Quanto Options and Cross-Currency Derivatives

A quanto option references an asset denominated in one currency but pays a fixed amount in another, eliminating realized FX conversion from the holder’s payoff while embedding a correlation-dependent quanto adjustment in its price. Fixing the conversion rate removes one visible risk and changes the risk-neutral drift of the foreign asset.

Contract mechanics

Suppose foreign equity S_f is quoted in foreign currency and the FX spot X is domestic currency per unit of foreign currency. A domestic-currency quanto call with fixed conversion q pays

payoff = q * max(S_f(T) - K, 0)

Its payoff differs from a composite call, which pays X(T) * max(S_f(T)-K,0). The composite retains terminal FX exposure; the quanto fixes it. Contract sheets must also state the equity index return convention, dividend treatment, payout currency, and FX fixing source.

Structure	Terminal domestic payoff	FX exposure
quanto call	q * max(S_f-K, 0)	indirect, through correlation
composite call	X(T) * max(S_f-K, 0)	direct terminal FX
FX-translated equity	max(X(T)S_f-K, 0)	direct and nonlinear

The quanto drift adjustment

Assume under an appropriate foreign measure:

dS_f/S_f = (r_f - q_div)dt + sigma_S dW_S
dX/X     = (r_d - r_f)dt + sigma_X dW_X
correlation(dW_S, dW_X) = rho

Changing to the domestic pricing measure gives the foreign equity a drift adjustment:

mu_quanto = r_f - q_div - rho * sigma_S * sigma_X
F_quanto = S0 * exp(mu_quanto * T)

With X defined as domestic per foreign, positive equity-FX correlation lowers the domestic-measure quanto forward. The sign reverses if FX is quoted inversely, so derive it from the chosen numeraire rather than memorizing a formula. The domestic value is then a Black-style price using domestic discounting and F_quanto.

import numpy as np
from scipy.stats import norm

def quanto_forward(spot, r_for, dividend, rho, vol_asset, vol_fx, T):
    drift = r_for - dividend - rho * vol_asset * vol_fx
    return spot * np.exp(drift * T)

def black_forward_call(F, K, vol, T, df):
    st = vol * np.sqrt(T)
    d1 = (np.log(F/K) + .5*vol**2*T) / st
    return df * (F*norm.cdf(d1) - K*norm.cdf(d1-st))

Market inputs and calibration

The price needs domestic and foreign discount curves, foreign dividends or carry, equity implied volatility, FX implied volatility, and equity-FX correlation. Curves must reflect collateral and cross-currency funding conventions; FX forwards alone may include basis. The mechanics of forward FX and NDFs are discussed in FX forwards and NDF trading.

Correlation is usually the weakest input. Historical correlation is not automatically risk-neutral correlation, particularly under stress when both volatilities and dependence change. Infer a correlation range from liquid quanto, composite, or related index/FX options when available; otherwise apply conservative reserves and scenario limits. A single realized correlation estimate is not a calibration.

Risk decomposition and trading

The quanto premium is approximately driven by

quanto_adjustment approximately -rho * sigma_S * sigma_X * T

Thus a long quanto option carries correlation sensitivity as well as equity delta/vega and FX-volatility vega. Delta hedging the foreign equity and FX forward may control first-order spot moves, but it does not remove correlation and cross-gamma. When equity falls and its implied vol rises, an FX move can alter both the hedge ratio and the pricing measure adjustment.

Risk	Practical hedge	Important residual
foreign equity delta	local stock/index future	dividends and basis
FX delta	FX forward	FX smile and funding
equity vol	local equity options	skew mismatch
FX vol	FX options	quote convention
correlation	quanto/composite relative value	illiquidity, jump dependence

Cross-currency derivatives also have counterparty exposure in two currencies. Collateral currency and FX resets affect discounting and valuation adjustments; a simplified domestic discount factor can be wrong even if the option formula is correct.

Model limitations

Constant volatilities and correlation produce a convenient closed form but cannot represent correlation skew, equity leverage, or FX smile. For barriers and other path-dependent quantos, simulation under a joint local-volatility, stochastic-volatility, or hybrid model may be required. Correlation should be term-structured and stress-tested, while the joint model must reproduce the vanilla surfaces it consumes. General volatility-model risk is illustrated by Heston stochastic volatility.

Key takeaways

• A quanto fixes payout conversion but retains a priced equity-FX correlation exposure.
• The drift adjustment’s sign depends on the FX quotation and chosen pricing measure.
• Calibrate curves, carry, equity vol, FX vol, and correlation consistently.
• Hedge spots and vegas separately; correlation and cross-gamma are material residual risks.
• Closed-form pricing is a baseline, not evidence that constant correlation is adequate.