Agency Costs

Agency costs are the costs that arise when the people who run an organisation (agents) do not share the interests of the people who own or fund it (principals). They include money spent to watch over and constrain managers, money managers spend to reassure owners, and the value lost because, despite these efforts, managers still make choices owners would not. Agency costs are the measurable price of the principal-agent problem.

The concept was formalised by Michael Jensen and William Meckling in a 1976 paper in the Journal of Financial Economics, "Theory of the Firm: Managerial Behavior, Agency Costs and Ownership Structure", one of the most cited works in finance. It is a cornerstone of corporate finance and of the economic theory of the firm, and it underlies debates over executive pay, board independence, leverage and dividend policy.

Imagine you own a small café but live in another city, so you hire a manager to run it. You pay an accountant to check the books and install cameras at the till; those are costs of watching. The manager agrees to a contract with penalties if the café fails an inspection; that is a cost of reassuring you. And even with all that, the manager may still order the expensive coffee machine they like or close early on quiet days.

All three together, the checking, the promises and the leftover waste, are agency costs. They exist only because the person running the business is not the person who owns it.

Jensen and Meckling split total agency costs into three parts:

• $\text{AC}$ is the total agency cost borne by the owners.
• $M$ is monitoring expenditure by the principal, such as audits, board oversight and compliance systems.
• $B$ is bonding expenditure by the agent, such as reporting commitments, guarantees and contract restrictions the agent accepts to win trust.
• $L$ is the residual loss, the value still lost because the agent's decisions differ from those the principal would have made.

Suppose a company spends $2 million a year on external audit and board oversight ($M$), management spends $0.5 million on extra investor reporting it has promised ($B$), and executives still pursue an unprofitable expansion that destroys $3 million of value ($L$). Total agency costs are $2 + 0.5 + 3 = 5.5$ million dollars a year. In the model, owners keep adding monitoring and bonding until a further dollar spent reduces the residual loss by less than a dollar, so agency costs are never zero.

Jensen and Meckling showed that a manager who owns all of a firm bears the full cost of any perk, but one who sells part of the firm to outside investors bears only part of it, so selling equity creates agency costs. Debt creates different ones:

• Asset substitution. Because limited liability caps shareholders' losses, a heavily indebted firm may switch to riskier projects that benefit shareholders if they succeed and creditors bear the loss if they fail.
• Debt overhang. Stewart Myers argued in 1977 that firms with heavy debt may pass up profitable investments, because most of the gain would go to repaying a creditor.
• Free cash flow. Jensen argued in 1986 that managers of cash-rich firms tend to waste surplus cash on low-return projects, and that paying it out or taking on debt disciplines them. This view helped justify the leveraged buyout wave of the 1980s and the private equity model.

• Incentive pay. Equity awards, share options and performance-linked bonuses tie managers' pay to owners' outcomes.
• Clawbacks. Under SEC Rule 10D-1, adopted in October 2022 under the Dodd-Frank Act, US-listed companies had to adopt policies by 1 December 2023 to recover incentive pay awarded on the basis of financial results that are later restated.
• Say on pay. Shareholders in the US have had advisory votes on executive pay since 2011, and UK shareholders have had a binding vote on pay policy since 2013.
• Board structure and audit. Independent directors, audit committees, an independent auditor and the internal-control rules of the Sarbanes-Oxley Act of 2002 strengthen monitoring.
• Outside pressure. Institutional investors, activist investors, the takeover market and debt covenants all constrain managers.

Agency costs are hard to measure, especially the residual loss, so researchers often rely on proxies such as expense ratios or asset turnover. Critics also argue that tools meant to cut agency costs can create new ones: large stock-based awards may encourage short-termism, earnings management or share buybacks timed to hit pay targets, and heavy leverage can lead to asset stripping. Supporters of stakeholder capitalism argue that the framework's focus on shareholders as the only principals is too narrow, while defenders reply that clear accountability to owners is what keeps managers' discretion in check.

• SEC: Compensation Recovery Listing Standards and Disclosure Rules
• The Nobel Prize: Economic Sciences 2016