How Options Are Priced: The Intuition and the Limits

This article gives the intuition without the mathematics, because the mathematics is not what a reader needs and the intuition is what makes the Greeks and implied volatility make sense. It then names the model's assumptions and their documented failures, because a model presented without its limits invites exactly the misplaced confidence this pillar exists to prevent.

The intuition: five inputs and one idea

Start with the inputs, all of which are now familiar. The price of the underlying and the strike together determine how far the option is from paying off. Time to expiry determines how much opportunity remains. Expected volatility determines how much the underlying is likely to move in that time. And the interest rate matters because money has a time value — buying an option instead of the underlying leaves cash uncommitted, which is worth something. Dividends enter as an adjustment, since a shareholder receives them and an option holder does not. Four of those five are observable. The fifth — expected volatility — is not, which is the entire reason implied volatility exists as a concept: given a traded price, you can solve backwards for the volatility that justifies it. Now the idea that made the framework work, and it is genuinely elegant. The insight was that an option's payoff can be replicated by continuously adjusting a mix of the underlying and cash. If a portfolio can be constructed that produces exactly the same outcome as the option under any price movement, then the option must cost the same as that portfolio — otherwise a riskless profit would be available by buying one and selling the other. This is a no-arbitrage argument, and it has a striking consequence: the option's value does not depend on whether you expect the underlying to rise or fall. Direction drops out entirely. What matters is how much the underlying moves, not which way. That is counter-intuitive on first encounter and it explains something readers often find puzzling — why two people with opposite views on a company can agree on what an option is worth. It also explains why vega is positive for buyers of both calls and puts: more movement is worth more, regardless of sign. The framework also delivers the Greeks as a by-product, since if you have a formula for value in terms of five inputs, the sensitivity to each input follows by differentiation. That is what the Greeks are — not separate discoveries but the derivatives of a pricing function, which is why they inherit every one of the model's limitations.

The assumptions, and how they fail

The framework rests on assumptions that are known to be false, and the honest position is that it remains useful anyway — as a common language and a reference point rather than as a description of reality. Five assumptions and their failures. Returns are lognormally distributed with constant volatility. They are not. Real return distributions have fatter tails than the model assumes — extreme moves happen considerably more often than it implies — and volatility varies over time and clusters, with turbulent periods following turbulent periods. This is the most consequential failure, because it means the model systematically underprices the probability of large moves, which is exactly the scenario in which a position matters most. The crisis pillar documents what happens when models built on this assumption meet markets that are not. Trading is continuous and frictionless. The replication argument requires continuous adjustment with no transaction costs and no gaps. Real markets have spreads, limited depth, and price gaps — a share can open substantially below its previous close, and no continuous adjustment is possible across that gap. Volatility is known. It is not observable at all, which is why the model is used in reverse more often than forwards: practitioners feed in the traded price and read out implied volatility rather than feeding in volatility and getting a price. Interest rates are constant and borrowing is unlimited at one rate. Neither holds. And the original framework assumes European exercise and no dividends, both of which required extensions — which is why American-style options need different approaches, and why numerical methods such as binomial trees and simulation are used where closed-form solutions do not exist. The clearest evidence that the model is imperfect is visible on any option chain. If a single true volatility existed, every option on the same underlying and expiry would imply the same figure. They do not — the skew or smile is the market pricing in fatter tails and asymmetric risk that the model omits, and practitioners accommodate it by using a different volatility input for each strike. That is a workaround, openly acknowledged as such, and it tells you what the framework really is: a shared language for quoting and comparing option prices, whose parameters are adjusted to fit observed reality rather than a law that determines it. The Nobel-recognised achievement was real, the mathematics is correct, and the assumptions are wrong — all three at once. What follows for a reader is a specific piece of caution. Every model-derived figure a platform shows you — every Greek, every implied volatility, every "fair value" — inherits these limitations. Those numbers are useful and they are not measurements. Treating them as precise, particularly in the conditions where precision would matter most, is the error the model's own history most warns against.