arXiv:2611.04431
Reducing the arithmetic cost of quantum derivative pricing with language-model circuit design
Monit Sharma, Hoong Chun Lau
Abstract
Quantum amplitude estimation can estimate expected discounted payoffs with quadratically fewer oracle calls than Monte Carlo, but each call requires coherent computation. We use language-model-assisted circuit design to reduce reversible payoff arithmetic for a three-asset worst-of autocallable and a target accrual redemption forward (TARF). The constructions retain compact integer representations of market paths alongside contract history, reconstructing quantities temporarily for contract decisions and payoff-angle evaluation. Relative to matched internal references, the factor-sum autocallable and windowed TARF circuits reduce the product of Toffoli count and peak logical qubits $(N_{Tof} Q_{max})$ by factors of 2.39 and 4.84, respectively. At a second autocallable setting, fresh search recovers a 1.35-fold saving after the tested fixed adaptation loses its advantage. For the factor-sum circuit, we prove exact coherent execution of its integer specification for every input, incoming accumulator and measurement record, with clean workspace and phase cancellation. A separate lattice calculation gives a numerical angle-tolerance-violation probability of 6.248e-8 and evaluates pricing bias under the declared loading law, with bounded omission but no certified numerical enclosure. The results connect compact representations to lower-cost payoff arithmetic and separately assessed logical and financial correctness. Their resource scope excludes loading, payoff rotation and estimation.
AI Summary
The paper asks how to reduce reversible payoff-arithmetic costs in amplitude-estimation-based derivative pricing while separately checking coherent implementation and financial approximation error. It studies fixed, discretized benchmarks: a three-asset worst-of autocallable and a target accrual redemption forward (TARF), using language-model-assisted circuit search with deterministic resource evaluation and simulation. The circuits retain compact integer market-state representations and contract history, reconstructing weighted quantities temporarily. Against matched internal references, the autocallable circuit reduces Toffoli count from 16,810 to 7,551 and peak logical qubits from 556 to 518, a $2.39$-fold reduction in $N_{\mathrm{Tof}}Q_{\max}$. Windowed TARF reduces these counts from 28,428 to 8,381 and from 384 to 269, respectively, giving a $4.84$-fold product reduction. At a second autocallable parameter setting, a tested fixed adaptation loses its advantage; fresh search finds $1.35$- and $1.31$-fold savings. Computational certificates establish the factor-sum autocallable’s exact coherent integer action for every encoded input, incoming accumulator and measurement record, including workspace cleanup and phase cancellation. A separate recombining lattice calculation estimates the probability of exceeding $10^{-5}$ radians angle error as $6.248\times10^{-8}$ under the declared loading law; omitted mass is bounded, but numerical errors lack a certified enclosure. Other circuits rely principally on sampled validation, including million-path holdouts. The resource scope excludes loading, payoff rotations, reflections and estimation. General parameter transfer, search reliability and complete-pricing advantage are not established; rare financial failures remain, and TARF lacks an all-input correctness proof.
authors supplied the tasks and controls, Claude proposed and implemented candidate designs, and deterministic tools fitted constants and evaluated programs.
Written by gpt-6.1-sol from the paper's full text. Automated: it can be wrong.