MONOPOLY.jpg
VoxEU Column Financial Regulation and Banking Global crisis

Understanding bank payouts during the financial crisis of 2007-2009

Banks' payout decisions at the beginning of the financial crisis of 2007-2009 were particularly controversial as the crisis eroded the capital of many banks. Concerns were raised that banks may have engaged in wealth transfer to shareholders, or that they may have been reluctant to reduce dividends to avoid negative signalling. This column examines these arguments using a large dataset on US bank holding companies. Cross-sectional tests do not provide clear-cut evidence of active wealth transfer. Similarly, the evidence on signalling is mixed.

Banks have long been known for paying high dividends. Their payout decisions in the financial crisis of 2007-2009 received considerable coverage in the press. Acharya et al. (2012) document that, despite mounting losses, many of the largest US bank holding companies kept dividends constant, and in some cases even increased them, until the end of 2008. While this behaviour could have been motivated by the desire to transfer wealth from their creditors (or the government), banks may have been reluctant to cut dividends, fearing that dividend reductions would lead to uncertainty about their fundaments, and cause subsequent refinancing problems.

The debate on bank payouts during the financial crisis of 2007-09

A number of academic studies have furthered our understanding of bank payouts during the crisis. Floyd et al. (2015) compare the payout policy of banks and non-financial firms in the US from 1980 to 2012 and find that dividends play a more important role for banks than for non-financial. Floyd et al. (2015) and Hirtle (2014) document that at the beginning of the crisis adjustments in payouts mainly occurred through reducing share repurchases rather than dividends. Both papers argue that their findings are broadly consistent with the idea that signalling through dividends is important for banks and banks feared adverse reactions to dividend cuts in the crisis. In our recent study (Cziraki et al. 2016), we add to this literature by establishing an empirical model of payouts, and using cross-sectional variation in bank characteristics to evaluate the wealth transfer and the signalling hypotheses.

Wealth transfer or signalling strength?

To understand whether banks’ payout policy at the beginning of the crisis differs from normal times, we need a reference that takes into account changes in banks’ fundamentals. To obtain such a reference point, we estimate a model of payout policy for the period of 1995-2006. The payout policy that the model predicts provides a reference point for what banks and investors could consider normal. A payment in excess of this amount is more difficult to reconcile with the argument that banks feared negative market reactions, and would be more consistent with the hypothesis that banks engaged in wealth transfer.

We find that, compared to the payout pattern in previous periods, the average bank pays out excessively in 2007. This is true for dividends as well as for total payout – the latter defined as the sum of dividends and repurchases. In contrast, in 2008, controlling for fundamentals, dividends per share as well as total payouts do not look significantly different from normal times. However, the average dividend yield and the total payout yield in 2008 are still excessive, suggesting that payouts did not drop as quickly as share prices. 

Having established a measure of abnormal payouts, we examine cross-sectional heterogeneity to understand whether certain groups of banks were more likely to maintain high payouts. Finding that large and worse capitalised banks paid out excessively would be consistent with the wealth transfer hypothesis to the extent that worse capitalised banks arguably have higher incentives to engage in wealth transfer, and larger banks are more likely to be bailed out.  We do not find that banks with low regulatory capital have higher abnormal payouts over the course of the crisis. Instead, they have negative abnormal dividends and total payout per share in 2008, while banks with high regulatory capital have high abnormal payouts in 2007 and 2008. Larger banks have high abnormal dividends, but not high abnormal total payouts.

To see whether banks’ dividend decisions could be explained by the desire to signal future strength, we analyse how dividend changes relate to future performance in the crisis. We find no relation between dividend changes and future performance in 2007. However, banks that reduce dividends in 2008 have lower ROA in 2009. We also look at how the market reacts to dividend changes in normal times and during 2007-2008. We do not find a significant abnormal stock price reaction to dividend cuts until 2009.

Finally, we relate managerial ownership and insider trading to banks’ dividend policy. First, we examine whether banks with a higher level of managerial ownership at the outset of the crisis follow payout policies that benefit shareholders at the expense of creditors. Second, we explore whether insider-trading patterns in banks that are reluctant to cut dividends differ from those in banks that reduce dividends. We find that banks with high levels of managerial ownership have lower abnormal payout measures in the crisis, consistent with the idea that for managers with high ownership stakes preserving capital was more important than extracting payouts. Comparing the insider trading behaviour across banks with different dividend policies (banks increasing, reducing, or keeping dividends constant), we do not find a difference in 2006. However, we detect significant differences for the years 2007 to 2009. Compared to the pre-crisis period, insiders of banks that increase dividends do not change their insider trading behaviour significantly, which does not lend support to the wealth transfer hypothesis. However, insiders of banks that hold dividends constant, and those that decrease dividends, buy significantly more stock compared to the pre-crisis period. These patterns can be consistent with a number of alternative explanations such as contrarian trading (Jenter 2005, Piotroski and Roulstone 2006), or regulatory and public scrutiny (Cziraki 2011).

Conclusion

While banks may have decreased their payouts too little, and too late (as the payout yields suggest), we do not find evidence that banks used their payout policy to engage in an active and deliberate wealth transfer as a response to the crisis, anticipating bankruptcy, or betting on bailouts. While payouts in 2007 appear excessive, our cross-sectional tests suggest that wealth transfer is not the main driver of payouts. Furthermore, insiders of banks that increased dividends did not sell more shares than before, and insiders of banks that kept dividends constant sold significantly less stock. The link between future performance and dividend policy became stronger in 2008, not weaker, in particular for banks with worse future performance. Banks that held dividends constant or increased them had better future performance than banks that reduced dividends. We would not expect the link between future performance and dividend policy to become stronger in the crisis for weaker banks, if wealth transfer considerations are a main motive of dividend policy. According to the wealth transfer story, it is poorly performing banks that have the most to gain by maintaining high payouts.

With regard to the argument of signalling, our evidence is mixed. We find that the correlation between dividend changes and future performance is only significant for dividend changes in 2008, as banks that reduce dividends in 2008 have a lower return on assets in 2009 compared to other banks. This finding can be consistent with signalling in the sense that better performing banks do not cut dividends – as well as prudence, as banks that anticipate worse performance cut back on payouts to preserve liquidity. However, the relationship between changes in payouts and future performance is not significant for other specifications of performance or total payouts. Moreover, we not find a significant adverse market reaction to announcements of dividend cuts before 2009. Thus, while banks may still have feared such a reaction in the crisis, our evidence does not provide a foundation for it.

References

Acharya, V V, I Gujral, N Kukarni, and H S Shin (2012), “Dividends and bank capital in the financial crisis of 2007-2009”, NBER Working paper, no. 16896.

Cziraki, P (2011), “Trading by bank insiders before and during the 2007-2008 financial crisis”, Working paper, available at SSRN 1962719.

Cziraki, P, C Laux, and G Loranth (2016), “Understanding bank payouts during the crisis of 2007-2009”, CEPR Discussion Paper 11453.

Farre-Mensa, J, R Michaely, and M Schmalz (2015), “Financing payouts”, Ross School of Business Paper No 1263.

Floyd, E, N Li, and D J Skinner (2015), “Payout policy through the financial crisis: The growth of repurchases and the resilience of dividends”, Journal of Financial Economics.

Hirtle, B (2014), “Bank holding company dividends and repurchases during the financial crisis”, Federal Reserve Bank of New York, Staff Report No 666.

Jenter, D (2005), “Market timing and managerial portfolio decisions”, Journal of Finance 60, 1903-1949.

Piotroski, J D, and D T Roulstone (2005), “Do Insider Trades Reflect both Contrarian Beliefs and Superior Knowledge about Future Cash Flow Realizations?”, Journal of Accounting and Economics 39, 55-81.

2,310 Reads