A0694
Title: Sequential monitoring for the relationship between global financial stability and the Fed
Authors: Cindy Shin Huei Wang - HSBC Business School, Peking University (China) [presenting]
Jingyuan Yao - Peking University (China)
Abstract: The aim is to propose an online cumulative sum of squares (CUSQ)-type test for monitoring the stability in the relationship between the Fed's communication of its forward-looking policy stance and global systemic risk via an autoregressive (AR) approximation of a mixed panel. This approach is easy to use because its limiting distribution follows the Brownian bridge, being free from model parameters and thus without the need for the bootstrap procedure and long-run variance framework to find the critical values, even the exact form of each series in this panel is unknown and varies across each of them. In particular, test statistics do not necessitate estimating the number of common factors and common factors and resulting factor loadings, which could mitigate the misleading outcomes caused by the issue of the spurious number of common factors and spurious breaks found by conventional retrospective tests in current literature. Monte Carlo simulations confirm the promising finite sample performance of our online test with various data-generating processes. The test is applied to examine the sequential online patterns in the relationship between the global systemic risk and monetary policy under an extension of a prior study's circumstance by using the event-study approach. Empirical evidence shows that the likely dates on which the structure of the relationship breaks are identified and consistent with the timing of common events occurrence
