top of page
Working Papers
Electoral Entrenchment and Municipal Bond Prices
Job Market Paper
Abstract:
Political parties often initiate policies designed to entrench themselves and limit the power of the opposition. I show that municipal bond investors value electoral entrenchment. An increase in electoral entrenchment lowers municipal bond spreads, regardless of which party benefits. The evidence indicates that rising agency frictions between constituents and their representatives drive the effect, particularly in states with substantial pension underfunding. Overall, my results suggest that a reduction in politicians' responsiveness to the will of the voters, caused by electoral entrenchment, allows politicians to make unpopular public finance decisions that bondholders value.
Pension Risk and State Borrowing Costs
Draft available upon request.
Abstract:
I show that increased investment risk-taking by US state public pensions is reflected in higher borrowing costs for the state. I measure borrowing costs as municipal bond tax-adjusted spreads and credit default swap spreads. I use the pension funds' 5 % value at risk as a measure of risk. A one standard deviation increase in the value at risk of a state's pension system results in a 4.6 basis point increase in the municipal bond tax adjusted spread and a 9.5 basis point increase in the CDS spread. The relationship I identify is robust and statistically significant, controlling for the funding level of the plans, the state's revenue and expenses, and revaluing the funds' liabilities to reflect finance principles better. Additionally, within municipal bonds, I demonstrate an interaction between the plan's funding status and value at risk such that the worse a pension system is funded, the more risk-taking matters for the total effect on the tax-adjusted spread.
Publications
How Much Can Collective Defined Contribution Plans Increase Risk-Sharing?
With Deborah Lucas
Journal of Investment Management Vol. 18, No. 4 (2020): 1-27
Abstract:
Collective Defined Contribution (CDC) plans have been suggested as an attractive and sustainable alternative to public sector DB plans. A CDC plan is a hybrid structure, designed to provide more predictable retirement benefits than a traditional DC plan while operating at the lower cost of a DB plan. It does this by sharing investment risk across worker cohorts and centralizing asset management. We develop a model of an unsubsidized CDC plan and use it to characterize the risk-sharing rules and investment policies that maximize a “scheduled benefit” for retirees that is almost always achieved or exceeded. We compare the outcomes under the CDC system with those from an otherwise similar options-augmented DC model, where participants have access to self-financing strategies that involve trading in one-year put and call options. The ability to effectively trade long-dated options in the CDC framework delivers a somewhat higher scheduled benefit than can be achieved by self-insuring in an options-augmented DC plan. However, under current contribution policies, the scheduled benefit in the CDC plan falls short of what most would consider an adequate retirement income
Work-in-Progress
Local Fiscal Policy
With Nancy Xu
How Sustainable Are Public Pension Plans? A Stochastic Analysis of Policy Alternatives
Abstract:
How concerned should policymakers be over public pension underfunding, and which remedies offer the best path forward? In this paper, we argue for using stochastic analysis of pension plan cash-flows over time to address these questions. We construct a detailed Monte Carlo simulation of 40 state and local pension plans using plan-provided data for 2017 and 2022. Compared to a stock measure of the funding level—the ratio of assets to reported liabilities- we can quantify the timing and severity of distress events. Contrasting our results with a deterministic cash-flow analysis, we confirm previous research that most public pensions are sustainable on a cash-flow basis on average. Still, because our analysis allows us to construct the entire distribution of outcomes, we show that the likelihood of entering a distress scenario is still meaningful for many plans. Finally, we use our model to test out some commonly proposed remedies to reduce the probability of default.
bottom of page