The Skew: Why Downside Protection Always Costs More

The Anomaly

Strip all options on a given stock index that share the same expiration date and strip out the implied volatility the market is paying for each one using the Black-Scholes model the standard tool traders use to convert the price of an option into a volatility number. The model assumes there is an underlying asset so it should return a volatility. Each strike should imply the same figure

In practice it doesn't work like that. Lower strikes imply greater volatility than higher strikes consistently in almost every stock index listed. The plot involves volatility against the strike price and the line has a downward slope sometimes steep. That downward slope is what traders call the skew

This is not a rounding error. In stock indices a put option well below the current price and a call option at the same distance above it can trade with implied volatilities separated by ten points or more. Same underlying. Same expiration date. Same number of days for something to go wrong. A very different insurance price depending on which direction it is insured

What the Skew Is Saying

Black-Scholes assumes that stock returns have a lognormal distribution which in simple terms means that the model represents a bell curve symmetrical around the current price with as much room above as below. Stock markets don't behave that way. They tend to rise gradually and fall sharply. A 20 percent drop in a single month occurs periodically and everyone remembers it. A 20 percent rise in one month is so rare that when it happens peopleHe's still talking about it years later

Bias is the market correcting that mismatch within a model that was never built to support it. Options that pay off in a crisis are priced as if the left tail of the return distribution were wider than the bell curve says it should be because it is. The model only has a dial for volatility so the only way to express the idea that an accident is more likely than the model thinks is turning that dial on specifically for strikes that deliver results when an accident occurs

The Demand Side

Price theory explains why the correction manifests itself as a change in volatility and not through some other channel. It does not fully explain the specific shape of the curve because there is also a simple supply and demand story underneath it

Pension funds insurers and long-term asset managers collectively own huge portfolios of stocks and a significant proportion of them want insurance against a downside. They buy puts quarter after quarter largely regardless of price. On the other side of that trade there is no equivalent population of investors who want to buy puts on the same scale. The persistent unilateral demand for downside protection pushes the price of that protection higher than a symmetrical model would ever produce

Something similar happens in reverse at the top of the ladder. Investors who already own the stock or index often sell calls against their position to earn income a strategy known as covered call writing. That selling adds upside option supply to the market and keeps implied volatility on the buy side lower than it would otherwise be. Structural buying pressure on puts and structural selling pressure on calls point in the same direction and both are built into the bias

Reading the Steepness

Bias is not a fixed number. It intensifies when the market is nervous and flattens when it is calm and traders often watch the direction of that change more closely than its absolute level in a single day

The reversal in which call options trade at a higher implied volatility than put options is rare in broad indices and much more common in individual stocks caught up in takeover speculation. When a buy rumor begins to circulate around a company the options that pay off if the deal goes through the options made above the current price suddenly become expensive and the entire curve flips

The Smile Versus the Smirk

Currency pairs and many commodities show a completely different shape called smile where both far strikes trade with higher implied volatility than the intermediate strikes. Those markets can move wildly in either direction a currency can crash or soar oil can do the same so both ends of the distribution carry a significant premium

Stocks often show a one-sided version of that curve sometimes called a smirk instead of a smile because fear is genuinely one-sided. Hardly anyone buys protection against their own portfolio growing too quickly. If that happens they're just happy

A Worked Example: Pricing the Skew

Here's what bias actually costs made with a clear illustrative example rather than actual quotes specific to a trading desk. Suppose a stock index is around 5,000 and consider options that expire in three months. To isolate the effect of bias itself this example ignores interest rates and dividends. They are important on a real pricing desk but here they would only add noise without changing the point

Suppose the market is quoting the following implied volatilities on the downward-sloping strike scale as a skew is actually seen in an equity index and consider what a standard Black-Scholes calculation converts each of them to:

Now let's ask the counterfactual question that a flat volatility model would answer. If the same 16 percent of at-the-money implied volatility were applied to each strike holding everything else fixed how much would the same put options cost?

The gap is the whole story. The 10 percent of the money put option costs $59.44 on the true and skewed surface and would cost only $16.65 if volatility were flat on each strike a difference of $42.79 on a single option or about 3.6 times the flat surface price. The 5 percent of the money put shows the same pattern on a smaller scale $94.40 versus $61.44 or so.1.5 times. Move over to the calls side and the sign changes: The 5,250 and 5,500 calls cost cheaper on the true and biased surface than they would on a 16 percent flat surface because selling covered calls keeps their implied volatility low

Expand that to something closer to a real hedging decision. Suppose a pension fund has $500 million of stock index exposure and wants to buy three months 10 percent of the money to protect it. Dividing $500 million by the index level of 5,000 gives 100,000 units of exposure for hedging (real index option contracts have a fixed multiplier but what matters here is the rationot the mechanics of a single contract).At the actual skewed price of $59.44 per unit that protection costs about $5.94 million for the quarter or about 1.19 percent of the portfolio value. At the hypothetical fixed price of $16.65 per unit the same protection would cost about $1.67 million or about 0.33 percent. The bias alone nothing else in the market changed adds about$4.28 million to the hedging bill for that quarter alone about 0.86 percentage points of portfolio value paid simply because the call and put options are not priced the same

That gap is not mispricing waiting to be arbitraged away. It has persisted for decades because it reflects something real about how index returns actually behave which is the topic of the next section

Case Study: The 1987 Crash and the Birth of the Skew

The clearest evidence that the bias reflects a real revaluation of risk rather than a permanent peculiarity of market structure is that it has not always existed. Before October 1987 the implied volatility surface in S&P 500 index options remained almost flat in all exercises. Call and put options at different distances from the money traded with very similar implied volatilities more or less consistent with what Black-Scholes actually assumes

On October 19 1987 since remembered as Black Monday the Dow Jones Industrial Average fell about 22 percent in a single trading session still the worst one-day percentage drop in the index's history. The S&P 500 fell about 20 percent that same day. A move of that magnitude in a single session was not supposed to occur under a lognormal return distribution outside of an astronomically unlikely tail event. It happened.in public anyway watched by all the traders who would continue pricing options for the rest of their careers

Part of what made the crisis worse according to subsequent analyzes including the government's Brady Commission report was a strategy called portfolio insurance which used dynamic hedging selling futures on stock indexes as prices fell to synthetically replicate a put option without actually buying it. When prices began to fall that mechanical selling added more selling pressure to an already falling market which then triggered more of the same selling in a feedback loop. The strategy built toMimicking drop protection ended up amplifying the very fall it was designed to protect against

After the crisis the options market never returned to a flat surface. Traders had just seen a 20 percent move in one day in front of them and they permanently repriced the probability of it happening again. Since then out-of-the-money puts have had a persistent premium over equally distant calls a pattern some market historians call crashophobia. In other words bias is not an abstract mathematical curiosity. It is scar tissue that formed on a specific date andIt never completely disappeared

Why It Matters for Anything Practical

Two consequences follow directly from all of the above. The first is that hedging costs more than a model based on the volatility of money would suggest and the gap between that naive estimate and the actual bill widens exactly when protection is most needed which is exactly the wrong time to discover it

The second is that any strategy based on selling low strike price options is charging a high premium for a specific reason not for free. That extra income compensates the seller for real exposure and the exposure is concentrated precisely in the scenario in which the rest of a portfolio also falls. Charging a rich premium quarter after quarter can seem for a long period like money for nothing. It is not. It is a payment for taking on a risk that rarely appears and is serious when it does

Where This Breaks: The Skew Is Not a Free Lunch

It would be easy to read the last section as an argument to permanently sell short puts to take advantage of the skewed premium. Take the other side seriously before doing that because the arguments for caution are at least as strong as the arguments for collecting the premium

The distribution is truly asymmetrical which is the only reason the skew exists in the first place. Selling a put option that is priced high because the market fears a crash is by definition taking the flip side of that fear. If the crash never comes the strategy earns its premium and looks slick for a long time.can earn more than the bonus received) and is effectively unlocked on the downside

This is not a hypothetical risk. Strategies built around persistently selling stocks' downside protection to collect the skewed premium have suffered rapid outsized losses more than once most visibly during the 2008 financial crisis and again during the rapid sharp sell-off in March 2020 when years of collected premiums were erased in a matter of days. The lesson from those episodes is not that the skewed premium is false. It's that the premium is a trade-off.genuine for a risk that is rare and serious correlated with everything else going wrong at the same time and much more difficult to manage in practice than a backtest conducted on a quiet stretch of history would suggest

There is a second quieter way that can be misleading. The entire framework above assumes that skew is a reasonably stable characteristic that can be measured and relied upon on a day-to-day basis. During true market stress liquidity in out-of-the-money options can evaporate bid-ask spreads widen considerably and the implied volatility at which a trader can trade can be much worse than what is displayed on a screen. The deviation that can be observed and the deviation that can be traded at a reasonable size is notThey are always the same deviation

How I Actually Use This

My read after spending time comparing quoted bias between different index maturities is that the absolute level of bias matters less than many retail comments suggest and the change in that level matters more

The way I would actually use this is as a sentiment indicator rather than a trading signal on its own. A bias that is becoming pronounced while the underlying index is still near its highs tells me that institutional buyers are quietly looking for protection before anything obviously bad has manifested itself in the price itself which is a very different situation from a bias that is pronounced because the index has already fallen and everyone is buying put options after the fact. The first case is genuinely informative. The second is mainly confirmation bias.what the price already told you

I would also be wary of treating a pronounced skew as evidence that puts are overvalued in some exploitable sense. It's tempting to look at 24 percent implied volatility on a put next to 13 percent implied volatility on an equidistant call and conclude that the put must be the better put. Sometimes that reading is correct. Often it's simply the market that correctly prices a real asymmetry and the trade only seems free until the quarter when it doesn't.it is

The Bottom Line

The bias exists because stock returns are not symmetrical and because the demand for downside protection is structurally one-sided: pension funds and insurers buy puts quarter after quarter largely regardless of price and no comparable population buys puts on the same scale. It is the market writing its own correction to a model that assumes a bell curve directly on the price grid one point of implied volatility at a time. The crisis of 1987 is the clearest evidence that this reviewA simple 10 percent three-month currency hedge can cost multiples of what a flat volatility model would adding real measurable dollars to the cost of protecting a portfolio. Selling that protection generates a real premium but the premium is payment for a risk that is rare and serious and has put an end toMore of a popular volatility selling strategy in a single bad stretch. Steepness of the Curve remains one of the few sentiment readings drawn from real money changing hands rather than a survey which is exactly why it's worth following although it will never tell the full story on its own