Unlevered beta

From CEOpedia
Unlevered beta
See also


Unlevered beta - is the beta of a firm as it would be if the firm carried no debt at all, a measure of the systematic risk of its operating business with the effect of financial leverage stripped out. Aswath Damodaran describes it simply as "the beta of the firm without any debt"[1]. Practitioners also call it asset beta, because it belongs to the firm's assets rather than to its shares.

Why bother? Because the beta you can look up for a listed company is an equity beta. It mixes two things: how sensitive the underlying business is to the market, and how much debt sits between that business and the shareholders. Two cement producers with identical plants and customers can show very different equity betas simply because one borrowed heavily and the other didn't. Unlevering separates those two layers. The business risk stays; the financing risk is removed. Once that is done, betas of different companies become comparable, and an analyst can build a beta for a firm that has no share price of its own.

Levered and unlevered beta

In the capital asset pricing model the beta of a share measures how its returns move with the market portfolio. Risk that can be diversified away (see diversifiable risk) doesn't count; only market risk does. A beta of 1.0 means the stock moves roughly in line with the index, 1.5 means it tends to swing half as much again.

Debt amplifies these swings. Interest has to be paid in good years and bad, so a fixed slice of operating income goes to lenders first, and the shareholders' residual becomes more volatile. Damodaran's explanation is that "the obligated payments on debt increase the variance in net income", which pushes the equity beta up as leverage rises[1]. What remains when that effect is removed reflects two things only: the kind of business the firm is in (cyclical or defensive, discretionary or staple) and its operating leverage, i.e. the share of fixed costs in its cost base.

Put simply:

  • Levered (equity) beta - observed in the market, reflects business risk plus financial risk.
  • Unlevered (asset) beta - reflects business risk only; it is what the equity beta would be at a debt-to-equity ratio of zero.
  • Debt beta - the systematic risk carried by the firm's lenders; often assumed to be zero, which is reasonable for investment-grade borrowers and much less so for distressed ones.

Background

The idea grew out of two older pieces of finance theory. The first was the Modigliani-Miller theorem (1958, with the 1963 correction for corporate taxes), which showed how the value and the cost of equity of a firm change with its capital structure. The second was the capital asset pricing model developed in the mid-1960s, which linked expected returns to beta.

Robert S. Hamada, working at the University of Chicago, put the two together. His 1972 article in The Journal of Finance, "The effect of the firm's capital structure on the systematic risk of common stocks", derived the relation between the beta of a levered firm's equity and the beta the same firm would have without debt, and tested it on US data[2]. The resulting formula is still known as the Hamada equation. It turned a theoretical point into a working tool: if leverage raises beta in a predictable way, the effect can be backed out and added back again at any other level of debt.

Later work questioned one of Hamada's assumptions, namely that the amount of debt is fixed forever. Real firms tend to keep a target ratio of debt to value rather than a fixed sum of debt, and that changes how risky the tax savings on interest are. James Miles and John Ezzell (1980) and Robert Harris and John Pringle (1985) worked out the consequences, and their formulas are the main alternatives in use today[3][4].

Formula of unlevered beta

Hamada equation

The standard version, used in most textbooks and in Damodaran's industry data, assumes that debt is risk-free (debt beta of zero), that the amount of debt is fixed, and that interest is tax deductible at rate t. The levered beta then equals:

\[ \beta_L = \beta_U \left[ 1 + (1 - t) \frac{D}{E} \right] \]

and, rearranged, the unlevered beta is:

\[ \beta_U = \frac{\beta_L}{1 + (1 - t) \frac{D}{E}} \]

where:

  • \( \beta_L \) - levered (equity) beta, usually estimated by regression of the stock's returns on index returns,
  • \( \beta_U \) - unlevered (asset) beta,
  • \( t \) - corporate tax rate (marginal rate in most applications),
  • \( D/E \) - debt-to-equity ratio, preferably in market values (see market value of equity).

Damodaran's dataset documentation gives exactly this expression: "Unlevered Beta = Beta / (1 + (1- tax rate) (Debt/Equity Ratio))"[5]. The (1 - t) term is there because the tax deductibility of interest makes debt cheaper for the firm and absorbs part of the risk that would otherwise fall on the shareholders.

Adding a debt beta

For heavily indebted firms the assumption that debt carries no market risk breaks down. Lenders to a highly levered company share in its downturns through higher default risk. If the debt beta \( \beta_D \) is not zero, the Hamada relation is extended to:

\[ \beta_L = \beta_U \left[ 1 + (1 - t) \frac{D}{E} \right] - \beta_D (1 - t) \frac{D}{E} \]

Damodaran notes that equity betas computed on the zero-debt-beta assumption "will overstate the true betas because debt does bear market risk especially at higher debt ratios", and recommends this expanded equation as the fix[1]. Solving for the unlevered beta:

\[ \beta_U = \frac{\beta_L + \beta_D (1 - t) \frac{D}{E}}{1 + (1 - t) \frac{D}{E}} \]

Harris-Pringle and Miles-Ezzell

When a firm keeps its leverage constant as a proportion of value, rebalancing debt continuously as the firm grows or shrinks, the interest tax shields become as risky as the operating assets themselves. Under that assumption the tax rate drops out of the beta relation altogether. The Harris-Pringle formula is just a value-weighted average of the betas of equity and debt[4]:

\[ \beta_U = \frac{E \cdot \beta_E + D \cdot \beta_D}{D + E} \]

With \( \beta_D = 0 \) this simplifies to \( \beta_U = \beta_L / (1 + D/E) \), which gives a lower unlevered beta than Hamada for the same inputs.

Miles and Ezzell assumed instead that debt is reset to the target ratio once per period. The tax shield of the next period is then known in advance, while later shields depend on future value. The result sits between the two previous cases; in one common form:

\[ \beta_L = \beta_U + \frac{D}{E} (\beta_U - \beta_D) \left[ 1 - \frac{t \cdot K_D}{1 + K_D} \right] \]

where \( K_D \) is the cost of debt[3].

Which formula is right? It depends on the firm's financing policy, not on taste. Pablo Fernandez has catalogued the many formulas found in textbooks and practice and showed that each corresponds to a particular assumption about how debt and tax shields behave; mixing the assumptions (for example, unlevering with Hamada and then discounting with a WACC that assumes constant leverage) produces internally inconsistent valuations[6].

Formula Debt policy assumed Tax term in beta
Hamada (1972) Fixed amount of debt, risk-free \( (1 - t) \) applied to D/E
Hamada with debt beta Fixed amount of debt, risky \( (1 - t) \) applied to D/E
Harris-Pringle (1985) Constant D/V, rebalanced continuously none
Miles-Ezzell (1980) Constant D/V, rebalanced once a period \( t \cdot K_D / (1 + K_D) \)

Correction for cash

An unlevered beta computed from a company's equity beta reflects all its assets, cash included. Cash has a beta close to zero, so a firm sitting on a large cash pile will look less risky than its operating business really is. Damodaran's data therefore also reports an unlevered beta corrected for cash[5]:

\[ \beta_{U,operating} = \frac{\beta_U}{1 - \frac{Cash}{Firm\ value}} \]

His own illustration: an unlevered beta of 1.20 for a firm whose cash is 20% of firm value becomes 1.20 / 0.80 = 1.50 for the operating assets alone.

Use in valuation

The bottom-up (comparable companies) approach

The main practical use of unlevered beta is estimating the cost of equity for something that has no traded shares: a private company, a division of a conglomerate, a new project, or a business about to be taken private in a buyout (see private equity investments). The procedure, which Damodaran calls a bottom-up beta, runs as follows[1]:

  1. Identify the business or businesses the target is in.
  2. Find listed comparable companies in each business and collect their regression betas, market debt-to-equity ratios and tax rates.
  3. Unlever each comparable's beta.
  4. Take the median or average unlevered beta of the group. If the target operates in several businesses, weight the business betas by the market value (or a proxy such as operating income or revenue) of each.
  5. Relever the result at the target's own debt-to-equity ratio, or at a target ratio set by management or typical for the industry.
  6. Plug the relevered beta into CAPM.

Averaging across many companies is the quiet strength of this method. A single regression beta is noisy; an average over twenty or thirty firms is far more stable. Damodaran makes this point directly, noting that averaging across regression betas "reduces the noise in the estimate"[1].

Use in CAPM

Once relevered, the beta goes into the familiar CAPM equation for the cost of equity:

\[ K_E = R_f + \beta_L \times ERP \]

where \( R_f \) is the risk-free return and ERP is the equity risk premium. The resulting rate becomes the shareholders' required rate of return and, combined with the cost of debt, the weighted average cost of capital used to discount free cash flows. Graphically, the firm moves along the security market line as leverage changes its beta.

The unlevered beta itself also has a direct use. In the adjusted present value method, the unlevered cash flows are discounted at the unlevered cost of equity (\( R_f + \beta_U \times ERP \)), and the value of tax shields is added separately. The same logic helps when testing an optimal capital structure: hold the asset beta fixed, relever at different debt ratios, and watch how the cost of equity and the cost of capital respond.

Example of unlevered beta

The following example is hypothetical; company names and figures are invented for illustration.

A family-owned logistics company is being valued for a sale. It has no share price. Management plans to keep debt at 50% of equity (D/E = 0.50), and its marginal tax rate is 25%. The analyst picks four listed peers:

Peer Levered beta D/E Tax rate 1 + (1 - t) D/E Unlevered beta
A 1.30 0.40 25% 1.3000 1.000
B 1.10 0.20 21% 1.1580 0.950
C 1.45 0.65 25% 1.4875 0.975
D 0.95 0.10 19% 1.0810 0.879

Peer C has the highest equity beta, but most of that comes from debt. Once unlevered, it is no riskier than peer A.

Step 1. Median. The four unlevered betas sorted are 0.879, 0.950, 0.975 and 1.000. The median is the average of the middle two: (0.950 + 0.975) / 2 = 0.962, rounded to 0.96.

Step 2. Relever at the target structure.

\[ \beta_L = 0.96 \times [1 + (1 - 0.25) \times 0.50] = 0.96 \times 1.375 = 1.32 \]

Step 3. Cost of equity. With a risk-free rate of 4.2% and an equity risk premium of 5.0%:

\[ K_E = 4.2\% + 1.32 \times 5.0\% = 10.8\% \]

For comparison, the same business financed entirely with equity would have a cost of equity of 4.2% + 0.96 × 5.0% = 9.0%. The 1.8 percentage-point gap is the price the shareholders charge for carrying the planned debt.

Sensitivity. Two quick checks show how much the assumptions matter:

  • If the peers' debt is assumed to have a beta of 0.20, the unlevered betas rise (to 1.046, 0.977, 1.040 and 0.894; median 1.01), and the relevered beta at D/E = 0.50 becomes 1.01 × 1.375 - 0.20 × 0.375 = 1.31. The cost of equity barely moves, to about 10.8%, because the higher asset beta is offset by the debt beta on the relevering side.
  • Using Harris-Pringle (no tax term, zero debt beta) gives unlevered betas of 0.929, 0.917, 0.879 and 0.864, a median of 0.898, a relevered beta of 0.898 × 1.5 = 1.35 and a cost of equity of about 10.9%.

Here the methods differ by only about a tenth of a percentage point, because the peers carry modest debt. With heavier leverage the gap widens quickly, and on a long-lived business valued by discounted cash flows (see net present value) even small differences in the discount rate move the result.

Damodaran's industry data

Many analysts skip the peer-collection step and use published sector averages. The best known source is Aswath Damodaran of NYU Stern, who updates a free dataset of levered and unlevered betas by industry every January[7]. The US edition with data as of January 2026 covers 5,994 firms in total and reports, for each industry, the number of firms, average beta, market D/E ratio, effective tax rate, unlevered beta, cash as a share of firm value and the unlevered beta corrected for cash. A few lines from that table:

Industry Firms Beta D/E Unlevered beta Unlevered beta corrected for cash
Semiconductor 66 1.52 2.59% 1.49 1.50
Software (System & Application) 309 1.28 5.58% 1.23 1.25
Air Transport 23 1.19 91.17% 0.70 0.76
Utility (General) 14 0.24 81.48% 0.15 0.15
Total Market 5,994 0.91 35.17% 0.72 0.76

The pattern is instructive. Air Transport and Semiconductor have similar-looking equity betas on the surface, yet airlines carry far more debt, and once that is removed their business risk turns out to be much lower than that of chip makers. Utilities sit at the bottom: stable regulated demand and a lot of debt. Damodaran publishes similar tables for Europe, Japan, emerging markets and a global sample on the same site.

Advantages of unlevered beta

  • Makes companies comparable. Firms in the same business but with different financing can be put on one scale.
  • Works without a share price. Private firms, divisions, projects and start-ups can get a market-based beta from listed peers.
  • Reduces estimation noise. Averaging unlevered betas across many comparables gives a steadier estimate than one regression on one stock.
  • Follows changes in capital structure. After a recapitalisation, an acquisition financed with debt or a planned deleveraging, the beta can be relevered at the new ratio instead of waiting years for a regression to catch up.
  • Handles business mix. A conglomerate's beta can be built as a value-weighted average of the unlevered betas of its segments, which fits naturally with a sum-of-the-parts valuation.

Limitations of unlevered beta

  • Debt beta assumption. The standard Hamada formula treats debt as risk-free. For highly levered or speculative-grade firms this overstates equity risk when relevering and understates asset risk when unlevering.
  • Tax shield assumptions. Hamada assumes a fixed amount of debt; Harris-Pringle and Miles-Ezzell assume a constant ratio. The choice changes the answer, and few firms behave exactly like either model. Loss-making firms that can't use their interest deductions break the (1 - t) logic altogether.
  • Estimation error in the input betas. Regression betas depend on the index, the return interval (daily, weekly, monthly) and the period chosen. Unlevering doesn't remove that error; it only averages it out if the peer group is large enough.
  • Mismatched leverage. The D/E used to unlever should be the one that prevailed over the regression window, not just the latest figure. Damodaran points out that regression betas lag recent leverage changes for this reason[1].
  • Choice of comparables. Peers rarely match on size, geography and operating leverage. A small firm with high fixed costs may deserve a higher asset beta than its industry median.
  • Book versus market values. Using book debt and book equity can distort the ratio badly, especially for firms whose share prices are far from book value.
  • Inherits CAPM's weaknesses. Unlevered beta captures only market-wide risk measured against one index. Undiversified owners of private firms bear total risk, which beta alone doesn't price.

Related approaches

  • Regression (top-down) beta - estimated directly from a stock's historical returns. Simple for listed firms, unavailable for private ones.
  • Adjusted (Blume) beta - a regression beta shrunk towards 1.0 to correct for mean reversion; often applied before unlevering.
  • Accounting beta - computed from changes in earnings instead of stock prices; useful when no market data exist, but based on few, smoothed data points.
  • Total beta - market beta divided by the correlation of the stock with the market, used for owners who hold most of their wealth in one business. Damodaran defines it in the same dataset documentation[5].
  • Multi-factor models - replace the single beta with sensitivities to several factors; the leverage adjustment then has to be applied factor by factor.

References

Footnotes

  1. ↑ Damodaran A., Estimating Risk Parameters, Stern School of Business, New York University, sections "Degree of Financial Leverage" and "Bottom Up Betas"
  2. ↑ Hamada R.S., The effect of the firm's capital structure on the systematic risk of common stocks, The Journal of Finance, 1972, pp. 435-452
  3. ↑ Miles J.A., Ezzell J.R., The weighted average cost of capital, perfect capital markets, and project life: a clarification, Journal of Financial and Quantitative Analysis, 1980
  4. ↑ Harris R.S., Pringle J.J., Risk-adjusted discount rates - extensions from the average-risk case, Journal of Financial Research, 1985, pp. 237-244
  5. ↑ Damodaran A., Variable definitions, Damodaran Online, entries "Unlevered Beta", "Unlevered Beta adjusted for cash" and "Total Beta"
  6. ↑ Fernandez P., Levered and Unlevered Beta, SSRN Working Paper, 2002
  7. ↑ Damodaran A., Betas by Sector (US), data as of January 2026, NYU Stern
Unlevered beta — recommended articles
Systematic risk — Debt-to-equity ratio — Modigliani-Miller theorem — Operating leverage — Required rate of return — Adjusted present value — Optimal capital structure
Retrieved from CEOpedia