Financial Crisis Early Warning Based on Panel Data and Dynamic Dual Choice Model

Abstract

Based on the research of currency crisis pressure index, bank crisis pressure index, and asset bubble crisis pressure index, this paper introduces an external shock pressure index reflecting the impact of global economic changes on economy and synthesizes systemic financial crisis pressure based on the above four pressure indexes; then, all the alternative early warning indicators and the systemic risk pressure index constructed in this paper were tested for Granger causality. We build financial systemic risk pressure indexes, including currency crisis pressure (CCP) banking crisis pressure (BCP) index, bubble crisis pressure (PBP) index, and external shock pressure (ESP) index to predict financial crises. Finally, four indicators that have a significant impact on the systemic financial crisis pressure index were selected, namely, the stock price index change rate, industrial added value growth rate, domestic and foreign real deposit interest rate differential, and foreign direct investment as a percentage of GDP. A dynamic Logit model with lagging binary variables is constructed, and compared with the traditional static Logit line, the actual dynamic fitting effect is better than the static Logit model. The dynamic Logit model is used to predict the early warning status of systemic financial crisis in 2020, and the forecast of various early warning indicators is realized by the ARIMA model. The final prediction results show that the probability of a systemic financial crisis in China in 2020 is extremely low, almost zero. This is in line with the overall improvement in the international economic situation in 2020 and the steady growth of the domestic economy.

Predicting Financial Fragility in Nigeria: Does the Filter Method Matter?

“…Comparably, the AUC and distance to corner also show that credit-to-GDP gap outperforms Non-performing loans, loan to deposit ratio and asset prices as EWIs of financial crisis no matter which filter was used. This result supports several earlier studies including Du [5], Sally and Katsiaryna [6] for advanced economies, Drehmann and Yetman [3], Beutel, List and Schweinitz [44], Giordani and Simon [45], Geršl, and Jašová [46] among others. However, the result of the present study is at variant with Sally and Katsiaryna [6] whose study indicated that asset prices do better for emerging economies and Ihejirika [4] whose findings show that Creditto-GDP Gap performs poorly in Nigeria.…”

“…This study focuses on asset prices, Non-Performing Loans and Loan to Deposit Ratio. The literature on credit-to-GDP ratio and the derived gap has received much attention from scholars [1,2,3,4,5,6] among others).…”

“…Du [5] set out to predict financial crisis in China using "stock price index change rate, industrial added value growth rate, domestic and foreign real deposit interest rate differential, and foreign direct investment as a percentage of GDP". The forecast result show that the probability of a systemic financial crisis in China in 2020 was extremely low, almost zero.…”

“…The forecast result show that the probability of a systemic financial crisis in China in 2020 was extremely low, almost zero. Du [5] used the traditional static Logit line to compare with the dynamic Logit models where the lagging binary variables were constructed. The actual dynamic fitting effect was better than the static logit model.…”

“…When will the next financial crisis occur? The answer to this question has been on the mind of researchers [1][2][3][4][5][6]. In search of early warning indicators of financial crisis across financial jurisdictions the world over.…”