2013年-世界发展银行全球_Duration_Dependence_and_Change-Points_in_the_Likelihood_of_Credit_Booms_Ending_48页_738kb

Summary of "Duration Dependence and Change-Points in the Likelihood of Credit Booms Ending"

Core Content

This paper investigates the duration dependence and change-points in the likelihood of credit booms ending, using a Weibull duration model over 71 countries from 1975q1 to 2010q4. It aims to understand whether the probability of a credit boom ending is influenced by its length and whether there are structural breaks in this relationship over time.

Main Goals

1. To determine if the likelihood of a credit boom ending is duration dependent.
2. To analyze whether there are change-points in the duration dependence parameter.
3. To differentiate between "good" and "bad" credit booms based on their outcomes (soft landing vs. systemic banking crisis).

Key Findings

• Duration Dependence: The likelihood of credit booms ending increases over time, indicating positive duration dependence. This means that as a credit boom persists, the probability of its end rises.
• Change-Points: The paper finds evidence of a change-point in the duration dependence parameter. Specifically, increasing positive duration dependence is observed in booms lasting less than eight to ten quarters, but this effect diminishes or disappears for longer booms.
• Robustness: The results are robust across different identification criteria (Mendoza-Terrones and Gourinchas-Valdes-Landarretche) and subgroups of countries (industrial vs. developing).
• Systemic Banking Crises: For bad credit booms (those followed by systemic banking crises), the findings also support the presence of a change-point. The likelihood of ending is positively duration dependent for shorter booms, but becomes less predictable for longer ones.
• Country Differences: Credit booms in industrial countries tend to be longer on average than those in developing countries.
• Time Periods: For the MT-criteria, the average duration of credit booms is 2.5 years in the pre-1990 period and 3.75 years in the post-1990 period. For the GVL-criteria, the duration ranges from 2.5 to 3.75 years.

Methodology

• Model Used: A continuous-time Weibull duration model is employed to analyze the likelihood of credit boom episodes ending.
• Data: Quarterly gross capital flows data for 71 countries from 1975q1 to 2010q4.
• Identification Criteria:
  • MT-criteria: Credit boom is identified when the deviation of real credit per capita from its trend exceeds 1.75 times its standard deviation.
  • GVL-criteria: Credit boom is identified when the deviation of the credit-to-GDP ratio from its trend exceeds 1.5 times its standard deviation or when the year-on-year growth in the credit-GDP ratio exceeds 20%.
• Censoring: Observations are censored if the sample period ends before the turning point is observed (i.e., $c_i = 0$), and not censored if the turning point is observed within the sample (i.e., $c_i = 1$).

Main Views

• Predictability: The presence of duration dependence suggests that credit booms are predictable, especially in the early stages. However, as booms grow longer, their predictability decreases.
• Importance of Duration: The duration of a credit boom is a crucial factor in predicting future financial and economic outcomes, including the likelihood of a systemic banking crisis.
• Economic Implications: Longer credit booms are associated with greater financial fragility and a higher probability of ending in a crisis, highlighting the need for early intervention and monitoring.

Conclusion

The study contributes to the understanding of credit boom dynamics by introducing the concept of change-points in duration dependence. It shows that the risk of a credit boom ending increases with its age, but only up to a certain point. Beyond that, the relationship becomes less clear, suggesting that other factors may play a more significant role. The findings also indicate that credit booms in industrial countries are more persistent than those in developing countries. These insights are valuable for policymakers in managing financial stability and economic growth.