An Option on an Option: Geske and the Hidden Leverage of Equity

Robert Geske's 1979 paper looks, from the title, like a piece of technical housekeeping: how do you price an option whose underlying asset is itself an option? A curiosity for exotics desks, surely.

It is much more than that. Buried in the paper is an explanation for one of the most robust empirical facts in all of finance, the fact that a company's stock gets more volatile as its price falls, and with it, an early theoretical account of why the volatility smile exists at all. Geske got there by taking one idea from Merton seriously and following it to its logical end.

The problem: Black-Scholes assumes constant volatility, and that cannot be right

Black-Scholes needs one volatility number, fixed for the life of the option. Feed in 25 percent and it stays 25 percent whether the stock is at 100 or at 10.

Merton had already planted a subversive idea: a company's equity is itself an option. Think of it this way. A firm has assets worth some amount, and it owes bondholders a fixed sum at some future date. If the assets are worth more than the debt when it comes due, shareholders pay off the bondholders and keep the surplus. If the assets are worth less, shareholders walk away, hand the firm to the creditors, and lose nothing further (that is what limited liability means). Shareholders therefore hold a call option on the firm's assets, struck at the face value of the debt.

Geske's move was to ask: if the stock is an option on the firm's assets, then a listed stock option is an option on an option. And once you see it that way, an important consequence follows immediately, one that Black-Scholes has no way to accommodate.

The key idea via analogy: the wobbling ladder

Imagine a firm's assets as solid ground, moving up and down with modest volatility, say 20 percent. The debt is a fixed hurdle. The equity is what is left over above the hurdle.

Now, equity is a leveraged claim on the assets. Suppose the assets are worth 100 and the debt is 60, so the equity is worth roughly 40. If the assets fall by 10, to 90, the equity falls to roughly 30: a 10 percent asset move produced a 25 percent equity move. The equity has amplified the asset's wobble by a factor related to how much debt sits underneath it.

Here is the crucial part. As the stock falls, that amplification increases. If the assets fall to 70, the equity is worth about 10, and now a further 10-point drop in assets wipes the equity out entirely: a 14 percent asset move becomes a 100 percent equity move. The closer you get to the debt hurdle, the more violently the equity swings for a given move in the assets.

So even if the firm's assets have perfectly constant volatility, the firm's stock does not. Stock volatility rises as the stock price falls, automatically, purely because of leverage. Nobody has to panic. Nothing has to change about the business. The mathematics of limited liability does it on its own.

This is what is now called the leverage effect, and it is one of the most reliably observed patterns in equity markets. Geske derived it from first principles in 1979.

What the model actually gives you

Working this through, Geske produced a closed-form formula for a compound option, a call on a call, expressed in terms of the bivariate normal distribution rather than the single normal distribution that appears in Black-Scholes. The extra dimension appears because there are now two moments that matter: the date the first option can be exercised, and the date the second one expires, and the outcomes at those two dates are correlated. The formula is more elaborate than Black-Scholes but it is exact and computable.

The consequences are what count:

• Volatility is not a constant, it is a function of the stock price, and specifically a decreasing one. This makes Geske's model an early example of what would later be called a local volatility model, fifteen years before Dupire formalised the idea.
• It generates a volatility skew. If low stock prices come with high volatility, then downside protection is worth more than a constant-volatility model would suggest. Price out-of-the-money puts under Geske's model and they look expensive relative to Black-Scholes: exactly the skew shape that appeared in equity index options after the 1987 crash and has never left.
• It ties option markets to credit. In Geske's framework, the value of a stock option, the value of the firm's debt and the probability of default are all computed from the same underlying object, the value of the firm's assets. Equity, credit and volatility are three views of one thing.

Why it mattered

• It gave the volatility skew a cause, not just a name. Most explanations of the skew are descriptive: the market is scared of crashes, so it pays up for puts. Geske's is structural: leverage mechanically makes stocks more volatile as they fall, so the skew should exist even in a market populated entirely by unemotional robots. That is a much stronger claim, and it is testable.
• It is a founding text of structural credit modelling. The Merton-Geske line of thinking, treating equity as an option on firm assets, is the basis of the models used to infer default probabilities from stock prices. Moody's KMV, and a great deal of the capital-structure arbitrage industry, run on this logic.
• It priced a genuinely useful class of contract. Options on options really do trade: installment options, where you pay premium in stages and can abandon at each stage, are compound options. So are many real options in corporate investment, the option to start a project that then gives you the option to expand it. Geske gave all of these a formula.
• It showed the field how to handle multiple exercise dates. The bivariate-normal technique became the template for pricing Bermudan-style and multi-stage contracts analytically.

The honest limitations

• The firm's asset value is unobservable. This is the model's central practical problem. You can see the stock price. You cannot see what the company's assets are truly worth, nor their volatility, so both must be backed out from observable prices via a fiddly, sometimes unstable, calibration.
• The capital structure it assumes is a cartoon. One zero-coupon bond, one maturity date, no coupons, no rollover, no covenants, no multiple seniority classes. Real balance sheets are nothing like this, and the approximations required to squash a real firm into the model are heroic.
• Leverage explains part of the skew, not all of it. The skew observed in index options is steeper than pure leverage can account for. Crash fear, the premium investors pay for protection against systemic collapse, is doing real work that Geske's mechanism does not capture. Both stories are true; neither is sufficient alone.
• It still assumes the firm's assets follow smooth lognormal motion. All the usual Black-Scholes idealisations are still in there, just moved down one level. Assets can jump, and firms can default suddenly, which this framework struggles with.
• Structural credit models are famously bad at short horizons. Because assets move continuously, the model says a healthy firm essentially cannot default next week, so it predicts near-zero short-term credit spreads. Real short-term spreads are visibly positive. Reality has jumps; the model does not.

The one-line takeaway

Geske priced options on options, and in doing so showed that because equity is itself a leveraged option on a firm's assets, a stock's volatility must rise mechanically as its price falls, which produces a volatility skew from pure arithmetic rather than from fear, and which links equity options, corporate debt and default probability into a single framework.