The myths of marked-to-market accounting

It is no magical pension key and might undermine system

Marked-to-market pension accounting is not gaining popularity these days.

Financial statements that use the principles of marked-to-market pension accounting show growing pension liabilities that are virtually meaningless. Funding regulations that use the principles require painful contribution increases while these resources could be used more productively elsewhere.

Overall, marked-to-market pension accounting creates significant difficulties for plan sponsors.

Regulators have taken notice. For corporate defined benefit plans, the discount rates mandated by the legislation Moving Ahead for Progress in the 21st Century, signed into law by President Barack Obama July 6, are based on 25-year average corporate yields with a “corridor” — a clear rejection of MMPA. For public defined benefit plans, the Governmental Accounting Standards Board in adopting new pension accounting standards June 25 declined to implement fully the marked-to-market approach for measuring pension commitments. In Europe, regulators in several countries are taking steps to mitigate the negative impact of marked-to-market pension accounting. It is becoming increasingly clear that MMPA might be detrimental to the health of retirement systems.

In light of these developments, it is informative to inquire why a seemingly good idea in theory — to account for current pension obligations as the market prices of matching bond portfolios — has not worked well in practice. Clearly, it is reasonable to benchmark a plan that intends to settle its pension commitments by investing in a matching bond portfolio to the market price of this portfolio. The proponents of marked-to-market pension accounting, however, wish to expand it to all defined benefit plans — public and private, ongoing and frozen, etc. — regardless of the intentions of these plans. This expansion is based on several myths, including:

c Marked-to-market pension accounting is good public policy. It is hard to find any evidence of MMPA's beneficial impact on retirement systems. To the contrary, retirement systems that implement MMPA normally experience numerous plan closures, freezes and terminations. Moreover, there is some evidence to suggest the management of pension assets implied by the MMPA mindset leads to suboptimal allocation of capital.

c MMPA promotes better pension plan management. MMPA produces confusing and inexact measurements of the quality of pension plan management. In particular, the movements of interest rates that are a major factor in the MMPA measurements are outside of control of pension plan managers.

c Matching assets and MMPA liabilities is in the best interests of plan participants. Optimal policy portfolios generated by asset-liability matching generally neither maximize the safety of benefits, nor minimize the cost of funding these benefits. Instead, these portfolios minimize the volatility of the plan sponsor's financial statements, which is not necessarily the most sensible goal of pension investing.

But the biggest myth is that the right pension accounting reform is the solution to all problems that plague defined benefit plans. The origins of this myth are unclear, even though many appear to believe in it. To those who have faith in the special healing powers of “economically proper” pension accounting, e.g., marked-to-market pension accounting, the following reminders should be in order.

Conventional accounting is an inherently backward-looking undertaking. In contrast, a pension valuation is an inherently forward-looking undertaking. No accounting reform can resolve this conflict. In this context, the phrase “pension accounting” is an oxymoron.

Over the years, there have been numerous fine-tunings of the reporting and funding requirements for defined benefit plans. Some of them moved these requirements toward marked-to-market pension accounting; some went in the opposite direction. Every one of them had a reason du jour: to ensure or improve solvency, transparency, predictability, comparability, stability, etc. Yet, the only tangible result of these activities appears to be the growing negative sentiment toward defined benefit plans among the majority of plan sponsors.

Objectives matter. If the objective of a reporting framework is to determine winners and losers in the “pension game,” then there is a real risk that the game would be finally over as all players exit it. At some point, creative regulators might design a perfect reporting framework for defined benefit plans and discover that it applies to no one. An entirely different framework might be required if the objective is to keep the players in the pension game indefinitely.

In a healthy retirement system, the objective of a defined benefit plan is twofold: to maximize the safety of pension benefits and minimize the cost of their funding. The system's reporting and funding requirements should be based on the measurements of the ability of the existing assets and future contributions to fund the benefits. Other objectives should be of lesser importance.

The vast majority of defined benefit plans endeavor to fund their pension commitments by investing in risky assets. Defined benefit plans need major advances in risk management rather than pension accounting. Any attempt to account transparently for the future behavior of risky assets is futile. The future is not transparent. There is nothing anyone can do about it. n

Dimitry Mindlin is president of CDI Advisors LLC, Oak Park, Calif.

This article originally appeared in the October 29, 2012 print issue as, "The myths of marked-to-market accounting".