Skip to content
Financial Context Wiki

Basket Options: Composition, Rebalancing, Correlation, and Settlement

Analyze basket options by separating legal claim, constituents, weights or quantities, rebalancing, currencies, payoff, settlement, joint-distribution assumptions, and hedge risk.

Updated

For educational purposes only; not investment advice. Investing may result in loss.

Direct answer

A basket option applies an option payoff to one contractually defined aggregate of several reference assets. For basket level B_T, strike K, and multiplier or quantity M, a simple European basket call and put pay M × max(B_T − K, 0) and M × max(K − B_T, 0). The option is not a strip of options on the constituents: aggregation occurs before the convex payoff is applied.

The basket formula is a separate contract field. With fixed initial return weights w_i that sum to one, one common normalized definition is B_t = B₀ × Σ_i w_i(S_i,t ÷ S_i,0). A currency-converted version can replace each price ratio with X_i,t S_i,t ÷ (X_i,0 S_i,0). A fixed-quantity basket instead uses a form such as B_t = Σ_i q_i X_i,t S_i,t + C_t, so its live percentage weights drift. A self-financing target-weight basket rebalances quantities and compounds successive period returns; it generally cannot be reconstructed from constituent endpoints alone.

Basket also does not identify the legal product. A bilateral over-the-counter option under an ISDA framework, an option on a published index or eligible listed underlying, a FLEX contract with permitted customized fields, and an unsecured basket-linked note issued or guaranteed by a financial institution create different claims. FLEX specifications do not by themselves create an arbitrary bespoke underlying basket. Counterparty, issuer, guarantor, clearing, collateral, priority, close-out, exercise, settlement, liquidity, and calculation-agent terms must be established independently.

Seven-step basket definition, valuation, and control process

  1. Classify the legal claim and documents. Record whether the position is bilateral OTC, exchange-listed, FLEX, cleared or uncleared, an option on an existing index or security, or embedded in a note. Identify counterparty, issuer, guarantor, clearinghouse, priority, collateral, netting, close-out, governing definitions, confirmation, prospectus, pricing supplement, amendments, calculation agent, and valuation agent.
  2. Freeze the constituent perimeter. List every legal name, identifier, share class, index version, fund, futures contract, commodity, rate, currency, exchange, initial level, price or total-return field, quantity or weight, cash component, source, and unit. Distinguish a basket of securities from an index, ETF, portfolio account, worst-of test, or informal watchlist.
  3. Write the weight and maintenance engine. Specify fixed initial return weights, fixed quantities with drifting weights, market-cap or price weights, or target weights reset on stated dates. Define reference, announcement, rebalance and effective timestamps; self-financing quantity changes, residual cash, caps, floors, turnover, rounding, fees, substitutions, mergers, delistings, insolvency, nationalization, extraordinary events, and successor assets.
  4. Build the observation and currency timeline. Align trade, initial, fixing, averaging, rebalance, final valuation, exercise, settlement, and payment dates across sessions, timezones and holiday calendars. Record stale, missing, delayed, corrected and disrupted-price rules. Define dividends and distributions, unhedged currency conversion, separate FX hedge, or fixed-rate quanto treatment; these produce different basket returns.
  5. Specify payoff, exercise, and settlement. Write call or put, long or short, K, M, notional, premium, participation, cap, floor, buffer, barrier, averaging and early-exercise terms. Separate American, European, Bermudan or automatic exercise from cash, physical, futures or mixed settlement, official basket settlement value, deliverable, rounding, last trading time, instruction deadline, and payment.
  6. Value and validate the joint model. Build matched forward curves from rates, dividends, borrow, storage, carry and currency; calibrate each constituent volatility surface and any quoted basket surface at compatible timestamps. Check covariance or correlation matrices for symmetry, units and positive semidefiniteness. Compare moment matching, conditioning, PDE, tree or Monte Carlo methods, independent code, limiting cases, path and timestep convergence, and sensitivity to risk-neutral correlation, skew, jumps, defaults and tail dependence.
  7. Stress hedge, liquidity, credit, and lifecycle cash. Shock correlation toward one and through regime changes, single-name and common jumps, volatility surfaces, FX, corporate actions, rebalances, missing markets, borrow and liquidity. Reconcile component Delta, Gamma, Vega and cross-risk, premium, margin, collateral, hedge trades, residual cash, fees, taxes, accounting, issuer or counterparty default, unwind value, model version, approvals, and final settlement.

Worked examples

  • Aggregation, a constituent strip, and worst-of are different claims. Let B₀ = 100, fixed return weights w = [50%, 30%, 20%], and maturity performance ratios R = [120%, 90%, 80%]. Then B_T = 100 × (0.50 × 1.20 + 0.30 × 0.90 + 0.20 × 0.80) = 103. With K = 100 and M = $1,000 per basket point, the basket call pays max(103 − 100, 0) × $1,000 = $3,000. Applying a strike-one call to each normalized constituent before weighting gives 100 × [0.50 × max(1.20 − 1, 0) + 0.30 × max(0.90 − 1, 0) + 0.20 × max(0.80 − 1, 0)] = 10 basket points, or 10 × $1,000 = $10,000. A worst-of level is min(120, 90, 80) = 80, so a corresponding strike-100 call pays zero. The same names do not make the payoffs interchangeable.
  • Fixed shares and self-financing rebalancing can diverge with identical endpoints. Assets A and B each start at 100. A fixed-share basket holds q_A = q_B = 0.5, so its initial value is 100. At the intermediate date A is 120 and B is 80; at the final date both are 96. Fixed shares finish at B_2^fixed = 0.5 × 96 + 0.5 × 96 = 96. A self-financing rebalance at the intermediate value of 100 resets each sleeve to 50, buying q_A = 50 ÷ 120 = 0.4166667 and q_B = 50 ÷ 80 = 0.625. Using unrounded quantities, final value is B_2^reb = (50 ÷ 120) × 96 + (50 ÷ 80) × 96 = 40 + 60 = 100. Ignoring costs and taxes, a call with K = 98 and M = $1,000 per point pays $0 on the fixed-share basket and (100 − 98) × $1,000 = $2,000 on the rebalanced basket.
  • Unhedged currency and quanto returns are not the same basket. A basket starts at 100 with 60% in a U.S. stock and 40% in a euro stock. The U.S. stock moves from 100 to 110; the euro stock moves from 80 to 88; EUR/USD moves from 1.20 to 1.08. The U.S. return ratio is 1.10, while the euro stock’s unhedged U.S.-dollar ratio is (88 × 1.08) ÷ (80 × 1.20) = 0.99. Thus B_T^unhedged = 100 × (0.60 × 1.10 + 0.40 × 0.99) = 105.60. A stated fixed-rate quanto treatment based on local returns instead gives B_T^quanto = 100 × (0.60 × 1.10 + 0.40 × 1.10) = 110. For K = 105 and M = $1,000 per point, cash payoffs are (105.60 − 105) × $1,000 = $600 and (110 − 105) × $1,000 = $5,000. The exact quanto hedge and valuation still depend on rates, volatilities and asset-FX dependence.
  • Correlation changes basket volatility, but pairwise inputs must form a valid matrix. Two assets each have volatility 30% and fixed beginning weight 50%. With correlation ρ = 0.20, variance is 0.25 × 0.30² + 0.25 × 0.30² + 2 × 0.25 × 0.30² × 0.20 = 0.054 and volatility is sqrt(0.054) = 23.237900%. With ρ = 0.80, variance is 0.081 and volatility is sqrt(0.081) = 28.460499%. This is a matched local risk comparison, not an executable option quote. For three assets, pairwise correlations [0.90, 0.90, −0.90] are each individually within bounds but imply det(R) = 1 + 2(0.90)(0.90)(−0.90) − 0.90² − 0.90² − (−0.90)² = −2.888, so the matrix is not positive semidefinite and is invalid for simulation.

Risks and validation controls

  • Identify the exact legal claim, obligor, guarantor, priority, collateral, netting, close-out, clearing, custody, and governing document version.
  • Match every constituent’s identifier, share class, index version, contract month, legal issuer, venue, currency, source, unit, and eligibility.
  • State whether each input is a fixed initial return weight, live percentage weight, fixed quantity, target weight, or performance-dependent weight.
  • Reproduce self-financing rebalances, quantity changes, residual cash, rounding, turnover and effective timestamps rather than resetting weights abstractly.
  • Preserve caps, floors, substitution, delisting, merger, insolvency, nationalization, successor, and extraordinary-event rules.
  • Distinguish price return from total return and record dividends, distributions, withholding, manufactured payments and reinvestment.
  • Define unhedged FX conversion, currency hedge or quanto rate, fixing source and time, settlement currency and asset-FX dependence.
  • Align observation source, field, session, timezone, holiday calendar and asynchronous closes across all constituent markets.
  • Retain averaging, stale, missing, delayed, corrected, disrupted and calculation-agent fallback provisions.
  • Separate exercise style, last trading time, instruction cutoff, official settlement value, deliverable, settlement date and payment.
  • Quantify calculation-agent and index-sponsor discretion and conflicts, including replacement, correction, valuation and disruption determinations.
  • Distinguish bilateral counterparty exposure, issuer or guarantor credit in a note, and clearinghouse or member exposure in a cleared product.
  • Build forward curves consistent with rates, dividends, borrow, storage, carry, currency and every observation or rebalance date.
  • Calibrate constituent and basket volatility surfaces at compatible timestamps and stress strike, tenor and forward-smile dynamics.
  • Check covariance and correlation matrices for symmetry, units, positive semidefiniteness, conditioning and stable numerical factorization.
  • Do not substitute historical correlation for risk-neutral dependence; stress correlation skew, regime change, contagion and tail dependence.
  • Model single-name and common jumps, default, trading halts, stochastic volatility and cross-currency shocks where material.
  • Validate approximation and simulation bias, dimensionality, random seeds, path and timestep convergence, confidence intervals and independent code.
  • Test component hedge liquidity, market-hour mismatch, borrow, transaction size, slippage, forced rebalance and capacity under stress.
  • Reconcile premium, margin, collateral, hedge cash, residuals, fees, taxes, accounting, model reserves, exit price, disputes and final settlement.

Common misconceptions

  • “A basket is automatically diversified.” Concentrated weights, common factors, currency exposure and correlations approaching one can remove diversification.
  • “Basket volatility is the weighted average of constituent volatilities.” Squared weights and every covariance enter, and the matrix must be mathematically valid.
  • “A basket option is the same as constituent options, a worst-of, an index option, or an ETF option.” These claims aggregate, exercise, settle and expose holders to legal rights differently.
  • “Weights stay constant, and endpoints determine every basket.” Fixed quantities drift, while rebalanced baskets depend on the path, timing, cash conservation and costs.
  • “Historical correlation or one implied-correlation number is sufficient for pricing.” Valuation requires a risk-neutral joint model by maturity and strike, including skew, jumps, tails and model uncertainty.

Authoritative sources

Navigation

Search the wiki...