Value a replicable financial claim

For traded contingent claims, an appropriate no-arbitrage model can infer value from a position reproducing the claim’s payments. In a one-period binomial model without an intermediate underlying payout, let current underlying value be S, up/down factors u and d, and accessible risk-free borrowing and lending growth R lie between d and u. The risk-neutral up weight is (R−d)/(u−d); discount the weighted state payoffs by R.

The reason is replication: choose an underlying holding and borrowing or lending so that both end-state payments match the option. Since the two available positions deliver the same payments under the model, different prices would allow an arbitrage. The worked call below recovers the actual holding and borrowing. The weight is not a forecast frequency, and substituting an analyst’s optimistic probability into that price calculation changes the model.

At an allowed early-exercise date, compare immediate exercise with the value of retaining the claim; at a date without that right, do not insert the exercise branch. For multiple periods apply the same local valuation backward under the model’s trading assumptions. Intermediate distributions, exercise restrictions, transaction frictions, borrowing limits, counterparty risk or path-dependent settlement can change both the payoff and replication. Obtain a model that treats those features when they matter instead of importing the simple price unchanged.

Match estimated inputs to the modeled quantity and period. Volatility of an underlying value, volatility of accounting profit and beta are different measures. An uncertainty estimate for the whole project including flexibility cannot automatically serve as the uncertainty of a fixed underlying activity to which that flexibility is then added. More finely spaced tree branches do not repair the wrong underlying definition or unavailable replication.