Economic resilience in California against two earthquakes

Abstract

Economic resilience in California to destructive earthquakes is considered as an indicator of regional sustainability. We measure the abnormal volatility in stock markets, during the 1989 Loma Prieta Earthquake and the 1994 Northridge Earthquake as a series, by event study of statistical model, auto regressive moving average (ARMA) and generalized auto regressive conditional heteroscedasticity (GARCH) of economic model, and long short-term memory (LSTM) of deep learning model. The test results show there is no effect on the whole market; the sectors of Basic Industries Companies, Capital Goods Companies and Energy Companies are more sensitive to destructive earthquakes; on the level of stocks, the test result of the statistical model fit that of the economic model with percentage over 50%, and the result of the deep learning model is different. The relation between the company headquarters of the affected stocks with macroseismic intensity is unclear; those located around San Francisco Bay are sensitive to destructive earthquakes in California.