Research

International Shadow Banking and Macroprudential Policy (Job Market Paper)

Abstract: The Great Recession featured a global collapse in real and financial economic activity that was highly synchronized across countries. Two unique precursors to the crisis were the rise in the shadow banking sector and increased securitization. I develop a model that is the first to explain the extent to which these factors contributed to the international transmission of the crisis that mostly originated in the United States. Using a two-country model with commercial and shadow banking sectors, I show that a country-specific financial shock leads to a simultaneous decline in real and financial aggregates in both countries. My model is the first to include both shadow and commercial banking in an open-economy framework. While commercial banks transfer funds from borrowers to lenders, shadow banks securitize loans and sell them to intermediaries internationally as asset-backed securities. Transmission occurs through a balance sheet channel, which is stronger when intermediaries hold more securities from abroad. I also consider the implications of capital controls on the transmission of a financial crisis. In general, I find that capital controls can reduce transmission.

Differences of Opinion, Liquidity, and Monetary Policy

Abstract: Liquidity plays a crucial role in the relationship between asset prices and monetary policy.  Differences of opinion among investors regarding the future value of an asset can affect the liquidity of the asset in question.  I consider a monetary framework in which money and a risky asset can facilitate trade, but unlike money, the risky asset is opinion-sensitive in that traders may disagree on its future value.  A pecking order theory of payments is established: optimists prefer to use money to facilitate trade, whereas pessimists prefer to use the risky asset.  In contrast to much of the asset pricing disagreement literature, the demand for the opinion-sensitive risky asset is composed of not just optimists, but also pessimists, where the latter is driven by the asset’s usefulness in facilitating trade.  Additionally, in support of Bernanke and Gertler (2000), I find that monetary policy aimed at reducing asset price volatility need not be welfare-maximizing.  Instead, the Friedman rule is welfare-maximizing.

Contagion and Shadow Banking

Abstract: On March 17, 2008, the Fed announced it would guarantee Bear Stearns’ bad loans and wanted JP Morgan to purchase Bear to prevent a bankruptcy.  Was the Fed justified in its concerns of this potential bankruptcy triggering severe adverse effects in the financial system as a whole?  I develop a model with multiple shadow banking sectors in which the deterioration of assets of a highly-leveraged shadow banking sector can result in a run not only in its own sector, but the entire shadow banking system. These shadow banking sectors hold entirely different assets, but obtain funds from an identical wholesale funding market. This degree of financial integration is sufficient for contagion and simultaneous runs so long as the sector holding the “toxic” assets is highly leveraged. I consider policy from a macroprudential perspective in the form of leverage restrictions and wholesale funding taxes, both of which can prevent runs altogether and “escape” a severe recession.

Fiscal Multipliers via Structured VARs with Recursive Ordering: Are they justified?

Abstract: The effectiveness of fiscal stimulus was greatly debated following the events of the Great Recession. Empirical estimates of fiscal multipliers suggest that where stimulus wasn't effective, it wasn't tried. However, there is little guidance as to which estimates carry the most credibility. The various estimates in the literature rely heavily on assumptions of recursive ordering, which could possibly be unjustified. I test the validity of recursive ordering in the context of estimating fiscal multipliers using a novel iterative projection instrumental variable identification mechanism. I find strong evidence rejecting every possible recursive ordering among common endogenous variables used in the literature. Additionally, I find new estimates of fiscal multipliers using US data between 1960 and 2015 that don't rely on a Cholesky ordering. My estimates are conservative relative to those in the literature. Specifically, my estimated fiscal multipliers are economically and statistically significant after four quarters, but disintegrate within eight quarters. I also find that output responds to monetary policy with a delay of up to four quarters.