Why Real-Probability Discounting Does Not Match Risk-Neutral Option Pricing
Summary
The document asks why a project option valued with real-world probabilities and the project’s required return differs from a risk-neutral valuation. Its binomial example has a successful project value of $10 million, an unsuccessful value of $2.7 million, an asset-sale value of $3 million, a 50 percent success probability, a 23 percent required return for comparable-risk assets, and a 5 percent risk-free rate. The stated risk-neutral up probability is 0.38, while the real-probability calculation discounts the option payoff at 23 percent.
The answer explains that option pricing comes from a replicating portfolio of the underlying and a risk-free asset, so the option’s price is determined relative to the underlying and uses risk-neutral pricing. Substituting the project’s expected return for the risk-free rate changes the discounting assumption and does not preserve the replication argument. The response is brief and does not work through the replication algebra, so it offers a conceptual explanation rather than a full derivation.
Key ideas
- A replicating portfolio prices an option using the underlying and a risk-free asset.
- Risk-neutral valuation reflects the option’s price relative to its underlying.
- Replacing the risk-free rate with the project’s required return changes the valuation assumptions.
- The example illustrates that real probabilities discounted at the project return need not match risk-neutral pricing.
Tags
Full text
# What discount rate to use when valuing binomial option with real probabilities # What discount rate to use when valuing binomial option with real probabilities We all know that we can use the argument of risk-neutrality and the law of one price, to get the option value without the real world probability. However, suppose if we use the real world probability and discount the option value with the project's risky rate, are we supposed to get same results as the risk neutral valuation? Today I came across a question (Krishnan 2006) with the following setup: There is a project with the following estimated values. \$10 million by the end of the first year, if the things work out well. and \$2.7 million if things do not turn out well. in the latter case, the company can sell the assets for \$3 million. There is a 50 percent chance that the business will succeed. Assets of comparable risk carry a required return of 23 percent. Risk free rate is 5 percent. If we use the risk neutral valuation, we calculate the PV of the project with the real rates, then calculate the risk neutral probabilities. P_up should equal 0.38. The option value is then 0.62 * 0.3 / 1.05. However, when I use real rates, the option value becomes 0.5 * 0.3 / 1.23 which is not equal to the risk neutral option value. I was wondering why is it the case? Shouldn't the law of one price give exactly one answer? > Reference: V. Sivarama Krishnan, Study Guide for Use with Principles of Corporate Finance ## Answer by Lennart_R (score 3) https://quant.stackexchange.com/a/37187 The Pricing equations are derived from duplicating portfolios consisting of underlying and a risk free asset. This means that the price of your option is relative only to the price of the underlying. In your case: Relative to the project, your option on the project does not command a risk-premium, which is basically the idea of risk-neutral pricing. Now, if i understood your question correctly, you exchanged the risk-free rate for the 23% expected return of the project, thereby implying a higher risk-free rate of return (in comparison to the 5%) to value your option based on the underlying. Obviously, the present value of the option decreases, as you increase the risk-free rate (see for example the risk-neutral pricing formula for a call as shown below).
Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)
This summary was written by Stratmill's research agent from the original; it is not a copy of the source.