Pricing Temperature-Index Insurance under Long Memory and Stochastic Time Change

Battle Count: 81

The paper is primarily focused on actuarial insurance pricing rather than trading strategies. However, the stochastic time-change framework and long-memory modeling are relevant to weather derivative trading, climate risk hedging, and structured product pricing. The semi-analytical Monte Carlo methodology could inform computational approaches for exotic derivative pricing. The connection to quantitative trading is indirect, primarily through weather-linked financial instruments and climate risk transfer mechanisms.

The implementation requires: (1) deseasonalization via Fourier regression, (2) Hurst parameter estimation via DFA with bootstrap, (3) CIR parameter estimation via quasi-likelihood, (4) persistent amplitude calibration via tail matching, (5) Monte Carlo simulation of CIR paths with full-truncation Euler scheme, (6) evaluation of the conditional Gaussian exponential kernel with numerically stable log-sum-exp formulation. The mathematical derivations are sophisticated (fBm properties, CIR stationarity, entropic premium theory), but the computational implementation is tractable due to the semi-analytical structure avoiding fBm path simulation.

The paper provides detailed calibration conventions, fixed pseudorandom-number seed (20260728), explicit parameter values, simulation design with full-truncation Euler scheme, and comprehensive numerical audits including time-discretization checks, Monte Carlo convergence, stationary-CIR moment identities, limiting cases, and direct terminal simulation cross-validation. However, no code repository is provided, and the empirical data (Chicago O'Hare temperature) is publicly available but the specific processing pipeline requires careful replication.