Michael Jensen's 1986 argument is that cash is dangerous in the hands of a management team with no positive-value uses for it. Executives are paid, promoted and written about in proportion to the size of what they run, and a shrinking company is a career risk in a way that a value-destroying acquisition is not.
The arithmetic is unforgiving. A rupee invested at a return of r in perpetuity, against a cost of capital k, is worth r/k rupees. Chandori's cost of capital is 13.5%. A diversification that earns 8% therefore destroys 1 − 0.08/0.135 = 0.407407 rupees of value for every rupee committed. Put ₹600 crore a year into it and ₹244.444444 crore evaporates annually. Do it for five years and the total is ₹1,222.222222 crore, which is 61.25% of the unlevered value of the entire company.
How does debt reduce the agency cost of equity?
Debt is a commitment device against exactly this. Interest and amortisation are contractual, so cash that would otherwise be available for an empire leaves the building on a schedule, and the discipline is enforced by a party with the legal power to act. Suppose leverage rising from 1.2 to 2.4 turns forces the annual discretionary spend down from ₹600 crore to ₹200 crore for five years. The value saved is 5 × 400 × 0.407407 = ₹814.814815 crore. The extra expected distress cost of that move, from the trade-off grid, is ₹92.589653 crore. The bonding is worth ₹722.225161 crore net.
That number is far too big to be believed as a point estimate and is the right order of magnitude for the mechanism. It is also another reason the naive trade-off model under-predicts leverage: it prices only the tax shield on the benefit side and ignores the discipline entirely.
The fingerprint is easy to screen for: high and stable free cash flow, low reinvestment opportunity, low leverage and a history of unrelated acquisitions. Leveraged-buyout sponsors have hunted that profile for forty years, and a buyout is the agency argument executed rather than written down.
Advertisement
The agency costs of debt: two conflicts, both with payoff tables
Once a firm has both debt and equity outstanding, the two claims have different payoff shapes on the same assets, and where the shapes differ the incentives differ.
Equity in a levered firm is a call option on the assets, struck at the face value of the debt. Below the strike the owners get nothing and cannot get less than nothing, which is what limited liability means. Above it, everything is theirs. Option holders like volatility, and that single fact generates both conflicts.
The first is asset substitution, sometimes called risk shifting. After the debt is priced and sold, the owners swap a safe project for a risky one with a lower expected value. Because their downside is already truncated, the extra dispersion is worth more to them than the lost expected value. The worked example on Braddock Chemicals runs the table: a switch that destroys $5 million of firm value raises equity by $25 million and cuts debt by $30 million. Of every dollar the lenders lose, 83.3333 cents reaches the owners and 16.6667 cents simply disappears.
The second is debt overhang, or underinvestment. A firm close to distress refuses a project with a positive net present value, because most of the value created accrues to lenders whose claim moves from doubtful to secure, while the owners fund all of it. Deccan Roadways is the case: a project with a net present value of ₹50 crore is turned down because the owners' share of the gain is ₹75 crore against an outlay of ₹100 crore, leaving them ₹25 crore worse off, while the lenders take ₹75 crore of a gain they did not pay for, which is 50% of the value the project creates.
How are both conflicts priced and contained?
Two properties of both conflicts are worth memorising. They exist only when default is a live possibility, which is why they are irrelevant to an investment-grade issuer and central to a stressed one. And the cost falls on the borrower, in advance. Lenders who can foresee the substitution price the debt as though it will happen, so in the Braddock case they lend $70 million against a $100 million face value, an implied yield of 42.857%. A covenant that credibly rules the switch out is worth the whole $30 million to the borrower, which is why covenants exist and why a borrower with a good story fights for the tightest set it can live with.
The remedies each conflict admits are different, and knowing which is which is the analytical payoff. Asset substitution is contained by restrictive covenants, security over specific assets, and shorter maturities that force a repricing. Overhang is contained by new money with priority over the existing claim, by a write-down of the existing face value, or by ring-fencing the project in a separate vehicle with its own claim. On Deccan's numbers, writing the face value down from ₹800 crore to ₹600 crore makes the owners exactly indifferent but costs the lenders ₹50 crore. Funding the project themselves with a priority claim leaves them ₹25 crore better off and the owners ₹25 crore better off, the two gains summing to the ₹50 crore of net present value that was going to be thrown away. Coordination beats concession, when the lenders can coordinate.