Job market paper
Connections as Jumps: Estimating Financial Interconnectedness from Market Data
Abstract: I develop a new methodology for measuring interconnectedness between financial institutions using readily available market price data. I argue that the classic endogeneity problem that arises when using contemporaneous price movements can be addressed by focusing on connections that trigger substantial spillovers upon default. As spillovers are statistically similar to jumps (sudden, discontinuous reductions in the underlying asset value of the firm), the effects of such connections are found in the jump-like default risk of a firm. Importantly, the remaining default risk is not exposed to such connections. Therefore, under appropriate identification assumptions, regressing jump-like default risk on non-jump-like default risk uncovers causal evidence of direct and indirect exposures. In my empirical work, I adapt existing techniques for estimating jump risk to a model in which the firm is a levered claim on a latent asset, and use equity, equity options, and credit default swap data on large US financial institutions to isolate jump-like default risk. Applying the methodology to the largest financial firms during the 2008 Financial Crisis, I find estimates of connections that are consistent with well-known developments during the crisis: firms change positions in the network in line with their risk and access to support programs. My estimates further provide evidence that market participants viewed the collapse of Lehman Brothers as a symptom, rather than the cause, of the crisis. The methodology I develop in this paper provides a new tool for monitoring the financial sector in real time using contemporaneous market price changes.
Research work in progress
Core-Periphery Networks as an Endogenous Outcome: Balancing Systemic and Idiosyncratic Risk (early stage)
Abstract: I develop a new mechanism to generate a core-periphery network structure from homogeneous financial agents (banks). In the model, banks have independent stochastic liquidity shocks, and wish to insure themselves against liquidity shortfalls by agreeing to share surpluses. To do so, they form pairwise linkages. A linkage gives the bank access to its counterparty's potential liquidity surplus, but also exposes the bank to the default risk of its counterparty. Numerical simulation suggests that such an environment supports a core-periphery network structure. Through its liquidity sharing, the fully connected core is relatively insulated from idiosyncratic shocks, but defaults tend to be highly correlated (systemic risk). The periphery is riskier, but its defaults are independent (idiosyncratic risk). Core members have an incentive to link to a periphery member to get an idiosyncratic source of liquidity during systemic events, and periphery members have an incentive to link to core members to get a relatively safe counterparty.
Unscaled Ridge Regression in a Sparse Setting
Abstract: I derive a new, data-driven methodology for choosing an optimal ridge regression parameter under sparsity assumptions. Given a sparsity structure (the probability that any particular coefficient is non-zero and the expected magnitude of non-zero coefficients), I derive a minimization problem for the ridge penalty parameter that yields the lowest expected mean-square-error of the estimated coefficients. Though not in closed form, this univariate minimization problem is simple to solve numerically. The optimal penalty depends on 1) the covariance structure of the predictor variables, 2) the estimated residual variance, and 3) the sparsity structure specified. The methodology does not rely on predictors being scaled to have unit variance, and is therefore readily applicable to cases where the magnitudes of the coefficients are directly comparable and the researcher wishes to minimize the mean-square-error in the original scale of the variables.