The implied volatility surface is not automatically arbitrage-free across multiple asset classes. Markets produce implied volatilities from option prices that must satisfy no-arbitrage constraints such as monotonicity in time and convexity in strike to avoid calendar and butterfly arbitrage. Seminal work by Emanuel Derman at Columbia University and Iraj Kani at Goldman Sachs formalized how implied trees can represent prices consistently, and Jim Gatheral at Baruch College summarized practical constraints in The Volatility Surface. In theory models can enforce arbitrage-free shapes, but real markets often violate those conditions.
Structural causes of arbitrage violations
Violations arise from microstructure, liquidity, and model mismatch. Different asset classes—equities, foreign exchange, commodities, and interest rates—have distinct drivers: equities face discrete dividends and corporate events, commodities exhibit seasonality and storage costs, FX reflects currency pegs and central bank intervention, and rates embed monetary policy. Bruno Dupire’s local volatility framework and subsequent literature show how a consistent surface can be constructed, yet calibration requires smooth, high-quality data. When quoted option prices are sparse or noisy, interpolation and extrapolation introduce spurious arbitrage even if market participants do not intend it.
Consequences for pricing, hedging, and risk
A surface that permits arbitrage leads to mispricing, erroneous hedges, and misleading risk metrics. Traders and risk managers relying on an inconsistent surface can under- or over-hedge exposures; model-driven desk losses during stressed conditions illustrate human and cultural dimensions, as market-making practices and risk tolerances differ by region and institution. Regulatory and reporting frameworks in different territories influence available liquidity and hence surface quality. Research and guidance from exchanges and regulators, including materials published by the Chicago Board Options Exchange, emphasize monitoring arbitrage conditions because of these consequences.
Practically, achieving an arbitrage-free implied volatility surface across asset classes is feasible only with careful modeling, high-quality market data, and adjustments for idiosyncratic features of each market. Techniques developed by academics and practitioners provide templates, but real-world implementation must account for liquidity, trading conventions, and occasional market stress that temporarily produce arbitrageable quotes.