Financial derivatives and bank risk: evidence from eighteen developed markets

Abstract

We examine the relationship between equity risk and the use of financial derivatives with a sample of 555 banks from eighteen developed markets from 2006 to 2015. Our main findings suggest that banks’ use of financial derivatives increased their risk. This increase in risk can be driven by banks’ use of derivatives for speculative purposes, by suboptimal hedging to obtain hedge accounting status, or from accounting mismatches that generate volatility in earnings. We also show that this relationship is nonlinear. Too-Big-To-Fail banks and those that employ a traditional retail banking business model are subject to lower idiosyncratic risk. We address endogeneity concerns using instrumental variables capturing the use of derivatives with portfolio ranking. Overall, our study contributes to understanding the impact of derivatives use on bank risk and the risk consequences of a bank’s business model choice.

Keywords:

• derivative accounting
• fair value
• financial derivatives
• risk management

Acknowledgement

We thank Abhinav Anand, Wolfgang Bessler, Justin Chircop, Thomas Conlon, John Cotter, Rong Ding, Minyue Dong, Michele Fabrizi, Jo Horton, Elisabetta Ipino, Anastasia Kopita, David Marginson, William Megginson, Giovanna Michelon, Yuval Millo, Gary J. Previts, Amedeo Pugliese, Richard Taffler, Georgios Voulgaris, Eamonn Walsh, two anonymous referees, and the editors for their helpful suggestions. This paper has also benefited from comments made by participants at the XII Workshop on Empirical Research in Financial Accounting (University of Exeter), and seminars in Ca’ Foscari University of Venice, Paris Dauphine University, University College Dublin, and University of Padova. Xing Huan acknowledges the support of the Irish Research Council for funding received under grant number REPRO/2015/109.

Disclosure statement

No potential conflict of interest was reported by the authors.

ORCID

Xing Huan http://orcid.org/0000-0003-1845-6367

Notes

1 For example, the trader known as the ‘London Whale’ within JP Morgan Chase gambled heavily on an obscure corner of the credit default swap (CDS) market and lost $6.2 billion (Financial Conduct Authority, 2013). Banks have also been fined for manipulating benchmark rates – including LIBOR, FOREX, and Isdafix, which determine the payout of derivatives – to benefit their own trading positions (Financial Conduct Authority, 2016).

2 The Basel Committee (2010, 2011) pointed out that one of the main reasons the financial crisis became so severe was that the banking sector of many countries had built up excessive on- and off-balance sheet leverage.

3 For instance, Altunbas et al. (2011) find that a strong deposit ratio and greater income diversification improve bank resilience. Ayadi et al. (2013) document that retail-oriented banks are less likely to default. Demirgüç-Kunt and Huizinga (2010) suggest that banking strategies that rely predominantly on attracting non-deposit funding or generating noninterest income are very risky. Mergaerts and Vander Vennet (2016) show that a strong reliance on retail banking activities is associated with higher profitability and stability.

4 While it was the regulation in place during the period under study, IAS 39 has been replaced by IFRS 9 since 1 January 2018. Key changes have been made to the following areas: classification and measurement of financial assets, classification and measurement of financial liabilities, impairment, and hedge accounting. See ‘IFRS 9: Financial Instruments – High Level Summary’ (Deloitte, 2016) for a detailed discussion.

5 TBTF banks refer to the global systemically important banks (G-SIBs) identified by the Financial Stability Board (in consultation with Basel Committee on Banking Supervision and national authorities). G-SIBs are subject to higher capital buffer requirements, total loss-absorbing capacity requirements, resolvability requirements, and higher supervisory expectations (Financial Stability Board, 2016).

6 The only exception is New Zealand. See Demirgüç-Kunt et al. (2014) for a complete list of countries with explicit deposit insurance schemes as of the end of 2013.

7 The charter value can be derived from factors related to entry into the industry and/or access to protected markets.

8 Level I inputs are quoted prices in active markets for identical assets or liabilities that the entity can access on the measurement date [IFRS 13:76]. Level II inputs are inputs other than the quoted market prices included in Level I that are observable for the asset or liability, either directly or indirectly [IFRS 13:81]. Level III inputs are unobservable inputs for the asset or liability [IFRS 13:86].

9 Both loans and deposits are scaled by bank total assets. The sample medians are 0.820 and 0.649 for deposits-to-total-assets and loans-to-total-assets, respectively.

10 We also perform Durbin-Wu-Hausman (Durbin, 1954; Wu, 1973; Hausman, 1978) test to compare the IV and OLS estimates. We reject the null hypothesis that both the IV and the OLS estimates are consistent at the 5% level, suggesting that OLS estimates are biased and that IV should be considered.

11 Combining the accounting data from Bankscope, stock data from CRSP, and data on holdings of the five types of derivatives obtained from the Federal Reserve Bank, we obtain a sample of 293 US banks (1,875 observations).

Additional information

Funding

This work was supported by Irish Research Council: [grant number REPRO/2015/109].