4
$\begingroup$

Hamada's formula is presented as follows:

$$\beta_{U}=\left[\frac{1}{1+\frac{D}{E}(1-\tau)}\right]\beta_{L},$$

where $\beta_{U}$ and $\beta_{L}$ are the unlevered and levered betas of a firm respectively. $D$ is the market value of debt. $E$ is the market value of equity and $\tau$ is the tax rate.

May anyone please provide a proof for this formula? I have found some sources on the internet, though they are not convincing.

Thanks to all in advanced.

$\endgroup$
5
  • $\begingroup$ I might but it's good to know what you found unconvincing about the other sources you found. Can you elaborate? $\endgroup$
    – Bob Jansen
    Apr 5, 2015 at 7:05
  • $\begingroup$ @BobJansen. Thank you for your time. I thought about it a little more and was able to figure it out. I posted my proof as an answer. I used the wikipedia source found here: en.wikipedia.org/wiki/Hamada%27s_equation $\endgroup$ Apr 5, 2015 at 12:47
  • $\begingroup$ That's the best :) Welcome to Quant.SE! $\endgroup$
    – Bob Jansen
    Apr 5, 2015 at 18:11
  • $\begingroup$ The Hamada equation contains some crippling assumptions that make it unusable in the real world. Please see researchgate.net/publication/… $\endgroup$
    – chaamjamal
    Mar 27, 2017 at 3:35
  • $\begingroup$ can you help me on these doubts, please? 1) In the (very nice) answer by Gus : a) why should Net Borrowing be computed for return on EL? Net borrowing is computed for FCFE but it is hard to see it as part of a return metric, specially considering MM theory that only tax shield creates value. b) why CAPEX Beta is assumed to be zero ? It is as intrinsic part of the business as EBITDA. c) How does it relate to this explanation (tax shield´s beta assumption)? faculty.babson.edu/goldstein/teaching/fin3520secvalfall2011/… 2) PREMIUM QUESTION - Why d $\endgroup$
    – Italo b
    Dec 28, 2023 at 17:59

2 Answers 2

9
$\begingroup$

Proof: Recall that

$$\beta_{i} = \frac{\mathrm{Cov}(r_{i},r_{m})}{\mathrm{Var}(r_{m})}.$$

Now, the returns on unlevered and levered equity are given by

$$r_{U} = \frac{\mathrm{EBIT}(1-\tau) - \mathrm{CAPEX} + \mathrm{Depreciation}}{E_{U}}$$ $$r_{L} = \frac{\mathrm{EBIT}(1-\tau) - \mathrm{CAPEX} + \mathrm{Depreciation} + \mathrm{Net\ Debt} - \mathrm{Interest}}{E_{L}},$$

respectively.

Therefore,

$$\beta_{U} = \frac{\mathrm{Cov}\left(\frac{\mathrm{EBIT}(1-\tau) - \mathrm{CAPEX} + \mathrm{Depreciation}}{E_{U}},r_{m}\right)}{\mathrm{Var}(r_{m})}$$ $$\beta_{L} = \frac{\mathrm{Cov}\left(\frac{\mathrm{EBIT}(1-\tau) - \mathrm{CAPEX} + \mathrm{Depreciation + \mathrm{Net\ Debt} - \mathrm{Interest}}}{E_{L}},r_{m}\right)}{\mathrm{Var}(r_{m})}.$$

Working with the $\beta_{U}$ equation,

$$ \begin{align} \beta_{U} &= \frac{\mathrm{Cov}\left(\frac{\mathrm{EBIT}(1-\tau) - \mathrm{CAPEX} + \mathrm{Depreciation}}{E_{U}},r_{m}\right)}{\mathrm{Var}(r_{m})} \\ &= \frac{\mathrm{Cov}\left(\frac{\mathrm{EBIT}(1-\tau)}{E_{U}} - \frac{\mathrm{CAPEX}}{E_{U}} + \frac{\mathrm{Depreciation}}{E_{U}},r_{m}\right)}{\mathrm{Var}(r_{m})} \\ &= \frac{\mathrm{Cov}\left(\frac{\mathrm{EBIT}(1-\tau)}{E_{U}}, r_{m}\right) + \mathrm{Cov}\left(\frac{-\mathrm{CAPEX}}{E_{U}}, r_{m}\right) + \mathrm{Cov}\left(\frac{\mathrm{Depreciation}}{E_{U}},r_{m}\right)}{\mathrm{Var}(r_{m})} \\ &= \frac{\mathrm{Cov}\left(\frac{\mathrm{EBIT}(1-\tau)}{E_{U}}, r_{m}\right)}{\mathrm{Var}(r_{m})} + \frac{\mathrm{Cov}\left(\frac{-\mathrm{CAPEX}}{E_{U}}, r_{m}\right)}{\mathrm{Var}(r_{m})} + \frac{\mathrm{Cov}\left(\frac{\mathrm{Depreciation}}{E_{U}},r_{m}\right)}{\mathrm{Var}(r_{m})}.\\ \end{align} $$

Since $E_{U}$ is the value of unlevered equity from the last financial year, it is constant. Hence,

$$ \begin{align} \beta_{U} &= \frac{\mathrm{Cov}\left(\frac{\mathrm{EBIT}(1-\tau)}{E_{U}}, r_{m}\right)}{\mathrm{Var}(r_{m})} + \frac{\mathrm{Cov}\left(\frac{-\mathrm{CAPEX}}{E_{U}}, r_{m}\right)}{\mathrm{Var}(r_{m})} + \frac{\mathrm{Cov}\left(\frac{\mathrm{Depreciation}}{E_{U}},r_{m}\right)}{\mathrm{Var}(r_{m})} \\ &= \frac{1}{E_{U}} \left[ \frac{\mathrm{Cov}\left(\mathrm{EBIT}(1-\tau), r_{m}\right)}{\mathrm{Var}(r_{m})} - \frac{\mathrm{Cov}\left(\mathrm{CAPEX}, r_{m}\right)}{\mathrm{Var}(r_{m})} + \frac{\mathrm{Cov}\left(\mathrm{Depreciation}, r_{m}\right)}{{\mathrm{Var}(r_{m})}} \right] \\ &= \frac{1}{E_{U}} \left[\beta_{\mathrm{EBIT}(1-\tau)} - \beta_{\mathrm{CAPEX}} + \beta_{\mathrm{Depreciation}}\right]. \end{align} $$

We next assume that that the correlation between the market, CAPEX, depreciation, net debt and interest is $0$. That is, we assume that $\beta_{\mathrm{CAPEX}} = \beta_{\mathrm{Depreciation}} = \beta_{\mathrm{Net\ Borrowing}} = \beta_{\mathrm{Interest}} = 0.$ Thus the equation for unlevered, and by similar computation, levered beta are given by the following formulas:

$$ \begin{align} \beta_{U} &= \frac{1}{E_{U}} \left[\beta_{\mathrm{EBIT}(1-\tau)}\right] \implies E_{U}\beta_{U} = \left[\beta_{\mathrm{EBIT}(1-\tau)}\right] \\ \beta_{L} &= \frac{1}{E_{L}} \left[\beta_{\mathrm{EBIT}(1-\tau)}\right] \implies E_{L}\beta_{L} = \left[\beta_{\mathrm{EBIT}(1-\tau)}\right] \\ \end{align} $$

Equating the equations yields

$$ \begin{align} E_{U}\beta_{U} &= E_{L}\beta_{L} \\ \implies \beta_{U} &= \frac{E_{L}}{E_{U}} \beta_{L}.\\ \end{align} $$

Now, for an unlevered firm it is known that:

$$ A_{U} = L_{U} + E_{U} \implies E_{U} = A_{U} - L_{U}. $$

Say that the assets and liabilities of this firm are fixed with the exception of new debt capital issued. That is, the firm is levered and therefore, $$ \begin{align} E_{L} &= \left(A_{U} + \mathrm{Tax\ Shield}\right) - \left(L_{U} + D\right) \\ &= \left(A_{U} - L_{U}\right) - D + \mathrm{Tax\ Shield} \\ \implies E_{L} &= E_{U} - D + \mathrm{Tax\ Shield}.\\ \end{align} $$

Thus we are left to calculate the tax shield. Assume that the pre tax cost of debt is $k_{d}$. Therefore,

$$ \begin{align} \mathrm{Tax\ Shield} &= \sum_{i = 1}^{\infty} \frac{k_{d}D\tau}{(1+k_{d})^{i}} \\ &= k_{d}D\tau \sum_{i=1}^{\infty} \frac{1}{(1+k_{d})^{i}} \\ &= k_{d}D\tau \cdot \frac{1}{k_{d}} \\ \implies \mathrm{Tax\ Shield} &= D\tau.\\ \end{align} $$

Therefore

$$ \begin{align} E_{L} &= E_{U} - D + D\tau \\ \implies E_{L} &= E_{U} - D(1-\tau) \\ \implies E_{U} &= E_{L} + D(1-\tau).\\ \end{align} $$

Recalling that $\beta_{U} = \frac{E_{L}}{E_{U}} \beta_{L}$ and substituting our newly created equation for $E_{U}$ yields

$$ \begin{align} \beta_{U} &= \frac{E_{L}}{E_{U}} \beta_{L} \\ &= \frac{E_{L}}{E_{L} + D(1-\tau)}\beta_{L} \\ \implies \beta_{U} &= \left[\frac{1}{1 + \frac{D}{E}(1-\tau)}\right]\beta_{L}, \\ \end{align} $$

as required. Thanks.

$\endgroup$
1
$\begingroup$

This formula is a direct result of the Modigliani-Miller theorem. After some searching I found a fairly simple proof of this here:

https://quantcoyote.com/2017/04/09/modigliani-miller-unlevered-betas/

$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.