VoxEU Column EU policies Financial Markets Global crisis

EU bank deleveraging

Following the financial crisis, European banks have taken steps to revise unsustainable business models by deleveraging. By this metric they have made substantial progress – but this column argues that improper management of the deleveraging process may threaten the recovery. The authors find that equity increases played a much larger role than asset decreases, and recommend increasing the disposal of bad assets.

Ever since the global financial crisis made it apparent that financial institutions had increased their leverage substantially (Figure 1), bank leverage has faced intense scrutiny. In the run-up to the crisis, the ballooning of banks’ balance sheets was primarily driven by both a significant increase in lending activities and an abundance of cheap funding. Many banks expanded dramatically, becoming too highly leveraged and ‘too-big-to-fail’, while at the same time accumulating substantial risks.

In response to these threats, policy initiatives have been taken with the aim of progressively abating bank leverage (BCBS 2014). Besides these regulatory initiatives, the downward adjustment observed over time at the global level has also reflected tighter conditions in debt and interbank markets – as well as the need to revise the unsustainable business models that had characterised a large part of banking activity before the crisis. In this perspective, policy makers have repeatedly expressed support for the idea that the restructuring of balance sheets could help re-establish adequate conditions for the banking business to serve the economy (Draghi 2014).

Nonetheless, proper management of the process is required in order to avoid both a disorderly sell-off of assets and an excessively prolonged adjustment period, which could undermine the economy (Cœuré 2014). Therefore from a policy perspective, it becomes crucial to ascertain how deleveraging is occurring and whether the observed development can be considered as being ‘good’, ‘bad’, or ‘ugly’ for the real economy.

Figure 1. Financial-sector debt burden (percentage of GDP)

Source: Based on ECB data.

The evolution and modalities of EU banks’ deleveraging

In our paper we analyse the main features of the deleveraging process undertaken by a sample of European banks, shedding light on its timing and components (Bologna et al. 2014). We also compare it with the adjustment carried out by the largest US banks, in order to assess its future prospects. The main dataset includes 43 of the 64 banks involved in the Transparency exercise carried out by the European Banking Authority (EBA) in December 2013. The banks selected belong to the following countries: Austria, Belgium, Germany, Denmark, Spain, France, United Kingdom, Italy, the Netherlands, and Sweden. They are the largest in Europe and account on average for about 79% of the banking systems’ assets of the countries considered, and for 94% of the total assets of the original EBA sample.1 Combining different data sources (EBA and SNL Financial), we are able to cover with semiannual data the years from 2007 to 2013, with more details available for the post-crisis period 2011-2013.

Leverage at time t is computed as the ratio of total assets on the bank’s balance sheet (A) to common equity (E). We deem common equity to be a reliable proxy for the actual level of a bank’s net worth because it is not affected by any regulatory definition and/or by any possible difference in its interpretation across jurisdictions.2

From December 2011 to June 2013, the average leverage ratio in our sample fell from 28.6 to 25.0. In general, the banks with a higher initial ratio were those that reduced it most over the period. Aggregating the sample banks by country of residence, we observe that the leverage fell in all cases (Figure 2). Interestingly, in June 2013 the highest leverage was that of German banks (32.2), followed by the French ones (30.2), while the lowest was that of Italian banks (18.0).

Figure 2. Leverage ratio of EU banks by country of residence

Source: Based on EBA data.

To disentangle the components of the observed deleveraging we use the shift-share technique of Dunn (1960), through which we are able to compute how much of a change in leverage is due to a change in assets and/or a change in equity.

In the period December 2011 to June 2013, deleveraging occurred in most cases by acting on both assets and equity, though equity increases played a much larger role. On average, common equity grew by 9.6%, while assets decreased by 4.3%. The scatterplot of the banks’ change in leverage, mapped on its two drivers highlights their relative share in the deleveraging process (Figure 3).3 Negative changes indicate a contribution to the decline of leverage, resulting from a reduction in assets, an increase in equity, or a combination of both. Positive changes instead point to contributions to higher leverage, as the result of either higher assets or lower equity, or both. The dots below the solid diagonal refer to the banks that deleveraged; the few above that line are those that increased their leverage.

Our second step involves the analysis of the types of asset class which contributed most to the observed path. The push to deleveraging from asset reduction reflected a decline of 18.7% in private sector securities and derivatives exposure and a contraction of 3.8% in customer loans (Figure 4). Interestingly, for most banks the reduction in loans to customers involved only performing loans with a very limited contribution from impaired assets, which instead generally increased. By contrast, the exposure to sovereign issuers – which grew by 68.2% – pushed leverage up in most cases.

Figure 3. Components of the changes in leverage by European Banks (December 2011 to June 2013)

Notes: Sector A: equity increase more than offsetting asset increase; Sector B: equity increase and asset reduction (equity increase prevails); Sector C: equity increase and asset reduction (asset reduction prevails); Sector D: asset reduction more than offsetting equity reduction.
Source: Based on EBA data.

Figure 4. Decomposition of European bank leverage dynamics (December 2011 to June 2013)

Source: Based on EBA and SNL financial data.

EU banks’ deleveraging in perspective: Is more deleveraging needed?

It is important – particularly from a policy perspective – to ascertain where European banks stand in their deleveraging process. We therefore compare our sample of European banks with the 10 largest US banks in order to identify possible further deleveraging needs. For both groups the leverage ratio decreased from its December 2008 peak (Figure 5), though the fall was of a smaller magnitude in the case of European banks (from 50.5 to 26.1) compared to the US banks (from 32.8 to 14.1).4 The different scales of deleveraging, together with the current environment in Europe – where the results of the Comprehensive Assessment carried out by the ECB and the national supervisory authorities ahead of the entry into force of the Single Supervisory Mechanism in November 2014 are forthcoming – hint at the possibility that EU banks might need to deleverage more.

The results of a what if analysis suggests that our sample banks should reduce their leverage by an additional 17.2% from the June 2013 level to hit a target ratio of 21.6 – a level consistent with the deleveraging achieved by the largest US banks. The additional deleveraging would translate into a further 17.3% contraction in assets or a 20.9% increase in equity. A more realistic scenario where banks concurrently modify assets and equity in the same proportion observed thus far would require a 6.1% decrease in assets and a 13.5% increase in equity.

Figure 5. EU and US banks’ leverage ratio

Source: Based on SNL financial data.


European banks have significantly reduced their leverage. Overall, deleveraging was more on the ‘good’ than the ‘ugly’ side, as it was largely obtained by raising common equity rather than by reducing assets. Nonetheless, it is possible to see further deleveraging by European banks in order to restore sustainable financial conditions and improve banks’ soundness.

We argue that an effective and viable strategy for achieving a lower level of leverage could entail increasing disposals of bad assets as balance sheet valuations become more aligned to market prices, with severing the bank-sovereign link also part of this strategy. At the same time, capital increases could continue to play a key role, as well as other asset disposals – particularly the non-strategic ones.

Finally, we argue that the maintenance of the current speed of deleveraging would allow European banks to complete their deleveraging by the first half of 2015, a point in time consistent with the period granted to banks to cover any capital shortfall resulting from the Comprehensive Assessment.

Disclaimer: The views expressed are those of the authors and do not represent those of the Bank of Italy.


Basel Committee on Banking Supervision (2011) Basel III: A global regulatory framework for more resilient banks and banking systems − revised version, June

Basel Committee on Banking Supervision (2014) Basel III leverage ratio framework and disclosure requirements, January

Bologna, P, M Caccavaio, A Miglietta (2014) “EU bank deleveraging”, Bank of Italy, Occasional Papers, n. 235

Coeuré, B (2014) “Monetary Policy Transmission and Bank Deleveraging”, speech at “The Future of Banking Summit”, Paris, 13 March 2014

Draghi, M (2014) “Bank restructuring and the economic recovery”, speech at the presentation ceremony of the Schumpeter Award, Oesterreichische Nationalbank,Vienna, 13 March 2014

Dunn, E S Jr (1960) “A statistical and Analytical Technique for Regional Analysis”, Papers and Proceedings of the Regional Science Association, 6:98-112

European Banking Authority (2013), EU-wide Transparency Exercise, December.


1 The remaining 6% is given by intermediaries from Cyprus, Finland, Hungary, Ireland, Luxemburg, Malta, Norway, Poland, Portugal and Slovenia.

2 Our calculation of the leverage ratio is not immediately comparable with that adopted by the BCBS and included in the Basel 3 package. First, the BCBS computes capital to assets, while we do the reverse. Second, the asset metrics, although broadly consistent, exhibit some differences due to a number of valuation adjustments explicitly required by the BCBS and not considered in this study. Third, the measure of capital used by the BCBS is the Tier 1 capital of the BCBS risk-based capital framework (2011).

3 More detailed results, with data presented also at banking name level, are available in Bologna et al. (2014).

4 Due to data availability, we compute leverage as the ratio of total balance sheet assets to tangible common equity, and no longer to the common equity measure available in EBA (2013). Notably, the comparability between the levels of the leverage ratio for US and EU banks remains potentially problematic due to variations in the financial reporting standards used in the two regions. Nonetheless, the leverage ratio dynamics over time should be less exposed to these caveats.

3,465 Reads