Modigliani-Miller Theorem

The claim in one line

In 1958, Franco Modigliani and Merton Miller published a paper in the American Economic Review with a result that ran directly counter to what every working banker believed. They claimed: under perfect markets, the total value of a firm is independent of how it is financed. Two firms with identical assets and identical operating cash flows must have the same value, whether one is funded entirely by equity and the other carries 90 percent debt. Capital structure is irrelevant.

That conclusion was so disorienting that the paper spent its first ten pages on the assumptions, not the result. The four core conditions:

1. No taxes (corporate or personal).
2. No bankruptcy costs — default, if it happens, is costless and instantaneous.
3. No information asymmetry — managers and outside investors agree on the firm's prospects.
4. Perfect markets — frictionless trading, no transaction costs, individuals borrow and lend at the same rate as firms.

Under those conditions, Proposition I says the levered firm and unlevered firm have identical value:

V_L = V_U
V   = E + D     (firm value = equity + debt)

Capital structure simply slices the firm's cash flows between two groups of claimants. The total cash flows aren't changed by how they're divided — and in a frictionless market, value follows cash flows.

The replication proof — why arbitrage forces irrelevance

The M&M proof is a clean arbitrage argument. Suppose, for contradiction, that two otherwise identical firms had different total values — say V_L > V_U because the market mistakenly prices the levered firm higher.

An investor holding a fraction α of the levered firm's equity earns α × (X − r_d × D_L), where X is operating cash flow. That same investor can replicate the cash flows by:

1. Selling the levered equity for α × E_L,
2. Borrowing α × D_L on personal account at rate r_d,
3. Using the proceeds to buy α of the unlevered firm.

The resulting payoff is identical: α × X − α × r_d × D_L. But the portfolio cost is α × V_U − α × D_L, while the original position is worth α × E_L = α × (V_L − D_L). If V_L > V_U, the replication is cheaper than the original — a riskless arbitrage. Investors do the trade in size, push V_L down and V_U up, and the prices equalize.

Crucially, the proof requires "home-made leverage" — the investor must borrow at the same rate r_d that firms borrow at. Drop that assumption (because retail investors can't borrow as cheaply as corporations) and Proposition I starts to crack.

Proposition II — why cheap debt doesn't lower WACC

If V is constant, what happens to the cost of equity as a firm levers up? Modigliani and Miller's Proposition II answers exactly:

r_e = r_a + (D / E) × (r_a − r_d)

Where r_a is the cost of capital for the firm's assets (the unlevered cost of equity), r_d is the cost of debt, and D/E is the debt-equity ratio. The cost of equity rises linearly with leverage. Why?

Equity holders are residual claimants — they get whatever is left after debt is serviced. Layering on more debt makes earnings less stable from the equity holder's perspective: a 10 percent decline in operating cash flow shrinks equity returns more than 10 percent, because interest is fixed. So shareholders demand a higher expected return to compensate for the amplified risk.

The increase in r_e is exactly calibrated to offset the use of cheaper debt. Plug it into the weighted-average cost of capital:

WACC = (E / V) × r_e + (D / V) × r_d
     = (E / V) × [r_a + (D / E) × (r_a − r_d)] + (D / V) × r_d
     = r_a

WACC equals r_a — the unlevered cost of capital — regardless of leverage. This is the financial equivalent of conservation of energy: you cannot reduce the cost of financing the firm's assets just by rearranging the claims against them.

Worked example: levering up a firm

Consider a firm with a $100 million asset base, generating $10 million per year in perpetual operating cash flow. Suppose r_a = 10 percent. Two financing options:

Scenario	D ($M)	E ($M)	r_d	r_e (Prop II)	WACC	V = E + D
Unlevered (100 % equity)	0	100	—	10.0 %	10.0 %	$100 M
20 % debt	20	80	5 %	11.25 %	10.0 %	$100 M
50 % debt	50	50	5 %	15.0 %	10.0 %	$100 M
80 % debt	80	20	5 %	30.0 %	10.0 %	$100 M

Add up E and D in every row: $100 million. The firm's WACC never moves. Equity holders in the 80 percent-debt case demand 30 percent — three times the unlevered cost — because residual earnings are wildly volatile. That's M&M.

With corporate taxes — the 1963 correction

The cleanest M&M conclusion was, of course, immediately interesting because it was so clearly wrong: firms did appear to care about capital structure, debt did appear to make some firms more valuable, and bond markets were not paid for nothing.

In 1963, Modigliani and Miller published a follow-up paper that fixed the most important omission: corporate taxes. Interest payments are tax-deductible; equity dividends are not. So every dollar of interest a firm pays to bondholders shields a dollar of pre-tax income from corporate tax at rate T_c. The present value of that perpetual tax saving on a perpetual debt issue is:

PV(tax shield) = (T_c × r_d × D) / r_d = T_c × D

V_L = V_U + T_c × D

A firm with $100 million of debt, in a country with a 25 percent corporate tax rate, is worth $25 million more than an otherwise identical all-equity firm. Leverage adds value through the tax shield. And in this revised world — with only the corporate-tax distortion — the optimum is 100 percent debt.

Personal taxes complicate the picture further. Miller's 1977 paper "Debt and Taxes" showed that, if personal taxes on interest income are higher than personal taxes on equity returns, much or all of the corporate tax shield is clawed back at the investor level. The effective shield depends on the difference T_c − [T_pe − T_pd × (1 − T_c)], which Miller showed could be near zero in his preferred parameterisation. The empirical magnitude is still actively debated.

Bankruptcy and trade-off theory

Empirically, firms do not lever to 100 percent debt — not even close. The S&P 500 median debt-to-assets ratio sits around 30 percent. What stops them?

The answer is bankruptcy costs. Default isn't free, and the probability of default rises sharply with leverage. The costs come in two flavours:

• Direct costs. Legal fees, court costs, bankruptcy administration, and fire-sale prices on liquidated assets. Empirically these run 3-7 percent of firm value in formal bankruptcies — large enough to matter but not dominant.
• Indirect costs. The costs of financial distress — customers defecting to competitors, suppliers tightening terms, key employees jumping ship, management distracted, debt-equity conflicts producing underinvestment ("debt overhang"), and risk-shifting toward high-variance projects ("asset substitution"). Indirect costs are an order of magnitude larger; estimates run 10-30 percent of firm value for distressed firms.

Trade-off theory packages this into a simple optimum:

V_L = V_U + PV(tax shield) − PV(bankruptcy costs)

The marginal tax-shield benefit per dollar of debt is roughly constant at T_c. The marginal bankruptcy cost is rising — because the probability of default and the severity of distress both grow with leverage. The optimum sits where the rising bankruptcy slope crosses the flat tax-shield slope. For most US firms in the modern era, that's somewhere between 20 and 50 percent debt.

The pecking order and information asymmetry

Stewart Myers (1984) noted that one M&M assumption — symmetric information — was uniquely fragile. Managers know more about the firm's prospects than outside investors. When a firm issues equity, investors infer that management thinks the current stock price is high enough to make the issue attractive — that is, the stock is overvalued. The announcement of a seasoned equity offering drops the share price by 2-3 percent on average. New equity is informationally expensive.

Myers and Majluf (1984) formalised this in a model where investors and managers have asymmetric information about asset value. The conclusion: firms minimise the negative signalling cost by following a hierarchy:

1. Retained earnings first. No new claims issued, no information revealed.
2. Debt second. Debt is less sensitive to inside information than equity (it's a fixed claim), so its issue carries a weaker negative signal.
3. External equity last. Issue equity only when retained earnings and debt capacity are exhausted.

This is the pecking order theory. It explains a fact trade-off theory struggles with: profitable firms tend to have lower leverage, not higher. Under trade-off theory, more profitable firms should lever up to capture more tax shield. Under pecking order, more profitable firms generate enough retained earnings to fund themselves, never need debt, and look "underlevered" by trade-off standards. Microsoft and Apple are textbook cases.

Agency costs and other extensions

Jensen and Meckling (1976) added a fourth class of frictions to the M&M baseline: agency costs. Managers do not necessarily maximise firm value — they may consume perquisites, take pet projects, or empire-build. Two M&M-related arguments come out of this:

• Debt as a disciplining device. Mandatory interest payments take cash flow off the table that managers might otherwise misuse. Jensen's (1986) "free cash flow theory" argues that high-debt firms have less slack for managerial waste; this is a positive role for leverage that doesn't show up in pure M&M.
• Debt overhang and risk-shifting. Highly levered firms underinvest (because gains accrue mostly to bondholders) and over-take-risk (because the upside accrues to equity, the downside to bondholders). Both are agency costs of debt.

Market timing (Baker and Wurgler 2002) is a fifth deviation: managers issue equity when valuations are high and repurchase when they're low, leaving capital structure as the cumulative result of past timing decisions rather than an optimum.

M&M versus its corrections

Framework	Key friction	Predicts optimal D/E	Empirical strength
M&M (1958, no tax)	None	Indeterminate	Pure benchmark; not testable directly
M&M (1963, corporate tax)	Tax-deductible interest	100 % debt	Refuted by observed leverage
Miller (1977, with personal tax)	+ personal tax wedge	Indeterminate at firm level	Mixed; depends on tax regime
Trade-off theory	Bankruptcy costs vs tax shield	Interior optimum	Matches cross-industry patterns; struggles with profitability-leverage relation
Pecking order (Myers 1984)	Information asymmetry	No optimum; financing hierarchy	Strong for profitable firms; weaker for small/high-growth
Agency (Jensen-Meckling 1976)	Manager-investor conflicts	Optimum balances disciplining vs overhang	Important for large mature firms
Market timing (Baker-Wurgler 2002)	Mispricing	Path-dependent; no static optimum	Persistent effect of historical valuations on leverage

No single theory wins. Modern empirical work treats the four frictions as complementary: trade-off explains long-run leverage targets, pecking order explains short-run financing choices, agency explains why levered firms behave differently, and market timing explains the path. M&M is the gravitational background — every other theory measures a particular departure from it.

Mirror theorems in macroeconomics and finance

The M&M argument is a special case of a more general principle: in a frictionless market, the way you slice up cash flows doesn't change their total value. Several other "irrelevance" results have the same logical shape:

• Ricardian equivalence (Barro 1974). In a world without distortionary taxes or liquidity constraints, financing government spending by taxes versus by debt has identical real effects, because rational households save more in anticipation of future taxes. Same M&M logic, applied to government finance.
• Wallace-Sargent neutrality. Open-market operations that swap one government liability (money) for another (bonds) are irrelevant in a frictionless world. The "M&M of monetary policy."
• Dividend irrelevance (Miller-Modigliani 1961). The same authors showed that, under similar assumptions, dividend policy is irrelevant to firm value — investors can manufacture any cash-flow pattern they want by selling shares.
• Hedging irrelevance. A frictionless firm gains nothing from hedging financial risk; investors can hedge on their own account. Real corporate hedging programs only make sense because of taxes, distress costs, or information frictions.

All four irrelevance results are useful for the same reason: they isolate the friction that's actually doing the work. If dividends matter for valuation, it must be because of taxes, signalling, or clientele effects — not anything intrinsic to dividends. If hedging adds value, it's because hedging affects expected taxes paid, distress probability, or investment opportunities. M&M-style reasoning is the cleanest way to find the actual answer.

What the data say about capital structure

Modern empirical work on capital structure produces a few robust facts that M&M and its successors are jointly trying to explain:

• Leverage varies by industry. Utilities and REITs lever to 50+ percent; software and biotech firms run with negative net debt. Trade-off theory explains this via differences in tangibility (collateral value) and operating-cash-flow volatility.
• Profitable firms are less levered. The negative profitability-leverage correlation is the strongest empirical fact in the field — and it's the one trade-off theory struggles with most. Pecking order handles it naturally.
• Leverage is persistent. Firms' debt ratios are highly autocorrelated. Targets exist but are moved toward slowly — half-lives of 3-5 years in mainstream estimates. This is hard to reconcile with frictionless dynamic trade-off models.
• Equity issues are rare. Net equity issuance is a small fraction of corporate financing; most external finance is debt. Buybacks now dominate equity flows for mature US firms. Consistent with pecking order.
• Tax changes shift leverage. The 2017 US TCJA cut the corporate rate from 35 to 21 percent, reducing the marginal tax shield. Studies find modest but real reductions in leverage in affected firms — consistent with the tax shield being one input, not the only one.

Common pitfalls

• Forgetting Proposition II when arguing for low WACC. A common student mistake: "we should issue more debt because debt is cheaper than equity." Under M&M, more debt raises r_e exactly enough to offset the cheap debt. The gain is the tax shield, not the substitution effect.
• Confusing book and market values. M&M is a market-value statement. Book D/E ratios drift mechanically with accounting policy and accumulated retained earnings, telling you little about the firm's economic capital structure.
• Ignoring personal taxes. The corporate-tax shield can be clawed back at the investor level. The "effective" tax shield can be much less than T_c × D when investors face high tax rates on interest income.
• Treating debt as costless. Even before formal bankruptcy, the costs of financial distress (lost sales, supplier retreat, management distraction) start showing up as leverage rises. Distressed but solvent firms underperform their industries by 10-20 percent.
• Conflating M&M with "leverage doesn't matter." M&M says leverage doesn't matter under specific assumptions. Drop any one assumption (taxes, bankruptcy, asymmetric information, transaction costs) and leverage starts to matter in characteristic, well-understood ways.
• Assuming the firm can borrow at r_d perpetually. The cost of debt itself rises with leverage as default risk grows. r_d is endogenous in any realistic setting, and that's part of how WACC eventually rises — bankruptcy-friction WACC curves are U-shaped, not flat.