind fat tails, the Greeks that drive your daily P&L, the risk models that should govern your sizing, and the pricing frameworks that tell you what the premium is actually worth. If you want to sell volatility, make sure you understand what you are selling.