What IV is, how it's computed, and why it is the central input in ASC 718 stock compensation, OPM backsolve, and DLOM analyses. For valuation professionals and finance students.
→ Compute implied volatility instantly with our free Black-Scholes calculatorThis article is educational reference material for finance and valuation practitioners and students. It is not valuation, legal, tax, or accounting advice, and should not be relied upon for any specific engagement, filing, or transaction. Standards and regulatory interpretations evolve; consult current primary sources and a credentialed professional for any real-world application.
The Black-Scholes formula takes six inputs — stock price, strike, time, risk-free rate, volatility, and dividend yield — and returns a theoretical option price. Of these six inputs, five are either directly observable or specified in the contract. Volatility is the exception: it cannot be read from a market screen or a balance sheet. It must be estimated.
Implied volatility inverts this process. Given a market-observable option price and the five known inputs, one solves for the single value of σ that makes Black-Scholes reproduce that market price exactly. In this sense, IV is not a forecast of future turbulence — it is a change of variables, a way of rescaling option prices into a common unit that makes them comparable across strikes, expirations, and underlyings. This framing, developed in the academic literature on option pricing, is more precise than the colloquial description of IV as "the market's expectation of volatility."
That looser description is not wrong, but it obscures an important subtlety: option prices embed both expected future volatility and a variance risk premium — the additional compensation option sellers demand for bearing tail risk. IV therefore tends to exceed subsequent realized volatility on average. Treating IV as a pure forecast conflates price with prediction.
Several properties follow from the inversion definition:
In business valuation, you are almost never computing IV from a market price — you are estimating a volatility input to use in a valuation model. Understanding IV as a rescaling of prices helps clarify what the volatility assumption in an ASC 718, 409A, or DLOM analysis actually represents: an analog to what the market would charge if the subject company's equity were publicly traded and options existed on it. The analyst's job is to construct a credible proxy from observable peer data.
The Black-Scholes formula is a smooth, monotonic function of σ for vanilla options: higher σ always produces a higher option price. This monotonicity is exactly what makes Newton-Raphson well-behaved here. The derivative of option price with respect to σ is Vega — a quantity the model already computes — so each iteration can be executed without finite-difference approximations.
The method converges quadratically when it converges: the number of correct decimal places roughly doubles with each iteration. In practice, the IV computation for a standard option typically converges in 5–10 iterations.
The algorithm as implemented in the homepage solver proceeds through six steps:
Step 1. Initialize σ = 0.25 (25% annualized). This is a neutral starting point that sits inside the feasible range for most listed equity options and produces a well-defined Vega.
Step 2. Compute the Black-Scholes price and Vega at the current σ.
Step 3. Apply the update rule: σ ← σ − (BS(σ) − Pmkt) / Vega.
Step 4. Clamp σ to [0.0001, 10] to prevent numeric blowups from large Newton steps in ill-conditioned regions.
Step 5. Check convergence: if |BS(σ) − Pmkt| < 1×10−8, the algorithm has converged. Return σ.
Step 6. Repeat from Step 2. If 100 iterations complete without convergence, return an error state — no valid IV exists for these inputs.
Three pathological cases require explicit guards and arise in practice:
Market price below intrinsic value. If Pmkt is below the option's intrinsic value (max(S−K,0) for calls, max(K−S,0) for puts), no σ ≥ 0 can produce that price. The solver must detect this condition and return an error before iterating.
Market price at or above the theoretical maximum. For a call, the upper bound is S·e−qT (the present value of the stock); for a put, it is K·e−rT (the present value of the strike). A market price at or above these bounds implies an arbitrage, not a valid IV. The solver rejects these inputs.
Vega collapse on deep out-of-the-money options. For options far from the strike, Vega approaches zero as the distribution assigns negligible probability to the option expiring in-the-money. Near zero, the Newton step becomes enormous and numerically unstable. A guard of |Vega| < 1×10−12 prevents division by near-zero and triggers an error return rather than a spurious result.
The pathological cases are not academic — they arise whenever a user enters a stale or erroneous market price, a price from a wide bid-ask spread, or attempts to solve for IV on a deeply OTM option with negligible market activity. Understanding the solver's failure modes is necessary for interpreting error states correctly: they indicate a data quality issue, not a model failure.
Realized volatility (also called historical volatility) is the annualized standard deviation of continuously-compounded returns over a chosen past window. Given a price series, it is unambiguous: choose a lookback window and a return frequency (daily, weekly) and the number is determined. Common conventions are 30-day, 90-day, and 252-day windows using daily returns.
Implied volatility is extracted from a current option price via inversion of Black-Scholes. It is forward-looking in the sense that the option price reflects the market's aggregate view about the distribution of future stock prices over the option's remaining life. However, because option sellers demand compensation for the risk of volatility being higher than expected — particularly in the left tail — IV incorporates a variance risk premium in addition to pure volatility expectations.
Academic literature has documented that IV tends to exceed subsequent realized volatility on average across equity markets. This systematic gap — the variance risk premium — represents the cost of the insurance that options provide. It means IV is an upward-biased predictor of realized volatility, and analysts using IV as an input to a private-company valuation should be aware that they are importing a market risk-premium component alongside the expected-volatility component.
In the ASC 718 context, SEC Staff Accounting Bulletin 107 (March 2005) describes expected volatility as a good-faith estimate that may consider historical and implied volatility, or a combination, as appropriate to the facts and circumstances. SAB 107 acknowledges both sources as potentially relevant: implied volatility from actively traded options on the company's own shares reflects current market conditions, while historical volatility provides a longer-dated empirical record. For companies without sufficient own-company option activity, the staff described using volatility of similar publicly traded companies in the same or similar industry as a substitute until the subject has at least two years of its own daily or weekly return history.
For private-company valuations where the subject has no traded options, realized volatility of the guideline public company peer group is the primary source. The analyst builds realized vol from historical return data over a lookback window matched to the valuation horizon — not from IV, which is unavailable. The academic evidence on the variance risk premium is relevant context for understanding the relationship between the peer-group realized vol estimate and what IV would be if the subject company were publicly traded and options existed.
If Black-Scholes's assumption of constant volatility were accurate, plotting IV against strike price for a single expiration would produce a flat horizontal line. In real markets, this plot is not flat — it is curved. The shape is called the volatility smile, though the curvature varies by asset class.
For equity options, the shape is more accurately described as a skew: IV increases monotonically as the strike moves below the current stock price, with OTM puts trading at meaningfully higher IV than ATM options, and OTM calls trading at equal or lower IV. This pattern did not exist in equity option markets before 1987. The crash in October 1987 caused a fundamental reassessment of left-tail risk, and the skew has persisted in listed equity markets since then. The literature attributes it to persistent demand for downside protection and to the market's recognition that large negative returns are more likely than a log-normal distribution predicts.
Foreign exchange options show a different shape — a true symmetric smile, with both OTM puts and OTM calls trading at higher IV than ATM, reflecting the symmetric nature of exchange rate risk.
IV also varies across expirations for a given underlying and strike. This volatility term structure is typically upward-sloping in calm markets — longer-dated options carry higher IV because there is more time for uncertainty to compound, and buyers pay an insurance premium for that additional exposure. During periods of elevated near-term uncertainty (earnings, macro events), the term structure can invert: short-dated IV spikes above long-dated IV as traders bid aggressively for near-term protection.
Combining these two dimensions — strike and expiration — produces the volatility surface: IV expressed as a function of both strike and time to expiration simultaneously. Practitioners calibrate option pricing models to the whole surface, not to a single ATM point.
Put-call parity, which must hold in any arbitrage-free market, implies that a call and a put at the same strike and expiration share the same implied volatility. If they did not, a costless arbitrage would exist. This is a useful consistency check.
Several modeling frameworks have been developed to account for the surface endogenously. Stochastic-volatility models — the Heston model (1993) being the most widely used — allow σ to evolve randomly over time according to a mean-reverting process, generating smiles and skews naturally. Local-volatility models, developed independently by Dupire (1994) and Derman and Kani (1994), treat σ as a deterministic function of stock price and time, calibrated to exactly match the observed surface. Both represent extensions of Black-Scholes, not replacements.
When selecting a peer-group volatility for a private-company analysis, the relevant question is which point on the public company's surface best corresponds to the subject valuation. An ATM IV from a 30-day listed option on a peer company is not the same as the long-dated, potentially OTM volatility appropriate for a 5-year ASC 718 grant or a 3-year Chaffe DLOM. Realized historical volatility over a matched lookback window is typically more appropriate for these long-horizon applications than a short-dated IV derived from listed options.
Private companies have no traded options and, by definition, no observable implied volatility. The standard approach — described in the AICPA Accounting and Valuation Guide: Valuation of Privately-Held-Company Equity Securities Issued as Compensation (commonly called the Cheap Stock Guide) — is to derive a volatility estimate from a guideline public company (GPC) peer set.
Guideline companies are publicly traded companies matched to the subject on industry, business model, and where possible, life-cycle stage and size. Broad market indices such as the S&P 500 or Russell 2000 are not appropriate substitutes for a subject-specific peer group: index volatility reflects a blended, diversified average that does not represent the undiversified equity risk of a single company in a specific industry and stage. Practitioner commentary in the field consistently emphasizes that the quality of peer selection — not the decimal precision of the volatility calculation — is the primary driver of the conclusion.
For each guideline company, the annualized standard deviation of log returns is computed over a lookback window that corresponds to the valuation horizon. For an ASC 718 grant with a 5-year expected term, a 5-year historical return series (or the longest available history if the company has been public for less than 5 years) is typical. For a Chaffe DLOM with a 3-year expected holding period, a 3-year lookback is more consistent with the valuation premise than an arbitrary 1-year window.
Equity volatility observed from a public company's stock price reflects that company's capital structure. A more leveraged company has higher equity volatility all else equal, because equity is the residual claim after fixed obligations are met. If the subject company's leverage differs materially from the median peer, the analyst may need to adjust for this difference.
The Hamada equation provides the standard framework: equity beta (and by analogy, equity volatility) can be de-levered to an asset-level measure and then re-levered at the subject company's capital structure. The intuition is straightforward — stripping out the leverage effect isolates the underlying business risk, which can then be relevered to reflect how that business risk is distributed between debt and equity at the subject company.
The following table presents annualized equity volatility by industry from Damodaran's January 2026 dataset (NYU Stern). These figures are computed from trailing stock return data for U.S. publicly traded companies and are updated annually. They serve as a reference for understanding the range of volatility levels across industries — not as substitutes for a subject-specific peer analysis.
| Industry | # Firms | σ (Equity) |
|---|---|---|
| Bank (Money Center) | 15 | 22.7% |
| Banks (Regional) | 568 | 22.7% |
| Building Materials | 41 | 35.3% |
| Business & Consumer Services | 155 | 41.1% |
| Auto Parts | 35 | 49.9% |
| Computer Services | 64 | 53.4% |
| Air Transport | 23 | 59.0% |
| Auto & Truck | 33 | 61.8% |
| Coal & Related Energy | 16 | 64.3% |
| Drugs (Biotechnology) | 496 | 75.7% |
| Drugs (Pharmaceutical) | 228 | 76.6% |
The range shown — from approximately 23% for money-center banks to over 76% for pharmaceutical companies — spans more than 50 percentage points. This spread illustrates why peer selection is the dominant source of variation in a volatility estimate, not the mechanics of the calculation. Two analysts using the same methodology but different peer groups can arrive at volatility assumptions that differ by 20–30 percentage points or more, with a corresponding impact on the valuation conclusion.
SAB 107 acknowledges the possibility of blending historical and implied volatility for companies where both sources are available — for example, a company that recently completed an IPO and has both own-company stock return history and traded options. For companies where no subject-specific option activity exists, historical peer volatility dominates. Standard practice is to ensure that the peer selection rationale, lookback window, and any leverage adjustments are clearly described and supportable from the facts and circumstances of the subject engagement.
FASB ASC 718 requires companies to measure the fair value of share-based payment awards at the grant date and recognize that fair value as compensation expense over the service period. For stock options, fair value is determined using an option-pricing model — Black-Scholes for simple grants, binomial or lattice models when the award has path-dependent features such as market conditions or reload provisions.
Volatility is one of six Black-Scholes inputs the grantor must estimate. SEC Staff Accounting Bulletin 107 (March 2005) provides the staff's interpretive views on applying this requirement. SAB 107 describes expected volatility as a good-faith estimate, identifies factors relevant to the selection of an estimation methodology, and addresses the situation of companies without sufficient own-company history. For those companies, the staff described using volatility of similar publicly traded companies in the same or similar industry — the guideline public company approach — until the subject has at least two years of its own daily or weekly return data.
SAB Topic 14.D, which addresses implied volatility specifically, identifies several factors relevant to its use: the volume of market activity in the underlying shares and any traded options, the length of time options have been available, the availability of options at strikes and expirations relevant to the grant, and whether the observed IV is consistent across the option series. Thinly traded options with wide bid-ask spreads may not provide a reliable IV signal even if they exist.
SAB 110 (December 2007) extended the availability of the simplified method for estimating expected term — not expected volatility — beyond its original 2007 sunset for plain-vanilla options granted by companies that lack sufficient own exercise history. This is a separate issue from the volatility estimation guidance in SAB 107; the two bulletins address different inputs and should not be conflated.
Section 409A of the Internal Revenue Code requires that compensatory stock options in private companies be granted at fair market value, with that value supported by a qualified appraisal. The Option Pricing Method (OPM), as described in the AICPA Cheap Stock Guide, is one of the primary frameworks for allocating total equity value across a complex capital structure with preferred stock, common stock, and options.
The backsolve method calibrates the OPM to market data rather than requiring an independent estimate of total equity value. The analyst identifies the most recent arm's-length financing round and solves for the total equity value that, when run through the OPM waterfall, produces a modeled preferred stock value equal to the transaction price. This calibrated total equity value is then used to allocate value to the common shares and other classes.
Volatility in this context is drawn from a peer group of publicly traded companies in the same industry and at a comparable life-cycle stage. Broad market indices are not appropriate substitutes for a subject-specific peer group; the relevant volatility is the undiversified equity volatility of companies that look like the subject. Leverage adjustment using the Hamada framework (described in §5) is particularly important here because early-stage companies and their public comparables may have materially different capital structures.
The expected time to a liquidity event — typically the time to an anticipated IPO, strategic sale, or other exit — is a critical input alongside volatility. These two inputs interact: a longer expected time to exit increases the value of OTM option claims, making the allocation more sensitive to both inputs simultaneously. Sensitivity analysis varying both the liquidity horizon and the volatility assumption is standard practice.
The Chaffe method, introduced in a 1993 study by David Chaffe, models the discount for lack of marketability (DLOM) as the cost of a European put option that would convert a non-marketable position into a marketable-equivalent position. The intuition: if a holder of restricted (non-marketable) stock could purchase a put option struck at the current marketable value, exercisable at the end of the restricted period, they would be fully protected against any decline in value during the restriction. The premium of that put option represents the value of marketability — and therefore the DLOM.
Chaffe's original 1993 analysis produced indicative DLOM estimates under this framework. For a 2-year holding period with equity volatility in the 60–90% range, the model produces DLOM estimates of approximately 28–41%. For a 4-year holding period with the same volatility range, the estimates rise to approximately 32–49%. These figures illustrate the dual sensitivity: both holding period and volatility materially affect the conclusion.
Chaffe acknowledged a significant limitation in the original study: the European exercise assumption understates the value of an American-equivalent marketability option. A holder of non-marketable stock might benefit from selling at any point during the restriction, not only at its expiration — a feature that European exercise cannot capture. The European put therefore produces a conservative (downward-biased) DLOM estimate. This limitation is well known in the practitioner literature.
Subsequent methods have addressed this limitation. Finnerty (2012) proposed an average-strike put model that better reflects the expected value of marketability over the full restriction period. The Finnerty method and its relationship to the Chaffe approach will be covered in a forthcoming article on DLOM methods.
Across all three frameworks, volatility is the input that typically produces the widest range of conclusions under sensitivity analysis. A 10 percentage point change in the assumed volatility — a plausible range when comparing different peer selection approaches — can move an ASC 718 fair value by 15–25%, a 409A common stock allocation by several hundred basis points, or a Chaffe DLOM by 5–10 percentage points. This is why peer selection methodology, leverage adjustment, and lookback window choice receive the most scrutiny in third-party review of these analyses.
VIX is not computed with Black-Scholes. The CBOE Volatility Index uses a model-free implied variance methodology derived from variance swap theory. The theoretical foundations were developed in work by Demeterfi, Derman, Kamal, and Zou (Goldman Sachs Quantitative Strategies, 1999) and by Britten-Jones and Neuberger (2000). Rather than extracting a single IV from a single option, the methodology aggregates a weighted portfolio of out-of-the-money call and put prices across the full available strike range, with weights proportional to the inverse square of the strike price. The result is a measure of the risk-neutral expected variance of the S&P 500 over the next 30 calendar days, expressed as an annualized volatility percentage.
What VIX measures. VIX is a 30-day forward-looking measure of expected volatility for the S&P 500, annualized and expressed as a percentage. A VIX reading of 20 does not mean the market expects a 20% move in the S&P 500 over the next month — it means the market is pricing 30-day variance at a level consistent with 20% annualized volatility. CBOE also publishes a family of related indices: VIX1D (a 1-day horizon variant), VVIX (volatility of VIX itself), and SKEW (a measure of tail-risk pricing in S&P 500 options).
VIX in a macro context. Federal Reserve research has linked option-implied volatility to macroeconomic conditions. Research published in FEDS Notes has documented that elevated option-implied interest rate volatility correlates with elevated downside risks to economic activity measures including housing starts, industrial production growth, and unemployment. Separately, research on global equity volatility has found strong cross-country correlation in option-implied volatility — a common factor that responds to global macro developments rather than country-specific news alone.
VIX is not a substitute for a subject-specific peer volatility estimate in any valuation context. An equity-sector peer group for a technology company will have materially higher volatility than the S&P 500 aggregate. However, VIX can serve as a rough sanity check: if the GPC-derived peer vol for a broad-market-correlated subject company diverges dramatically from VIX-implied levels in the same market environment, that divergence is worth investigating. It may reflect appropriate industry specificity, or it may reflect a peer selection issue.
1. Treating IV as a forecast. Implied volatility is a market price — specifically, the volatility that equates Black-Scholes to an observed option price. Because option prices embed the variance risk premium in addition to expected volatility, IV is a systematically biased predictor of subsequent realized volatility. Using a publicly traded peer company's near-term ATM IV as a direct input to a long-horizon private-company valuation imports that bias without adjustment. Realized historical volatility over a matched lookback window is generally more appropriate for this purpose.
2. Using ATM IV for an OTM valuation. The volatility surface is not flat. If the relevant option in the valuation is materially out-of-the-money — for example, a deeply subordinated common stock claim in a capital structure with a large liquidation preference — the appropriate volatility reference may differ from the ATM level. The skew means OTM options trade at higher implied volatility than ATM options; a valuation that ignores this may systematically misprice deep OTM claims.
3. Using a stale volatility input. Equity volatility is not constant over time — it varies with the market cycle, the company's own development stage, and macro conditions. A volatility assumption estimated during a low-volatility regime is not necessarily appropriate in a high-volatility regime, and vice versa. The relevant question is whether the volatility estimate reflects the conditions that will prevail over the valuation horizon, not just the conditions that prevailed during the lookback window used to compute it.
4. Ignoring the term structure. A 30-day IV or a 90-day realized volatility figure is not appropriate for a 5-year valuation horizon. Volatility term structure — the variation of volatility by expiration — means that short-dated and long-dated volatility measures are different quantities. A 5-year expected option term in an ASC 718 analysis calls for a volatility estimate over a 5-year horizon, which generally implies a longer lookback window and a longer-dated IV reference if one is available.
5. Confusing lookback window length with horizon. The lookback window for computing realized volatility is most analytically coherent when it corresponds to the valuation horizon — the time period over which the option or restricted interest is held and the uncertainty resolves. Using a 1-year daily return window as a universal default, regardless of the specific valuation horizon, is a convention without a principled basis. A 3-year Chaffe DLOM analysis is better served by a volatility estimate computed over a 3-year lookback than by an arbitrarily shorter window.
The homepage calculator uses a Newton-Raphson solver with 100-iteration convergence and pathological-case guards — the same algorithm described in §2 above. Enter an observed option price and the known inputs to extract the implied volatility instantly.
Open the Free Black-Scholes Calculator →This article is educational reference material for finance and valuation practitioners and students. It is not valuation, legal, tax, or accounting advice, and should not be relied upon for any specific engagement, filing, or transaction. All references to standards, regulatory guidance, and methodology descriptions reflect the sources as cited and as of the publication date noted above. Standards and regulatory interpretations evolve; consult current primary sources — including FASB ASC 718, SEC Staff Accounting Bulletins, and applicable AICPA guidance — and a credentialed professional for any real-world application. The volatility data referenced from Damodaran (NYU Stern) reflects a specific annual dataset and is updated periodically; current figures should be obtained directly from the source. Nothing in this article constitutes a representation that any methodology described is appropriate for any particular engagement, sufficient for any audit purpose, or consistent with any specific professional standard.