Fiscal risks and their impact on banks’ capital buffers in South Africa

South African Reserve Bank

Working Paper Series

WP/21/08

Fiscal risks and their impact on banks' capital buffers in South Africa

Konstantin Makrelov, Neryvia Pillay and Bojosi Morule

Authorised for distribution by Christopher Loewald

14 May 2021

© South African Reserve Bank

All rights reserved. No part of this publication may be reproduced, stored in a retrieval system, or transmitted in any form or by any means without fully acknowledging the author(s) and this Working Paper as the source.

South African Reserve Bank Working Papers are written by staff members of the South African Reserve Bank and, on occasion, by consultants under the auspices of the South African Reserve Bank. The papers deal with topical issues and describe preliminary research findings, and develop new analytical or empirical approaches in their analyses. They are solely intended to elicit comments and stimulate debate.

The views expressed in this Working Paper are those of the author(s) and do not necessarily represent those of the South African Reserve Bank or South African Reserve Bank policy. While every precaution is taken to ensure the accuracy of information, the South African Reserve Bank shall not be liable to any person for inaccurate information, omissions or opinions contained herein.

South African Reserve Bank Working Papers are externally refereed.

Information on South African Reserve Bank Working Papers can be found at https://www.resbank.co.za/en/home/publications/Papers/working-papers

Enquiries relating to the Working Paper Series can be addressed to:

Head: Economic Research Department

South African Reserve Bank

P O Box 427

Pretoria 0001

Tel. [phone]

Fiscal risks and their impact on banks' capital buffers in South

Africa

Konstantin Makrelov, Neryvia Pillay and Bojosi Morule*

14 May 2021

Abstract

South Africa's fiscal balances have deteriorated significantly over the last decade, while the economy has been recording disappointing economic growth rates even prior to the COVID-19 crisis. In this paper, we estimate a series of equations using the Arellano and Bond (1991) estimator to test how sovereign risk premia affect capital buffers, while controlling for variables identified in the literature, such as size of banks, the economic cycle, competition and equity prices. Unlike other studies, we use actual capital buffers provided by the South African Prudential Authority. We show that these are substantively different to the proxy buffers calculated using the common approach in the literature, indicating that results based on proxy measures should be interpreted with caution. Our overall results show a positive relationship between the sovereign risk premium and capital buffers, and the results are robust across different specifications. This suggests that banks are accumulating capital to mitigate against fiscal and other domestic policy risks, and the related financial stability issues. It is likely that this is contributing to higher lending rates.

JEL classification: C23, E62, H32, G28

Keywords: Fiscal policy, capital buffers, financial regulation, sovereign-bank nexus, South Africa

• Konstantin Makrelov is a lead economist at the South African Reserve Bank (SARB), Neryvia Pillay-Bell is a Research Fellow at the SARB and at Economic Research Southern Africa, and Bojosi Morule is an economist at the SARB.

1

1. Introduction1

South Africa's fiscal balances have deteriorated significantly over the last decade, while the economy has been recording disappointing economic growth rates even prior to the COVID-19 crisis (Burger and Calitz 2020; Loewald, Faulkner and Makrelov 2020). South Africa's credit rating has been reduced to 'junk status', while sovereign risk premia have increased significantly, driven by global risk aversion and deteriorating domestic fiscal and economic conditions.

In this paper, we assess how the deteriorating fiscal conditions are affecting banks' capital buffers in South Africa. Banks and the sovereign are closely interlinked through several channels. For example, fiscal dynamics are an important determinant of bank equity and solvency value as they depend on changes in the perceived solvency and market value of government debt (Dell'Ariccia et al. 2018). A drop in bond prices also reduces the market value of banks' bond holdings, eroding their collateral value.2 Government plays an important role as a backstop in the event of financial sector distress. Unsustainable fiscal balances reduce the ability of government to play this role and increase risks of bank runs and contagion if the banking sector faces solvency challenges (Kallestrup, Lando and Murgoci 2016). Higher fiscal risks can also increase the financial sector's funding costs and risk aversion, causing banks to hold more capital than required by regulators (Borio and Zhu 2012). Increasing capital buffers is costly for banks and can increase the cost of lending and reduce economic activity.3

In South Africa, fiscal risks have increased, while banks and the financial sector in general have increased their holding of government debt. Banks, pension funds and insurance companies held government debt of just over 40% of GDP in 2020/21, up

1

2

3

We would like to thank Hugh Campbell, Angeline Phahlamohlaka and Jaco Vermeulen for their help with our numerous Basel III queries. We also want to thank Vafa Anvari, an anonymous referee and participants in the SARB seminar series for their useful comments and suggestions.

Bonds are used for example as collateral in some transactions with the Reserve Bank. See for example Woodford (2010).

2

from 20% in 2012/13.4 At the same time, the government debt-to-GDP ratio deteriorated from 41% to 80.3%, and sovereign risk premia have increased steadily by over 200 basis points, indicating rising fiscal risks.

We estimated a series of equations using the Arellano and Bond (1991) estimator, controlling for variables such as size of banks, the economic cycle, changes in equity prices, and - most importantly - sovereign risk premia. The results are robust across different specifications and they show a strong relationship between the sovereign risk premium and the banks' capital buffers. For every one percentage point increase in the sovereign risk premium, the capital buffers increase by 1.7 to 1.9 percentage points, on average, across the banks.

Our contribution to the literature is threefold. Firstly, we provide the first estimates of how rising fiscal risks are affecting capital buffers in South Africa. Previous studies on other countries have relied on proxy measures for banks' capital buffers to study their determinants. In our analysis, we use the actual capital buffers provided to us by the South African Prudential Authority. Secondly, we highlight significant differences between the actual capital buffers and the proxy measures used in the literature. We show how the proxy measures are not a good indicator of actual buffers. Thirdly, we illustrate how potential growth revisions affect estimates of the cyclicality of capital buffers.

The results suggest that fiscal policy decisions need to take into account the adverse impacts of negative fiscal shocks on the financial sector, as these contribute to smaller fiscal multipliers. Macro- and microprudential regulators need to review the role of government debt as a low-risk asset in the regulatory framework and identify interventions to reduce bank exposure to sovereign debt instruments as fiscal risks increase. This will also support more prudent fiscal policy and enhance macroeconomic stability.

4 See the second edition of the 2020 Financial Stability Review available at https://www.resbank.co.za/en/home/publications/publication-detail-pages/reviews/finstab-review/2020/Second_edition_Financial_Stability_Review.

3