Long-term option pricing with a lower reflecting barrier

TL;DR

This paper develops a model for long-term option pricing on assets with a lower reflecting barrier, incorporating government intervention and local arbitrage, and explores hedging strategies and pricing implications.

Contribution

It introduces a pricing framework for options on assets with a reflecting barrier, accounting for intervention and arbitrage, and analyzes hedging and replication strategies.

Findings

Option prices can be expressed as compound options on the notional price.

Hedging with the observed price replicates payoffs of the notional price.

Pricing strategies differ for puts and calls based on replication costs and interim losses.

Abstract

This paper considers the pricing of long-term options on assets such as housing, where either government intervention or the economic nature of the asset is assumed to limit large falls in prices. The observed asset price is modelled by a geometric Brownian motion (the 'notional price') reflected at a lower barrier. The resulting observed price has standard dynamics but with localised intervention at the barrier, which allows arbitrage with interim losses; this is funded by the government's unlimited powers of intervention, and its exploitation is subject to credit constraints. Despite the lack of an equivalent martingale measure for the observed price, options on this price can be expressed as compound options on the arbitrage-free notional price, to which standard risk-neutral arguments can be applied. Because option deltas tend to zero when the observed price approaches the barrier, hedging with the observed price gives the same results as hedging with the notional price, and so exactly replicates option payoffs. Hedging schemes are not unique, with the cheapest scheme for any derivative being the one which best exploits the interventions at the barrier. The price of a put is clear: direct replication has a lower initial cost than synthetic replication, and the replication portfolio always has positive value. The price of a call is ambiguous: synthetic replication has a lower initial cost than direct replication, but the replication portfolio may give interim losses, and so the preferred replication strategy (and hence price) of a call may depend on what margin payments need to be made on these losses.