Econophysics Textbooks: Mantegna-Stanley vs Bouchaud-Potters
Which econophysics reference actually covers the method a trader needs?
Pick a reference from the calculation needed: tail risk, stochastic processes or market microstructure, then verify its assumptions against the data you have. A textbook?s coverage helps choose a method; it does not validate a trading edge.
Bibliographic guidance is descriptive and does not establish a return-producing strategy.
Evidence map
| Aspect | Finding |
|---|---|
| What it is | The two reference books of the field, written from its two ends. |
| Key result / formula | Mantegna & Stanley cover, in order, the efficient-market hypothesis and the random walk; Lévy-stable distributions and why they cannot fully describe returns (infinite variance); the truncated Lévy flight as a fix; scaling and its breakdown; ARCH/GARCH as the economists' answer; the analogy with turbulence and where it fails; cross-correlations between stocks and their hierarchical structure (the minimum-spanning-tree taxonomy); and a first pass at option pricing in idealised and real markets. |
| Why it matters for backtesting | Open Mantegna & Stanley when a scaling claim needs to be understood or reproduced: it explains what a stable law, a truncated Lévy flight or a correlation taxonomy is, and its worked examples are on index data, so it is the fastest way to see what the physicists' plots measure. |
What it is
Mantegna & Stanley (2000) is the short introduction by the Boston group behind the scaling-law literature: it teaches the vocabulary of statistical physics — random walks, stable distributions, scaling, correlations — and shows it on financial data. Bouchaud & Potters (2003, second edition) is the working manual of a quantitative firm: it goes from the same empirical facts to portfolio risk, option pricing and hedging under non-Gaussian returns.
Key result / formula
Bouchaud & Potters open with probability theory for fat-tailed variables and the statistics of real prices, then treat portfolio optimisation with non-Gaussian risk measures, futures and options — including an option theory in which the hedging residual risk stays above zero — risk management, random matrix theory for correlation matrices, and, added in the second edition, stochastic processes, Monte Carlo methods, the Black-Scholes limit, the yield curve and the Minority Game.
Why it matters for backtesting
Open Bouchaud & Potters when the question is what to do with a fat-tailed return distribution: how much a Gaussian value-at-risk underestimates, how to size a portfolio against a tail rather than a variance, why a hedge on a non-Gaussian underlying leaves residual risk, and how to clean a correlation matrix before inverting it. Neither book contains a strategy. Both insist that the empirical facts are distributional, and the residual-risk chapters of Bouchaud & Potters are the sharper reminder that a model reproducing the full list of stylized facts can still be useless for prediction.
Source
Mantegna & Stanley, An Introduction to Econophysics: Correlations and Complexity in Finance, Cambridge University Press, 2000; Bouchaud & Potters, Theory of Financial Risk and Derivative Pricing: From Statistical Physics to Risk Management, 2nd ed., Cambridge University Press, 2003. Primary source