I'm a fourth year PhD candidate in Finance under the supervision of Prof Franklin Allen and Prof David Miles. My research interests are systemic risk and macroeconomics. In my two recent papers, I study how a coordination problem can give rise to endogenous risk, and are of interest theoretically as well as relevant for policy. I did an MPhil in economics at the University of Oxford (2013-2015), and worked at the Bank of England to develop a new stress test, modelling liquidity hoarding (summer 2017).
Work in progress
Almost pure endogenous fluctuations
This paper explores under what conditions can a coordination problem, captured in the stylized model of Morris and Shin (2002), produce an amplification mechanism of fundamentals. Previous work by Angeletos and La'o (2013) has shown that if strategic complementarity is constant, the variance of equilibrium actions is bounded from above by the variance of fundamentals - that is, the coordination problem is not an amplification mechanism. I study this model, but allow strategic complementarity to vary with the state. I find that the variance of equilibrium actions is no longer bounded from above by the variance of fundamentals. In the extreme case, I show that in equilibrium there remains `non-negligible' variation in actions, even when variation in fundamentals is `negligible'. Furthermore, I embed the model in a macroeconomic setting with three frictions: monopolistic competition, nominal wage rigidity and lack of common knowledge. If returns to scale are constant, and the deviation from common knowledge is `small', best responses of firms are linear functions of both fundamentals and expectations of others - that is, the macroeconomic model can be mapped to the model explored in the paper. The result is that macroeconomic fluctuations can result from minute changes in fundamentals. Finally, I extend the model to an infinite horizon OLG, and study asset pricing implications.
Seniority and Systemic Risk
This paper investigates how seniority of interbank liabilities vis à vis short-term creditors, such as the one implied by insolvency netting, affect the probability of a financial crisis. Netting is often seen by regulators as enhancing systemic stability by weakening cascading losses through the financial system (domino-effect). I explore under what conditions this statement is true when the bank run channel is taken into view, and find that the effect of netting on the probability of bank runs is ambiguous. Intuitively, netting increases banks' recovery rates at the expense of other creditors such as depositors. This effect has to be considered in equilibrium, in which those banks in themselves have short-term creditors. To the extent netting effects a transfer only from bankrupt to bankrupt banks, financial stability is actually decreased by netting, because it increases expected welfare of short term depositors who choose to run on the bank, thereby incentivizing early withdrawal. In the more plausible case in which equilibrium encompasses states in which there both failed and non-failed banks, the effect of netting becomes ambiguous.
Measurement: Centrality Matters
My model provides performance based ranking of systemic importance measurements that rank nodes in the financial system according to observables. One of the surprising results is that traditional ways of assessing systemic importance under-performs almost any other importance measurement tested. Other than that the main message is that centrality measurements perform better than a random benchmark -- i.e. they contain useful information that can be utilized by a policy maker.
Application: Liquidity Hoarding, a Dynamic Stress Test model
This is an application made in association with the Bank of England. It develops a multi-asset dynamic stress test model in which banks make decisions about their liquidity buffer: both size and composition (how much of each security to hold). Decisions are made both on public and private information, and are essentially thumb rules that attempt to describe bank behaviour in the daily practice of liquidity management. Having decided how much of each security they wish to hold, banks engage in trade with one another and the outside world to accomplish their demand/supply. The model is aimed to provide quantitative estimation of stress from a liquidity perspective: how much reliance is there on the central bank?