Real-time credit gap signals financial imbalances in South Africa

Financial crises often stem from excessive risk-taking during extended growth periods, particularly when regulatory oversight is weak.

Banks tend to expand lending aggressively during economic booms, only to restrict it sharply during downturns, which can exacerbate recessions.

To mitigate this procyclicality, the Basel Committee on Banking Supervision advocates for the credit-to-GDP gap as a crucial early warning indicator.

This metric helps identify escalating credit risks and systemic instability.

When the gap surpasses predefined thresholds, the countercyclical capital buffer (CCyB) can be activated, compelling banks to build up larger capital reserves.

This strengthens the financial system against potential future downturns.

This paper specifically investigates various methods for identifying the South African credit gap in real time, aiming to enhance macroprudential policy effectiveness.

The study constructs a unique real-time dataset to evaluate a range of filtering techniques.

The objective is to effectively isolate the credit gap from the raw credit-to-GDP ratio and pinpoint the real-time estimate that most closely aligns with the ex post full-sample credit gap.

The research finds that the one-sided Hodrick-Prescott filter consistently produces a reliable real-time estimate.

This filter accurately predicted 75% of financial cycle turning points in South Africa since 1980, demonstrating strong consistency with full-sample credit-to-GDP gap estimates.

Consequently, the Hodrick-Prescott filter is deemed suitable as a practical guide for activating the CCyB, offering a robust tool for macroprudential policy.