PhD Defense Jacob Bjerre Skov

Title: Dynamic Term Structure Modeling and the LIBOR Transition

Abstract:

This thesis deals with the topic of term structure modeling with a focus on the current LIBOR transition. We begin by studying the SOFR futures market in the context of Gaussian dynamic term structure models. We show that, despite jumps and spikes in the overnight benchmark, a continuous Gaussian arbitrage-free Nelson-Siegel model is able to fit the term structure of SOFR
futures rates. Furthermore, using a shadow rate extension, we find that accounting for the zero lower bound has significant impact on volatility. Next, we develop a multi-curve model, which endogenously generates the spreads between different tenors of secured and unsecured rates. We show that the model is able to simultaneously fit the term structure of SOFR, Federal Funds, and Eurodollar futures rates as well as spot USD LIBOR and term repo rates. The framework models the spreads as roll-over risk, which we further separate into a credit and funding-liquidity component allowing us to decompose the LIBOR-OIS spread. The last paper considers the task of accurately modeling both overnight and term rates based the SOFR benchmark. We show that a continuous model is unable to simultaneously reflect the near piecewise constant dynamics of overnight SOFR and the
diffusive dynamics of SOFR futures rates. Instead, we construct a model of scheduled and unscheduled jumps in the short-rate reflecting jumps in the central bank policy rate following scheduled and unscheduled FOMC meetings. Accounting for jumps in the short-rate, the jump model is able to reconcile the overnight and futures rate dynamics.

Thesis for download

Assessment committee:
Jesper Lund Pedersen (chairperson)
Associate Professor, University of Copenhagen Anders Bjerre Trolle
Professor, Copenhagen Business School
Jens Henrik Eggert Christensen
Research Advisor, Federal Reserve Bank of San Francisco

Supervisor:
David Glavind Skovmand
Associate Professor, University of Copenhagen