The fading stock market response to announcements of bank bailouts

https://doi.org/10.1016/j.jfs.2013.01.004Get rights and content

Abstract

We analyze the effects on bank valuation of government policies aimed at shoring up banks’ financial conditions during the 2008–2009 financial crisis. Governments injected into troubled institutions massive amounts of fresh capital and/or guaranteed bank assets and liabilities. We employ event study methodology to estimate the impact of government-intervention announcements on bank valuation. Using traditional approaches, announcements directed at the banking system as a whole were associated with positive cumulative abnormal returns, whereas announcements directed at specific banks with negative ones. Findings are consistent with the hypothesis that individual institutions were reluctant to seek public assistance. However, when we correct standard errors for bank-and-time effects, virtually all announcement impacts vanish in Europe, whereas they weaken in the United States. The policy implication is that the large public commitments were either not credible or deemed inadequate relative to the underlying financial difficulties of banks.

Highlights

► This paper analyzes government announcements of recue plans for banks in the recent crisis. ► Traditional methods show announcements were priced by markets as abnormal returns. ► But these effects disappear with modern estimation methods. ► Conclusion: either announcements not credible or plans inadequate relative to problem.

Introduction

The financial tsunami of the 2008–2009 crisis produced massive expenditure commitments on the part of governments aimed at shoring up their national banking systems. Governments intervened massively and repeatedly to support banks during the crisis. At first, governments reacted to the sharp declines in equity prices with disjointed and ad-hoc interventions. The failure of Lehman on September 15, 2008 was a watershed and prompted policymakers in the next two months to implement programs addressing systemic problems, such as the $700 billion Troubled Asset Relief Program (TARP) in the United States and the £500 billion banking recapitalization program in the United Kingdom. The initial objective of purchasing sub-standard illiquid assets ran into difficulties because, without a market, governments were likely to either overvalue “toxic” assets, thus penalizing taxpayers, or undervaluing them, thus penalizing potential sellers. Governments then adjusted their policy by either recapitalizing financially distressed banks (e.g., in the United States) or nationalizing them (e.g., in the United Kingdom). In December 2008 and January 2009, governments tried to douse the fire of the crisis by targeting specific large banks (e.g., Commerzbank and Citigroup); they were unsuccessful. In February and March 2009, additional general measures were taken, this time with a focus to relieve banks of bad assets. At the same time, many indebted US banks began repaying the US government, while in Europe the number of banks that had signaled their intention for government assistance declined (Wilson and Wu, 2012).

In this paper, we examine the impact of these interventions by measuring the market's reaction to their announcements. Hence, we take the viewpoint of bank shareholders. To do so, we create an original dataset that distinguishes government announcements directed at the banking system as a whole (general announcements) from those directed at specific banks (specific announcements) in the spirit of the distinction made by Carvalho et al. (2010). Then, we apply event-study methodology to estimate the impact of government interventions on bank valuation. The maintained hypothesis is that the announcement of a rescue plan is credible if it affects rates of return of the targeted banks. We test for these effects by computing cumulative abnormal returns (CAR) and abnormal risks of the participating banks around a window that includes announcement dates.

We perform three separate tests on our sample of large banks. The first estimates the overall impact on banks’ equity value of the two types of rescue announcements; the second estimates whether bank size impacts on announcement effects; and the third considers announcements of different types. Our traditional parametric approach shows that general and specific announcements were priced by the markets as CAR and abnormal risks over the selected time windows. In particular, general announcements were associated with positive CAR and decreasing abnormal risks, whereas specific announcements were associated with negative CAR and increasing abnormal risks. However, when we apply more modern techniques to control for auto-correlation and cross-correlation dependence – that is, correcting for both bank and time effects – announcement coefficients lose statistical significance. This reversal is robust to different estimators, traditional as well as modern, and is not driven by sample selection, the length of the event window, or multiple announcements. The findings are consistent either with announcements being not credible or related to rescue programs of inadequate size relative to the underlying problem.

The paper is organized as follows. Section 2 reviews the relevant literature. Section 3 examines event-study methodology and describes our testable equations. Data are presented in Section 4. Sections 5 Traditional approach, 6 Recent approaches employ, respectively, traditional and recent event-study methodology to estimate the impact of government interventions on bank valuation. Section 7 presents findings using a mixed estimation method. Section 8 tests the robustness of results. Conclusions are drawn in the last section.

Section snippets

Recent literature

The recent event-study literature shows that announcements by governments or international institutions tend to have weak or mixed effects on bank valuation. During the Asian crisis of 1997, IMF program announcements increased bank shareholder wealth only marginally, with the exception of South Korea (Kho and Stulz, 2000); East Asian government announcements of debt guarantees, instead, exerted a stronger positive impact on bank stock prices. Klingebiel et al. (2001) argue that these

Methodology

The rescue of several large financial institutions in the United States and in Europe was sparked by the migration of liquidity risk from banks to other financial institutions and followed the rapidly expanding role of government as a market maker of last resort to support, not only big banking, but also big finance. We employ event-study methodology to estimate markets’ reaction to the announcements of government interventions.

Event-study methodology goes back to the 1930s (Dolley, 1933), but

Data

Our data set consists of daily rates of return (inclusive of dividends) of 19 national market indices and of 122 banks listed within these indices over the period from July 31, 2007 to December 31, 2009.1 The listed banks are shown in Table A1 of the Appendix; Bloomberg is the source of the data. July 31, 2007 is the starting point, a pre-crisis date.

Findings using dummies

The first test estimates the overall impact of 48 general and 130 specific announcements on banks’ returns using the entire panel of 115 banks. We test Eq. (1) by aggregating all announcements (ALL). We recall that G and S have seven-day and five-day windows respectively. We experimented with different window lengths (see Section 8). We have added a relative bank size measured by the US dollar capitalization value of bank i relative to capitalization of all banks (SIZEREL). This variable turns

Findings with clustered standard errors

The parametric dummy approach assumes that bank and time effects are time-invariant and common to all banks (Wooldridge, 2007). In the presence of relevant omitted variables, the independence assumption of classic linear regression is violated because the error term becomes correlated with the regressors.10

Mixed approach

Bank and time dummies correct the bias of OLS coefficients; more recent approaches, instead, adjust their SE.15 Since both bank and time effects could have a fixed and a variable component, the natural step would be to avoid simultaneously both the omitted variable bias and residual

Robustness

We check the robustness of our findings with five separate exercises. The first estimates our three equations with bank-and-time dummies and double-clustered SE over a restricted subperiod. The second estimates the same three equations with bank-and-time dummies and Driscoll–Kraay SE for the entire period, but with two different bank subsamples. The reason for switching from double clustering to Driscoll–Kraay has to do with the higher efficiency of the latter when the number of clusters is low

Conclusions

The great financial crisis of 2008–2009 prompted governments to inject vast sums of public funds into banks. Our paper has focused on the specific question of whether general and bank-specific announcements of government rescue plans were priced by the markets as cumulative abnormal returns and abnormal market risk during selected event-time windows. The paper also checks for the presence of too-big-to-fail and too-big-to-save policies. The headline result is that general and specific

Acknowledgements

We thank Matteo Cassiani for providing us with bank data. We are grateful to two referees and the editor of the journal for constructive comments and suggestions.

References (55)

  • A. Almazan et al.

    Why constrain your mutual fund manager?

    Journal of Financial Economics

    (2004)
  • H. Hakenes et al.

    Banks without parachutes: competitive effects of government bail-out policies

    Journal of Financial Stability

    (2010)
  • B.C. Kho et al.

    Banks, the IMF, and the Asian Crisis

    Pacific-Basin Finance Journal

    (2000)
  • P. Veronesi et al.

    Paulson's gift

    Journal of Financial Economics

    (2010)
  • L. Wilson et al.

    Escaping TARP

    Journal of Financial Stability

    (2012)
  • AÏt-Sahalia, Y., Andritzky, J., Jobst, A., Nowak, S., Tamirisa, N., 2010. Market response to policy initiatives during...
  • R. Ball et al.

    An empirical evaluation of accounting income numbers

    Journal of Accounting Research

    (1968)
  • D. Bayazitova et al.

    Assessing TARP

    Review of Financial Studies

    (2012)
  • N. Beck et al.

    What to do (and not to do) with time-series cross-section data

    American Political Science Review

    (1995)
  • Berndt, A., Douglas, R., Duffie, D., Ferguson. M., Schranz, D., 2005. Measuring Default Risk Premia from Default Swap...
  • J.J. Binder

    The event study methodology since 1969

    Review of Quantitative Finance and Accounting

    (1998)
  • Black, L.K., Hazelwood, L.N. The effect of TARP on bank risk-taking. Journal of Financial Stability,...
  • BNP Paribas

    To the Rescue. Report of Market Economics, Interest Rate Strategy, Credit Strategy

    (2009)
  • Cameron, A.C., Gelbach, J.B., Miller, D.L., 2006. Robust Inference with Multi-way Clustering. Working Paper, University...
  • Carvalho, D., Ferreira, M., Matos, P., 2010. Lending Relationships and the Effect of Bank Distress: Evidence from the...
  • R. Chandra et al.

    A synthesis of alternative testing procedures for event studies

    Contemporary Accounting Research

    (1990)
  • CNN Money, online....
  • C.J. Corrado

    Event Studies: A Methodology Review

    (2009)
  • De Jong, F., 2007. Event Studies Methodology. Lecture...
  • Demirguc-Kunt, A., Huizinga, H., 2010. Are Banks Too Big to Fail or Too Big to Save? International Evidence from Equity...
  • DLA Piper, online....
  • J.C. Dolley

    Characteristics and procedure of common stock split-ups

    Harvard Business Review

    (1933)
  • S. Donald et al.

    Inference with difference in differences and other panel data

    Review of Economics and Statistics

    (2007)
  • J. Driscoll et al.

    Consistent covariance matrix estimation with spatially dependent data

    Review of Economics and Statistics

    (1998)
  • Elyasiani, E., Mester, L., Pagano, M., 2011. Large Capital Infusions, Investor Reactions, and the Return and Risk...
  • E. Fama et al.

    The adjustment of stock prices to new information

    International Economic Review

    (1969)
  • E. Fama et al.

    Risk, return and equilibrium: empirical tests

    Journal of Political Economy

    (1973)
  • Cited by (0)

    View full text