Monitoring Systemic Risk Basedon Dynamic Thresholds

IMF Working Papers, June 1, 2012

Source details

Successful implementation of macroprudential policy depends on identifying and estimating systemic risk in real time.
Systemic risk is defined as the conditional probability of a systemic banking crisis.
The conditional probability is modeled in a fixed effect binary response model framework.
The model structure is dynamic and designed for monitoring because systemic risk forecasts depend only on data available in real time.
Several risk factors are identified and the level of systemic risk contains a predictable component that varies through time.
The paper shows how systemic risk forecasts map into crisis signals and how policy thresholds are derived within this framework.
An out-of-sample exercise shows the systemic risk estimates provided reliable early warning signals ahead of the recent financial crisis for several economies.

Modeling approach: fixed effect binary response model framework.
Dynamic design: forecasts rely exclusively on data available in real time (suitable for monitoring).
Risk definition: systemic risk = conditional probability of a systemic banking crisis.
Identified inputs/risk factors: (explicitly listed in the source keywords and subject headings) credit-to-GDP gap, credit-to-GDP growth, equity price growth, banking sector leverage, real effective exchange rates, foreign exchange, and other financial sector indicators.

The level of systemic risk contains a predictable component which varies through time.
Systemic risk forecasts can be converted into crisis signals.
Policy thresholds for action can be derived within the presented framework.
Out-of-sample results: systemic risk estimates provided reliable early warning signals ahead of the recent financial crisis for several economies.

Macroprudential policy implementation should be supported by real-time systemic risk monitoring based on the proposed dynamic framework.
Derived crisis signals and policy thresholds offer a structured way to translate risk forecasts into policy action.
Identified risk factors (e.g., credit-to-GDP gap, credit growth, equity price growth, banking sector leverage, real effective exchange rates) can guide indicator selection for surveillance and policy calibration.