First-hand information for everyone
This paper proposes a new method for describing the probabilities that financial markets use to price options. Instead of working with the usual probability density or cumulative distribution, the author directly models the quantile function—the inverse of the cumulative distribution, which maps a probability level to a price return. That choice gives a simple mathematical space to work in and makes it easier to build flexible but still small models that control the shape of the implied volatility curve. The implied volatility curve is the standard way traders summarize option prices across strike prices; it can show complex local features such as central concavity or multiple peaks.
The core idea is to parameterize the risk‑neutral quantile function with a few interpretable parameters that each control behavior in a specific range of moneyness (moneyness is a way to compare the option strike to the underlying price). The paper explains how option prices come from that quantile function by a change of variables and gives formulae that link the quantile, its derivative, and the usual implied volatility. The author also shows how to construct valid risk‑neutral quantiles from simple building blocks. Examples in the text include a family built by conic combinations of exponential quantiles that yields an asymmetric logistic quantile, and operations like monotone transformations and segmental construction that preserve validity.
The model is deliberately parsimonious and interpretable. It is designed to capture a wide variety of implied volatility shapes, including U‑shaped, W‑shaped, inverse‑U, and S‑shaped curves with local concavity that have become common in short‑dated option markets, sometimes around scheduled events. The paper reports empirical calibration to about a quarter of a million implied‑volatility curves taken from two years of Standard & Poor’s 500 index option data. The fitted parameters were reported as accurate and stable across different maturities (tenors) and strikes, and the author says those stable patterns allow safe interpolation across maturities and building a dynamic model of parameters without introducing static arbitrage (an internal inconsistency in prices).
Why this matters: many existing approaches either add many parameters (for flexibility) or stick to compact models that miss complex shapes. By working with quantiles, the method keeps the model small and interpretable while reaching richer shapes than classical parsimonious models. That can help practitioners fit market prices more reliably, interpolate missing maturities, and build time‑evolving models of option prices that respect no‑arbitrage constraints.
Important caveats. The paper is focused on a particular construction and a two‑year S&P 500 options dataset; broader tests on other assets or longer periods are not shown in the excerpt. Some technical conditions are required for a quantile to be risk neutral: the paper notes integrability bounds on parameters (for one example, a tail parameter must be less than 1). The excerpt also contains theoretical derivations about how tails of the distribution relate to implied volatility asymptotics (moment and wing formulas), which place limits on possible parameter choices. Finally, the available text is an excerpt and may omit further tests or limitations discussed in the full paper.