Save in Good Times, Rescue in Bad Times: Why Capital Buffers Pay Off?

The economy is often compared to the sea: sometimes it is calm, while at other times the waves grow higher and higher until they turn into a storm. The banking sector can likewise be seen as a passenger ship whose task is to carry its passengers safely to their destination, in other words, to provide loans to households and businesses. In calm waters, it may be tempting to take on as many passengers as possible and not worry too much about life jackets. After all, life jackets have to be purchased, maintained, and stored, which involves costs and leaves a little less room for passengers. For this very reason, banks are required to build up additional capital buffers during periods of economic expansion. As long as the weather remains fair, such requirements may seem unnecessary. But when a storm arrives, life jackets become indispensable. In the same way, capital buffers accumulated during good times can be used in times of crisis, allowing banks to continue carrying passengers, that is, financing the economy, even under difficult conditions, rather than cutting back on lending precisely when it is needed most.

We have previously discussed how optimal capital buffers can support long-term economic growth. In this article, we focus on the circumstances under which releasable capital buffers are built up and subsequently released.

Recognising the role of capital buffers across different phases of the financial cycle, this study focuses on three key stages – the standard risk phase, the risk build-up phase, and the crisis and downward phase. In the context of this assessment, it is assumed that the standard risk phase is characterised by the average rate of change in GDP and a moderate level of systemic risks and financial stress [1]. The risk build-up phase is characterised by a higher rate of change in GDP, increasing systemic risks, and low financial stress [2]. The crisis and downward phase, in turn, is characterised by decreasing GDP, the materialisation of risks, and high financial stress [3].

The figure shows that decisions to increase capital buffer requirements are typically made during the standard risk phase or the risk build-up phase. In much the same way, life jackets are procured when the sea is still calm or when storm clouds are only beginning to gather on the horizon. By contrast, when the economy enters a downturn or a crisis, capital buffer requirements are released in full or in part to help prevent a contraction in lending. Returning to our analogy, a well-equipped ship can continue sailing even through rough seas, while the risk of passengers drowning is significantly reduced.

Raising capital buffers during the standard risk phase is expected to have a modest negative impact on median future GDP growth, that is, on the rate of change in GDP under normal conditions. In the short run, it may also weigh on Growth-at-Risk (GaR), the rate of change in GDP under adverse conditions. By contrast, when capital buffers are increased during the risk build-up phase, their effect on the rate of change in GDP is expected to turn positive after a few years. Similarly, releasing capital buffers during a crisis does not have an immediate positive effect, but helps the economy over the medium term by softening the downturn and/or supporting a stronger recovery. Overall, the cycle of building up capital buffers in good times and subsequently releasing them in bad times has a positive effect on GDP growth.

The results indicate that during a crisis, the long-term benefits from releasing capital buffers are higher than the short-term costs incurred by increasing the capital requirements during periods of standard risk (Table) [4].

• If the capital requirements are increased during the standard risk phase in the short term [5], the median rate of change in GDP can decrease by 14–25 basis points, while during sudden shocks (as characterised by GaR), the annual rate of change in GDP would be lower by 41–113 basis points.
• If the capital requirements are increased during the risk build-up phase in the medium term,  an overall neutral or positive impact on GDP dynamics is expected. The average rate of change in GDP could improve to +122 basis points, while during sudden shocks – to +178 basis points. The positive impact can be explained by banks' higher risk absorption capacity, reduced risk accumulation, and a smaller probability of a crisis.
• When a crisis occurs and the financial cycle enters a downward phase, in which capital buffers are released, the assessed long-term benefits are significant: the average rate of change in GDP is 52–94 basis points higher, while during more severe shocks, it could increase by 60–136 basis points.

The findings regarding the long-term benefits of raising capital requirements in Latvia are consistent with the assessments of other studies. The studies by other authors also confirm the usefulness of releasable capital buffers during a crisis. Herrera et al. (2024)  have assessed that the total benefit of releasing the CCyB is more than twice as high as the costs related to increasing the requirement. Similarly, the study by Estrada et al. (2024)  has confirmed that raising CCyB rates during periods of economic growth results in a comparatively smaller drop in GDP growth, while their release during downward periods yields benefits that are 3–4 times higher, particularly at the lower end of the distribution of GDP growth rates.

Technical details

The GaR panel approach was used to assess the long-term effect of changes in capital requirements – it allows for the analysis of these effects during various phases of the financial cycle. This approach was selected in accordance with the basic principles guiding the CCyB rate. The CCyB rate is initially increased in the early stages of the cycle, when the cyclical systemic risk is not elevated, but is not significantly low either. In the risk build-up phase, when the cyclical risks accumulate, the CCyB rate is increased above the CCyB base rate of 1% depending on the risk intensity. While in the crisis and downward phase of the financial cycle, the CCyB rate can be reduced or fully released.

The GaR model was used in the studies by Estrada et al. (2024) [6] and Galan and Rodríguez-Moreno (2020) . The following equation was used for the assessment:
     

The authors would like to thank Jorge Galán, economist at Banco de España, for his valuable advice on methodological issues.

[1] The medians of the rate of change in GDP, CCI, and CLIFS.

[2] The 95th percentile of the rate of change in GDP and CCI, and the 5th percentile of CLIFS.

[3] The 5th percentile of the rate of change in GDP and CCI, and the 95th percentile of CLIFS.

[4] After assessing the impact of implementing the CCyB rate in Latvia, it was concluded that in 2025–2028 the average impact on the median rate of change in GDP is positive, but not statistically significant.

[5] Up to two years.

[6] Estrada, A. et al. (2024), Analysis of cyclical systemic risks in Spain and of their mitigation through countercyclical bank capital requirements, Documentos Ocasionales, N.º 2414, Banco de España.