Fast Calculation of Cheapest-to-Deliver Curves

Why It Matters

Faster CTD calculations enable real‑time pricing, XVA adjustments, and efficient collateral management, giving banks a competitive edge in multi‑currency trading.

Key Takeaways

• Multi-currency collateral adds optionality requiring CTD curves
• Monte Carlo accurate but computationally intensive
• New method uses Clark algorithm and Gauss‑Hermite quadrature
• Achieves faster, scalable calculations across any currency count
• Validated against realistic market scenarios with high precision

Pulse Analysis

Since the 2008 financial crisis, market participants have placed a premium on accurate collateral‑adjusted pricing. When a trade permits cash collateral in several currencies, the discount factor must reflect the optionality of choosing the cheapest‑to‑deliver (CTD) currency, giving rise to a CTD curve. Traditional Monte‑Carlo simulations can price this optionality almost exactly, but the computational load grows sharply with the number of eligible currencies and the frequency of re‑valuation. Consequently, banks and asset managers have sought faster analytic approximations that retain pricing fidelity while supporting real‑time risk dashboards.

The new approximation builds on two well‑established statistical tools. First, the Clark algorithm efficiently computes the distribution of the maximum of correlated normal variables, which captures the currency‑selection payoff embedded in the CTD curve. Second, Gauss‑Hermite quadrature provides accurate integration of the resulting Gaussian mixture with only a handful of nodes. By coupling these techniques with an optimized discretization of the underlying factor model and aggressive factor‑reduction, the method delivers near‑Monte‑Carlo accuracy at a fraction of the runtime. It scales linearly with the number of currencies, making it practical for portfolios spanning dozens of collateral options.

The speed gains translate directly into lower operational costs and more frequent CTD updates, which are critical for XVA calculations and collateral optimization. Traders can now run scenario analyses across multiple currency baskets in near real‑time, improving hedge effectiveness and regulatory reporting. Early adopters in major banks report integration times of weeks rather than months, and the approach is compatible with existing pricing libraries written in C++ or Python. As multi‑currency collateral becomes standard in emerging markets, the scalable analytic framework positions firms to capture pricing advantages while maintaining rigorous risk controls.