Why Implied Volatility Requires Numerical Root Finding
Summary
Implied volatility is the volatility input that makes a model option price match an observed market price. The document explains why there is no universally accepted closed-form expression for that inverse mapping, while noting that the answer depends partly on what counts as a closed form. An infinite series or a formula involving special constructions may qualify under some definitions, and a cited paper proposes one such expression. The Lagrange Inversion Theorem is also mentioned as a possible route if the Black–Scholes price is analytic in volatility.
The discussion places this mathematical issue in historical context: before cheap electronic computing, iterative numerical solutions were costly, encouraging approximations and simpler models with more tractable formulas. Bond yields are given as a parallel example of quantities routinely obtained by root solving. The comments are conceptual rather than a rigorous proof that no closed form exists; they also caution that a convenient formula can depend on simplifying assumptions that reduce a model’s realism.
Key ideas
- Whether an expression counts as closed form depends on the definition being used.
- Implied volatility can be obtained by numerically solving for the volatility that matches an option price.
- Series representations or inversion theorems may provide alternative analytic expressions under suitable conditions.
- Historical computational limits encouraged simpler models and approximations with easier formulas.
- A tractable closed-form result may reflect assumptions that sacrifice accuracy.
Tags
Full text
# Why is there no formula for implied volatility?
# Why is there no formula for implied volatility?
While there are methods, none is a closed form solution. My question is aimed at mathematic problems, rather than the practical.
## Answer by Frido (score 3)
https://quant.stackexchange.com/a/81017
Too long for a comment, too short for an answer:
It depends what you mean by closed form solution. Do you consider $e^x = \sum_{n=0}^\infty \frac{x^n}{n!}$ a closed form solution?
See for example this recent paper on a closed form solution for IV, which is I believe not the only 'closed form solution' for IV: https://papers.ssrn.com/sol3/papers.cfm?abstract_id=3573239
I am not sure if the BS call/put price formula is an analytic function of the IV actually, but if it is I suppose the Lagrange Inversion Theorem could also give you a closed form solution.
## Answer by Dimitri Vulis (score 1)
https://quant.stackexchange.com/a/81015
TLDR: because before electronic computers became widely available, a paper treating the output of a root finder as a market observable might have difficulty getting through peer review.
Random reflections about then and now: I was recently showing off my grandfather's books (https://search.worldcat.org/title/927133 et al ) and we made the following observation - back in those days, people tried hard to find closed-form solutions, because numerical calculations were expensive, often too expensive. As a consequence, people sometimes tweaked their models so they would represent the reality less accurately, but had a pretty closed-form solution.
Take bond yields, for example. They're so easy to get these days, people often forget that they're almost always the output from an iterative root solver. Before everyone had access to cheap computers, people used actual paper books where you looked up the nearest coupon and maturity in a table and interpolated.
You may like the book, When Computers Were Human
Conclusion: whenever I see a pretty closed-form solution, I can't but wonder a little whether someone took some shortcuts to oversimplify the model assumptions in order to reduce numerical calculations.Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.