Giovanni Dell’Ariccia, Deniz Igan, Paolo Mauro, Hala Moussawi, Alexander F. Tieman, Aleksandra Zdzienicka 04 March 2020
During the Global Crisis, governments rescued banks with capital injections, asset purchases, and guarantees. Until now, we have had no clear idea what happened to that taxpayer money. This column uses bank-level data to compile a comprehensive accounting of the costs of, and returns on, these interventions. While initial public support cost $1.6 trillion, the total fiscal impact has been $250 billion – on average less than 1% of GDP.
Public interventions in financial institutions were a hallmark of the response to the global financial crisis that began in 2007–08. Governments poured vast sums of money into their banking systems – through capital injections, asset purchases, and guarantees – to prevent further financial turmoil and shield economies from even sharper recessions. The interventions may have been necessary, but they were often unpopular. Public frustration emerged against the use of taxpayers’ money to rescue the financial institutions that many saw as the culprits of the crisis. This sentiment shaped some aspects of post-crisis financial reform: the Dodd-Frank Act aimed “to protect the American taxpayer by ending bailouts”1 while, in the EU, the Bank Recovery and Resolution Directive sought “to obviate the need for [using taxpayers’ money to save financial institutions] to the greatest extent possible”.2
More than a decade later, what has happened to that taxpayer money? What banking assets did governments purchase? What cash flows, if any, did they receive on these assets? Did they sell them back to the private sector? What assets remain on governments’ books, and what is their value today?
And, taking all this into account, what have the bank rescues cost in each country?
Answering these questions, even imperfectly, requires painstaking work. Not only because there is no definitive measure of the cost of a banking rescue (Harbert 2019, Lucas 2019), but also because we lack data that track government stakes in banks beyond the initial intervention. Previous work on systemic banking crises, notably by Laeven and Valencia (2018), has been primarily based on aggregate country-level statistics.
In a recent paper (Igan et al. 2019), we draw on a wide range of official reports on crisis response, court rulings, public letters between national agencies, and bank annual reports to piece together a dataset tracking 1,114 financial institutions in 37 countries – covering the epicentres of the crisis and representing 62% of global GDP – over 2007–17. The benefits of our bank-level database lie in its granularity. We compile data at the level of specific transactions in individual banks,3 which allows us to gauge the association between bank characteristics, the amount of support they received, and the way in which support was provided. We can also see the implications for bank-level outcomes – something that aggregate country-level data cannot do. Our database also allows us to track the evolution of the assets acquired during the global crisis, and to focus on those that remain in public hands. We can therefore compare their value to the cost of intervention.
There is still a sizeable public stake in banks
Direct public support to financial institutions adds up to $1.6 trillion for the countries in our sample. Governments recovered more than half of that amount over the next decade – $900 billion – by selling their stakes in those banks back to the private sector. This leave a net direct fiscal cost of intervention of $700 billion.
That is not the end of the story, however. At the end of 2017, governments still owned banking assets valued at $350 billion. The unwinding of direct support has been uneven, with only a few countries fully divesting the government stake in banks (Figure 1). Public equity holdings remain large in Luxembourg, Portugal, Greece, and Belgium. Public holdings of impaired assets are still substantial in Austria, Slovenia, and Germany.
Figure 1 Public asset holdings as a percent of GDP, 2017
Income received from government stakes in banks was higher – by almost $100 billion – than interest payments on the money governments borrowed to finance their acquisition. Taking both the value of the remaining assets and these cash flows into account, the total fiscal impact of interventions is $250 billion, or less than 1% of GDP.
This is an average, however, and the results vary widely across countries. The total impact is near 20% of GDP in Greece and Cyprus, 12% in Slovenia, and 9% in Portugal. Other countries saw total costs of 5% of GDP or less. Eleven countries had total costs below 1% of GDP – or even small gains.
The long shadow of the Global Crisis
Why are governments still holding the banking assets they acquired to stabilise the markets? We find that governments divested more of their bank asset holdings in countries in which the economy recovered faster, and in which the bank in question was better capitalised and more profitable before the crisis. We also find that in countries in which the government stake remained high, private investment and credit growth were slower, financial access, depth, and efficiency were worse, and financial stability improved less.
These bank interventions have cast a long shadow, and so future work could explore the interaction between bank size and government interventions, the macro-financial environment’s effect on recovery and divestment rates, the factors underlying the choice of deploying different instruments in asset purchases, and the long-term consequences of government interventions in the financial sector including on growth, stability, and market structure.
Our dataset exclusively focuses on direct government interventions and the stake governments have taken in banks as a result. In some cases, the size of government interventions was reduced by means of private sector burden sharing or bail-in (Portugal and Slovenia are examples, see Dell’Ariccia et al. 2018). The fiscal cost of public interventions would have been even larger in the absence of bail-ins of equity and debt-holders (for example, by converting to equity or writing off debt-holders). This is particularly relevant for future crises, given that the reformed resolution frameworks would resort more to such procedures to resolve distressed banks. Further analysis using this new dataset would allow for an evaluation of what worked and at what cost, thus informing future policies.
Equally importantly, our data contribute to greater transparency on the use of taxpayers’ money to support the financial sector. Data on government support to the financial sector and the associated fiscal implications are not always easily accessible and are often based on different methodologies across countries. In some cases, they remain confidential, even a decade after the crisis. We hope our paper will advance the discussion on data availability, transparency, and accountability.
Dell’Ariccia, G, M S Martinez Peria, E A Awadzi, M Dobler, D Igan, and D Sandri (2018), “Trade-offs in Bank Resolution”, IMF staff discussion note 18/02.
European Central Bank (2016), “Compilation guide on government assistance to the financial sector”, November.
European Commission (2018), “Eurostat supplementary table for reporting government interventions to support financial institutions”, background note, October.
Harbert, T (2019), “Here’s how much the 2008 bailouts really cost”, Ideas Made to Matter, MIT Sloan, 21 February.
Igan, D, H Moussawi, A Tieman, A Zdzienicka, G Dell’Ariccia, and P Mauro (2019), “The Long Shadow of the Global Financial Crisis: Public Interventions in the Financial Sector“, IMF working paper 19/164.
Laeven, L and F Valencia (2018), “Systemic banking crises revisited”, IMF working paper 18/206.
Lucas, D (2019), “Measuring the Cost of Bailouts”, Annual Review of Financial Economics 11(1): 85–108.
3 We supplemented the bank-level data with updated information on government acquisition of impaired assets and the financial costs and benefits from government asset holdings. Our data combines stock and flow data, similar to the approach adopted by EU countries in the Excessive Deficit Procedure Supplementary Tables and Financial Assistance Measures Tables (EC 2018, ECB 2016).