Advanced Financial Management · Application of option pricing theory in investment decisions

How to Estimate Volatility and Inputs for Real Options

Updated 11 October 2026 · Fact-checked

To value a real option with Black-Scholes, map the project to five inputs: PV of project cash flows (share price), investment cost (exercise price), time to decide, risk-free rate and volatility. Estimate volatility from the standard deviation of returns on a proxy company's shares or from historical data, then annualise it.

Understand Estimating Volatility and Inputs for Real Option Valuation

A real option is the right, but not the obligation, to take a business action later: delay, expand, abandon or switch. Black-Scholes was built for traded share options. To use it on a project, you treat the project like a share and the option like a call (or put) on it.

The mapping is the core skill. The share price (Pa) becomes the present value of the future cash inflows from the project, excluding the cost of investing. The exercise price (Pe) becomes the capital you must spend to take the action. The time to expiry (t) is how long you can wait before the opportunity disappears. The risk-free rate (r) is the government yield for a similar period. The volatility (s) is the annual standard deviation of the project's value.

Volatility is the hard input because projects do not trade. You cannot read their price history. So you borrow it. The usual approach is to find a listed company in the same line of business, take the standard deviation of its share returns, and use that as a proxy. If the question gives share prices or returns, you calculate the standard deviation yourself. If it gives a rate for a short period, you scale it up to a year.

Volatility has a big effect on option value. Higher volatility means a higher option value, because the option protects you from the downside but lets you keep the upside. So a small error in volatility can change the answer a lot. This is why examiners also ask you to comment on how reliable the estimate is.

Black-Scholes is only a guide for real options. A project is not traded, its volatility is an estimate, and volatility may change over time. Real options can often be exercised early, which the basic model does not allow for. Competitors may also act and shorten the option's life. Quote these limits when you are asked to evaluate the result.

Key rules to remember

Black-Scholes call value: c = Pa × N(d1) − Pe × e^(−rt) × N(d2); Pa = PV of project inflows, Pe = investment cost, r = risk-free rate, t = years to expiry. Use the formula sheet and normal distribution tables ACCA gives you.
d1: d1 = [ln(Pa ÷ Pe) + (r + 0.5 × s²) × t] ÷ (s × √t); s is annual volatility as a decimal, e.g. 0.25 for 25%.
d2: d2 = d1 − s × √t; Calculate d1 first and then deduct s√t.
Input mapping: Pa = PV of inflows; Pe = capital outlay to exercise; t = time to decide; r = risk-free rate; s = volatility of project value; Do not subtract Pe from Pa when you set Pa. The model does that through the formula.
Sample standard deviation of returns: s(period) = √[ Σ(x − x̄)² ÷ (n − 1) ]; x = periodic return, often ln(Pt ÷ Pt−1). Use n − 1 for a sample unless the question says otherwise.
Annualising volatility: s(annual) = s(period) × √(number of periods in a year); Weekly: × √52. Monthly: × √12. Daily: × √(trading days), usually 250 or 252 as told. Volatility scales with the square root of time, not time.
Option value and volatility: Higher s → higher option value (for calls and puts); Use this for comment marks on sensitivity.

How to solve Estimating Volatility and Inputs for Real Option Valuation questions

Use the same sequence for any question on real option inputs. It keeps the mapping, the calculation and the comment separate, so you pick up marks at each stage.

1Identify the option type from the scenario: delay or wait (call), expand (call), abandon (put) or switch. Say it in one line.
2Map each feature to an input. State Pa (PV of inflows), Pe (investment cost), t (time to decide) and r (risk-free rate). Give the figures with units.
3Decide how to get volatility. If given directly, use it. If given a proxy company's share prices or returns, calculate it. If it is for a short period, annualise it.
4Calculate periodic returns if needed, then the mean, the squared deviations and the standard deviation using n − 1. Multiply by the square root of periods per year.
5Calculate d1 and d2, read N(d1) and N(d2) from the tables, and compute the option value.
6Interpret the result. Compare the option value with any base-case NPV. Say what the extra value means for the decision.
7Comment on limits: proxy volatility may not match the project, volatility may change, exercise may be early, and competitors may act. Link each point to the scenario.

Quickest way: Map, annualise, plug in

When to use it: Use this when time is short and the question already gives you volatility data, or a proxy company, and asks for the option value or an assessment of it.

Write a five-line input box: Pa, Pe, t, r, s. Fill in all five before any calculation.
If s is not annual, convert it now: periodic s × √periods per year.
Compute d1 first, then d2 = d1 − s√t. Round d to two decimals so the table lookup is easy.
Calculate c = Pa × N(d1) − Pe × e^(−rt) × N(d2). Show every part so method marks are safe.
Close with two lines: the option value compared with the base NPV, and one limitation that is specific to the scenario.

Common mistakes in Estimating Volatility and Inputs for Real Option Valuation

Setting Pa as the project NPV, or as the PV of inflows minus the investment.
Students think the option is on the net benefit, so they subtract the cost before putting figures into the formula.
Fix: Pa is the PV of inflows only. The cost goes in as Pe. The model compares them itself. The option value is then added to, or compared with, the base NPV.
Using weekly or monthly standard deviation directly as annual volatility.
The number looks like a volatility, and the time basis is missed.
Fix: Check the data frequency. Multiply by √52 for weekly, √12 for monthly or √250 (or the stated figure) for daily data.
Multiplying by the number of periods instead of its square root.
Students scale volatility like an interest rate or a cash flow.
Fix: Variance grows with time, so standard deviation grows with √time. Write the square root in your working.
Using n instead of n − 1 when calculating standard deviation from a sample of returns.
Population and sample formulas are confused, especially under time pressure.
Fix: Use n − 1 for historical data treated as a sample, unless the question tells you otherwise. State your choice.
Using a proxy company's volatility without comment on whether it fits the project.
Students treat the calculation as the whole answer and skip the judgement.
Fix: Say whether the proxy is in the same business, has similar gearing and operating risk, and is of similar size. Note that equity volatility also reflects the proxy's financial risk.
Giving a generic list of Black-Scholes limitations with no link to the scenario.
Students memorise a list instead of applying it.
Fix: Pick two or three limits that matter here, such as an uncertain project volatility, the possibility of early exercise or a competitor entering, and explain the effect on the value.

Worked examples

Example 1

A company wants to value an option to expand a project in four years' time. It has no history of the project's value, so it uses a listed competitor as a proxy. Weekly continuously compounded returns on the competitor's shares over five weeks were 2%, 4%, −1%, 3% and 2%. Estimate the annual volatility, using the sample standard deviation and 52 weeks in a year.

Show the solution

Mean weekly return = (2 + 4 − 1 + 3 + 2) ÷ 5 = 10 ÷ 5 = 2%.
Deviations from the mean: 0, 2, −3, 1, 0.
Squared deviations: 0, 4, 9, 1, 0. Their sum is 14.
Sample variance = 14 ÷ (5 − 1) = 3.5. Weekly standard deviation = √3.5 = 1.8708%.
Annualise: 1.8708% × √52 = 1.8708% × 7.2111 = 13.49%.
Comment: five observations are far too few for a reliable estimate in practice. Use a longer series. The competitor's volatility also reflects its own gearing and mix of activities, so test whether it matches the project.

Answer: Estimated annual volatility is about 13.5%. Treat it as a rough proxy and say why it may differ from the project's true volatility.

Example 2

A company can invest $11 million in a project in two years' time, if conditions are right. The PV of the project's inflows today is $12 million. The risk-free rate is 4% a year. The proxy company's annual share volatility is 25%. Treat the right to invest as a call option. Identify the inputs and value the option using Black-Scholes. Use d values rounded to two decimals and the normal tables: N(0.65) = 0.7422 and N(0.30) = 0.6179.

Show the solution

Map the inputs: Pa = $12m (PV of inflows), Pe = $11m (investment cost), t = 2 years, r = 4% (0.04), s = 25% (0.25).
d1 = [ln(12 ÷ 11) + (0.04 + 0.5 × 0.25²) × 2] ÷ (0.25 × √2).
ln(12 ÷ 11) = ln(1.0909) = 0.0870. (0.04 + 0.03125) × 2 = 0.1425. Numerator = 0.2295.
Denominator = 0.25 × 1.4142 = 0.3536. d1 = 0.2295 ÷ 0.3536 = 0.65 (to two decimals).
d2 = 0.6491 − 0.3536 = 0.2955, which rounds to 0.30.
N(d1) = 0.7422 and N(d2) = 0.6179.
PV of exercise price = 11 × e^(−0.04 × 2) = 11 × 0.9231 = $10.154m.
c = 12 × 0.7422 − 10.154 × 0.6179 = 8.906 − 6.274 = $2.632m.
Comment: a static NPV, taking the investment now, is 12 − 11 = $1m. The option to wait is worth about $2.6m, so waiting has more value than investing immediately. The result depends on the 25% volatility, which comes from a proxy. It also assumes the option can only be used at the end of year 2, and competitors may enter before then.

Answer: The option is worth about $2.63 million, compared with a static NPV of $1 million. On these inputs, keeping the option open is worth more than investing now, subject to the reliability of the volatility estimate.

Exam tips

Always start with a clear input table. Marks are given for correct mapping even if your arithmetic later goes wrong.
Check the frequency of the data in the question. Annualisation is a very common trap, and the question rarely tells you to do it.
Show standard deviation working in steps: mean, deviations, squares, divide by n − 1, square root. Partial credit depends on it.
Write the comment part in the scenario's own terms. Say what the option value means for the decision, and tie each limitation to something in the case.
Remember the professional skills marks. A short, structured recommendation to the board is worth more than a long list of generic points.

Practice questions from Application of option pricing theory in investment decisions

Estimating Volatility and Inputs for Real Option Valuation: frequently asked questions

How do I estimate volatility for a real option in ACCA AFM?

Use a listed company in a similar business as a proxy. Calculate the standard deviation of its share returns and annualise it. If the question gives the volatility, use it directly. Always say how reliable the estimate is.

What are the five Black-Scholes inputs for a real option?

Pa is the PV of the project's future inflows. Pe is the cost of investing or exercising. t is the time you can wait before deciding. r is the risk-free rate. s is the volatility of the project's value, as an annual figure.

Why can't I just use the project's own cash flows to get volatility?

A project usually has no traded price history, and its forecast cash flows are estimates. A proxy company's share prices give observed data on how risky that type of business is. This is not perfect, but it is more objective than a guess.

What are the limitations of Black-Scholes for real options?

Volatility is hard to estimate and may change. The model assumes a single exercise date, while real options can often be exercised earlier. It assumes the underlying can be traded, and a project cannot. Competitors can also erode the option's value. Use these points to qualify your answer.