Managerial Risk Taking Incentives and Corporate Pension Policy

In our paper, Managerial Risk Taking Incentives and Corporate Pension Policy, forthcoming in the Journal of Financial Economics, we examine whether the compensation incentives of top management affect the extent of risk shifting versus risk management behavior in pension plans.

The employee beneficiaries of a firm’s defined benefit pension plan hold claims on the firm similar to those held by the firm’s debtholders. Beneficiaries are entitled to receive a fixed stream of cash flows starting at retirement. The firm sponsoring the plan is required to set aside assets in a trust to fund these obligations, but if the sponsor goes bankrupt with insufficient assets to fund pension obligations, beneficiaries are bound to accept whatever reduced payouts can be made with the assets secured for the plan.

Stockholders of firms approaching a state of distress, therefore, have incentives to underfund pension plans. Underfunding plans amounts to promising future benefits without funding them and is effectively a way of increasing leverage by borrowing from employees. These stockholders also have the incentive to make risky investments with plan assets. If the investments pay off and the firm survives, stockholders benefit from having to contribute less into the plan; if they do not pay off and the firm goes bankrupt, beneficiaries suffer. Thinking of beneficiaries as akin to debtholders, these are manifestations of the classic risk shifting incentives that stockholders of all leveraged firms have (Jensen and Meckling, 1976; Myers, 1977).

The regulatory environment of corporate defined benefit plans in the United States exacerbates these risk shifting incentives. Under the Employee Retirement Income Security Act of 1974 (ERISA), most defined benefit pension plans are insured by the Pension Benefit Guaranty Corporation (PBGC). If a plan sponsor goes bankrupt with an underfunded plan, the PBGC takes over the plan and makes up the funding deficit, up to a limit. This guarantee mutes the incentives of rank-and-file beneficiaries to monitor the management of their plan, fueling the moral hazard problem. Hence, strong reasons exist to expect US defined benefit sponsors to increase plan risk (both by underfunding and by increasing plan asset risk) as they approach distress.

The extant empirical evidence in this regard has, however, been surprisingly weak. While some studies find evidence of risk shifting, the majority finds a negative association between firm risk and pension risk, consistent with risk management (or risk offsetting), not risk shifting. Given the strong theoretical predictions for risk shifting, this conflicting empirical evidence creates a puzzle.

We propose one explanation for this puzzle: managerial risk aversion. While diversified stockholders have incentives to increase firm risk at the expense of debtholders, most corporate decision making is in the hands of managers, who prefer less risk than stockholders, out of concern for their reputation, undiversifiable human capital, or private benefits of control. The stockholder-manager conflict on risk could thus offset risk shifting incentives arising from the stockholder-debtholder conflict.

Compensation contracting is one of the primary means of altering managerial incentives, and firms vary in the extent to which executive compensation aligns managers’ risk preferences with those of stockholders. Equity-based compensation increases the sensitivity of managers’ wealth to stock price performance (delta), and so aligns managers closer to stockholders, but could also lead managers who are under-diversified in firm-specific wealth to avoid risk. Options add convexity to managers’ payoffs and, by increasing the sensitivity of managerial wealth to firm risk (vega), can offset the risk-avoiding tendencies introduced by delta and by reputation or human capital concerns. If pension funding and investing choices are attributable at least partly to managerial incentives, we would expect to find more risk shifting in firms in which top managers have high vega.

Top managers, however, also participate in the broad-based pension plans covered by ERISA (ERISA-qualified plans). If managers are concerned with safeguarding their own pensions, then their vega incentives need not necessarily exacerbate pension risk shifting, making the ultimate effect of managers’ equity incentives an interesting empirical issue. Hence, we first examine how chief executive officer (CEO) equity incentives (vega, option delta, and stock delta) affect risk shifting through pension funding and asset allocation.

Pension funding and investing decisions interact closely with core financing and investing decisions. For example, required contributions to pension plans affect resources available for financing other investment opportunities (Rauh, 2006). Decisions on how to invest pension assets affect not only future cash requirements, but also the firm’s overall risk profile. Finally, ERISA-qualified pension plans offer tax-saving opportunities, making their management a key component of firm tax planning. As leaders of the finance function, chief financial officers (CFOs) play a key role in budgeting and managing cash flows, risk management, and tax strategy, and they could thus affect pension investing and financing. Hence, we also examine whether CFO equity incentives affect pension choices.

We examine pension funding (asset allocation) with a sample of 5,748 (4,398) firm-years spanning 1999–2010. We find that managers’ equity incentives affect the extent of risk shifting versus risk management behavior in defined benefit pension plans. Risk shifting by underfunding plans (and, to a lesser extent, by investing plan funds in risky asset classes) is stronger when top managers have high wealth-risk sensitivity (vega) and weaker when they have high wealth-price sensitivity (delta). Conversely, risk shifting by underfunding plans is weaker when top managers have a larger stake in the plans that is at risk if the plan fails. These findings are stronger for CFOs than for CEOs, suggesting that pension policy falls within the CFO’s domain in most firms.

These results highlight an important driver of corporate pension policy in the US. They also identify plans in which risk shifting behavior manifests most strongly and the moral hazard fueled by PBGC insurance is of particular concern. Understanding further the governance factors that determine pension risk shifting, particularly through asset allocation practices (which remain little understood), is an interesting and important area for future research, especially at this time of deteriorating PBGC finances.

The full paper is available for download here.