Working papers

Abstract:   Although public debt in the US is at a historical high, the effects of changes in public debt supply are unclear. There are two main reasons why. First, debt supply is jointly determined with the rest of the economy - the Treasury takes multiple economic factors into account while implementing debt management policy. Second, debt supply changes combine two distinct phenomena - a change in the total level of debt, and a change in issuance across maturities for a given level of debt. In this paper, I use a novel identification design to estimate shocks to the level and maturity structure of public debt supply in the US. Using high-frequency changes in Treasury futures prices around auction announcements, I first isolate exogenous changes in public debt supply. Next, I use narrative evidence to identify subsets of announcements that have high information on either debt level or maturity. Exploiting differences in the variance of level and maturity shocks across subsets, I estimate a factor model and separately identify the two shocks. I find that an increase in debt level and maturity leads to higher bond returns and yields. An increase in debt level leads to lower output and employment while an increase in debt maturity leads to higher output and employment. 

Abstract:  We develop a heterogeneous agent New-Keynesian model featuring endogenous non-normal labour income risk that we use to explain the moments and dynamics of equity and bond returns. Within our framework, frictional labour markets introduce the possibility of employed individuals experiencing job loss and significant, uninsurable declines in their labour income. Adverse shocks not only decrease dividends but also elevate the risk of endogenous job loss, rendering equity a risky asset as it performs poorly during periods of heightened idiosyncratic labour income risk. Consequently, our model generates a realistic equity premium in general equilibrium for reasonable levels of risk aversion, while maintaining the smoothness of aggregate consumption observed in the data. The risk-free rate is also low in equilibrium due to the strong precautionary saving motive of households, avoiding the risk-free rate puzzle. Finally, we also demonstrate the model's ability to replicate empirical findings attributing the majority of the stock market's response to a monetary policy shock to changes in discount rates rather than cash flows. 
Abstract:  This paper investigates the cyclical behaviour observed in the US nominal yield curve across business cycles and proposes a novel explanation. I empirically establish that variation in the nominal yield curve is mainly driven by fluctuations in the real component of yields. Employing a real business cycle model with time-varying disaster risk, I demonstrate that during recessions, the yield curve steepens due to the perception of short-term real bonds being relatively safer than long-term bonds. As the likelihood of a disaster event increases, short-term yields fall as short-term bonds become more valuable, aligning with higher marginal utility. Over time, the probability of disaster reverts to its long-term average, leading to stabilized future expected short-term yields and a decline in long-term yields on impact, albeit to a lesser extent. This process results in a steepening of the yield curve during economic downturns and a flattening during periods of expansion. I analytically characterize intertemporal elasticity of substitution (IES) values compatible with empirical observations. Contrary to prior assumptions, an IES greater than 1 is neither necessary nor sufficient to generate the facts. Overall, this study provides insights into the mechanisms driving yield curve behavior and sheds light on the implications of disaster risk for bond market dynamics.

Works in progress 

1.  Why does the Fed move long-term real bond yields? A model of recursive preferences and disaster risk

Pre PhD publication

Policy papers