Rebalancing in Spain’s private sector is under way, but with more modest progress on reducing stocks. Spain is subject to significant spending pressures, reflecting unfavorable demographic trends and subdued growth prospects, and will require substantial structural reform. Priority reforms are needed to strengthen its fiscal framework. The study tries to infer the potential impact of the ongoing integration process on bank efficiency based on preconsolidation bank data. The reforms, such as those recently implemented, of employment protection and collective bargaining could help improve Spain’s inflation performance.

Abstract

Rebalancing in Spain’s private sector is under way, but with more modest progress on reducing stocks. Spain is subject to significant spending pressures, reflecting unfavorable demographic trends and subdued growth prospects, and will require substantial structural reform. Priority reforms are needed to strengthen its fiscal framework. The study tries to infer the potential impact of the ongoing integration process on bank efficiency based on preconsolidation bank data. The reforms, such as those recently implemented, of employment protection and collective bargaining could help improve Spain’s inflation performance.

IV. Will the Savings Bank Mergers Increase Efficiency? A Non-Parametric Analysis1

A significant consolidation and restructuring process of the Spanish savings bank sector is underway. Although this process is still ongoing, a non-parametric Data Envelopment Analysis is used to analyze whether the new configuration of the sector can be expected to improve the efficiency of the banking sector as a whole. The study tries to infer the potential impact of the ongoing integration process on bank efficiency based on pre-consolidation bank data. Since the present analysis constitutes only a partial assessment of the current, more complex and far-reaching reorganization of the savings bank sector, the results ought to be considered with caution. They can be considered as a benchmark case that, compared with the future observed efficiency frontier based upon actual data, will inform on the sources of efficiency changes. The results suggest that while the bulk of the mergers can be expected, ex ante, to produce significant efficiency gains, some mergers among small institutions do not seem best configured to deliver significant efficiency gains. This underscores the need, as planned, for substantial restructuring, reorganizing, and downsizing that could also prompt a second round of integration.

A. Introduction

1. The purpose of this note is to provide an overview of the ongoing consolidation process that has fundamentally reshaped the savings bank sector in Spain. Through a number of mergers and joint-ventures—the so-called Institutional Protection Schemes (Sistemas Institucionales de Protección, SIP)—the number of institutions has been reduced from 45 to 18 and their legal status transformed.2 This note uses a non-parametric analysis to assess whether the newly created institutions could potentially enhance the efficiency of the Spanish banking sector. Since the present analysis constitutes only a partial assessment of the current, more complex and far-reaching reorganization of the savings bank sector, the results ought to be considered with caution. They can be considered as a benchmark case that, compared with the future observed efficiency frontier based upon actual data, will inform on the sources of efficiency changes.

2. The note is organized as follows: section B provides a brief overview of the reform process that is re-shaping the Spanish savings bank sector; section B explains the methodology used to estimate the potential efficiency gains and the main results of the analysis; section C draws some conclusions.

B. Background

3. By the end of 1980s, all the institutional barriers, including geographic constraints, limiting the business activities of savings banks were lifted. Over time, savings banks gradually expanded beyond their “home” regions, broadened their range of activities, built extensive branch networks, and increased their staff, thereby becoming solid competitors of commercial banks (Figure 1).3

Figure 1.
Figure 1.

Spain: Commercial and Savings Banks Indicators

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A004

Sources: Banco de España; and IMF staff estimates.

4. The other side of the coin has been the build-up of excess capacity in the system. As of end-2009, there was almost 1 branch every 1,000 inhabitants in Spain, almost twice the density of the euro-area average (Figure 2). The extreme capillary of the branch network is confirmed by the low number of employees per branch compared with other European banking systems. Spanish savings banks, in particular, do not compare favorably in terms of assets-per-employee with euro-area average. Roughly speaking, reaching a dimension broadly in line with the average banking sector in the euro-area would imply the need for Spanish savings banks to reduce their staff by almost 30 percent (more than 37,000 positions) and halve the number of branches. It is worth noting that progress is already being made, with the number of branches down 14 percent since the peak in 2008 and employees down 7 percent for the system as a whole.

Figure 2.
Figure 2.

Banking Sector’s Selected Structural Indicators

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A004

Sources: European Central Bank; Banco de España; and IMF staff estimates.

5. With the financial crisis, the business model of the Spanish banks came under pressure, which was particularly acute in the case of savings banks. On the asset side, banks were hit by the collapsing real estate sector, to which they have been traditionally exposed. On the liability side, wholesale markets, which had become a primary source of funding, dried up. The prospect of more demanding Basel III capital requirements put additional pressure especially on the savings bank sector. Strong headwinds, reflecting the weak operating environment and the increase in non-performing assets, were expected to significantly reduce the internal generation capacity of many savings banks. And savings banks’ particular ownership structure severely limited their capacity to tap financial markets to bolster capital levels.4

6. The restructuring of the sector thus became urgent and it occurred relatively rapidly in three phases.

  • The first phase started with the creation of the Fund for Orderly Bank Restructuring (Fondo de Reestructuración Ordenada Bancaria - FROB) in June 2009. The main purpose of this fund is to assist and foster the reorganization of the Spanish banking industry as well as to provide a rapid and effective solution for ailing institutions. In May and June 2010, the Bank of Spain (BdE) approved seven mergers or acquisitions and five SIPs, some of which requested financial support from the FROB (Table 1). And in July 2010, the legal and regulatory framework of savings banks was fundamentally reformed. In particular, the new law gives savings banks a menu of options: (1) to maintain their existing structure but removing a number of legal impediments to the issuance of equity-like instruments (cuotas participativas); (2) to operate through a bank; (3) to become part of a SIP; or (4) to change their legal nature and become a foundation and a (potentially minority) shareholder of the bank to which it transfers its business. The corporate governance structures has been also enhanced by: (1) reducing the maximum voting rights for public entities’ from 50 percent to 40 percent; (2) prohibiting elected officials to be members of governing bodies; and (3) strengthening reputation and experience criteria for members of the Boards.5

  • The second phase was marked by the Irish crisis (November/December 2010) that brought the Spanish banking sector under the market spotlight again. To signal the commitment to a fully-fledged restructuring, the five SIPs, decided to increase their mutual support to 100 percent of capital and liquidity compared to a legal minimum of 40 percent.

  • In the third phase, to further strengthen market confidence on the Spanish banking system, in February 2011, the government adopted a series of measures including strengthening the level and quality of minimum capital requirements. In addition to the usual Tier 1 and capital adequacy ratio, a new solvency ratio was added, a “principal capital” ratio, where the definition of “principal capital” is an approximation of common equity Tier 1 under Basel III. The base requirement of 8 percent is raised to 10 percent for those credit institutions (primarily savings banks) that heavily rely upon wholesale funding (more than 20 percent) and have not placed a significant share of their capital (20 percent) with third parties. Reflecting the more demanding capital requirements and the new rules governing the FROB, the SIPs as well as some savings banks decided to spin-off their banking business to newly created commercial banks and started the process of listing these new entities on the stock exchange. 6 7

Table 1.

Spain: Deposit Insurance and FROB Support

article image
Sources: Banco de España; Confederación Española de Cajas de Ahorros; and IMF staff estimates.

Based on market share, defined in terms of Spanish credit institutions’ total assets (as of end-December 2009).

In case of successful IPO (at least 20 percent of its share capital is floated), the minimum additional amount needed to reach 8 percent core capital would be at least €1,795. Subsequently, BFA-Bankia reconfigured their corporate structure and FROB 1 is only a liability of BFA.

IPO resolution passed; it must place at least 20 percent of its share capital.

C. Assessing the Mergers

7. Mergers can be a powerful tool to restructure ailing institutions. They aim at achieving three main objectives:8

  • Rationalize the “production” (cost-saving);

  • Exploit economies of scale and scope;

  • Reduce internal inefficiencies (so called “X-inefficiency”).

8. While mergers tend also to be associated with lower competition, reduced credit availability and higher prices, this is less of a concern in the spanish context given the substantial overcapacity of the sector.9 Consolidation is also critical to create more robust financial institutions that can compete in a tougher environment (also from a regulatory point of view). On the other hand, consolidation may exacerbate moral hazard in the system by generating larger and more complex financial institutions, which in turn intensify systemic risk.

9. The question in the Spanish context is more whether the envisaged mergers can be expected to enhance efficiency. There are some negative a priori considerations, especially that all savings banks have broadly the same business model; therefore, in principle, opportunities for economies of scope are rather limited. Economies of scale could be still at play, although the objective is to downsize the sector and to rationalize costs.

10. Although the consolidation process is still ongoing, a simple analysis can be performed to investigate whether the new configuration of the savings bank sector could represent a potential improvement in the efficiency of the sector. To this end, a non-parametric Data Envelopment Analysis (DEA) is used to determine the efficiency of Spanish savings and commercial banks before the starting of the restructuring process. Then, based on these ex-ante results, savings banks have been “virtually” merged following the actual grouping of institutions to evaluate potential changes in efficiency.

The DEA methodology

11. DEA is a non-parametric linear programming methodology used to measure best practice technology and relative technical efficiency of decision making units (DMUs), in this case banks, using the same inputs and outputs (see Appendix). In this context, DEA can determine the set of banks that make up the technically efficient production frontier and others which lie within interior, inefficient points below the frontier. To identify the efficient frontier, an input- or an output-oriented model can be used: in the former, inputs are minimized while satisfying at least the given output levels; in the latter, output is maximized without requiring more of any of the observed input values. Each DMU will be associated with an “efficiency score” that ranges between 0 (inefficient) and 1 (efficient).

12. The main advantage of DEA is that, unlike typical regression analysis, no a priori model specification is required. Instead, DEA constructs a non-parametric envelopment frontier over the sample data such that observed points lie on or below the “efficient” production frontier. However, as DEA looks at relative efficiency within a particular sample of DMUs, the results cannot say anything about the absolute efficiency of Spanish banks vis-á-vis other countries’ banking sector. It also does not allow for random errors.

Empirical analysis

13. To model bank behavior, two approaches are usually considered: the production and the intermediation approach.10 In the former, banks are regarded as using labor and capital to generate deposits and loans. In the latter, banks are regarded as intermediaries in raising funds (deposits and other funds) and lending those funds in the form of loans or other investment to generate earnings.

14. This paper follows the intermediation approach to define input and output variables. Specifically, following Avikran (2006) and Banker, Chang and Lee (2010), inputs are represented by interest expenses and non-interest expenses, while outputs are represented by interest and non-interest income.11 Since the main objective of the ongoing restructuring process is to reduce operating costs and downsize the sector, the study assumes an input-oriented model.12

15. The sample of credit institutions comprises 43 savings banks (the two institutions that were intervened—CCM in 2009 and Cajasur in 2010—were excluded)—and 7 commercial banks. Both the input and the output variables were averaged over the 2008-09 period.

16. The results are reported in Table 2 (Figure 3) in which savings banks have been already grouped according to actual mergers and SIPs.13 The efficiency scores (θCRS) under the assumption of constant returns to scale (CRS) are reported in the second column, while the results (θVRS) of the variable returns-to-scale (VRS) model are listed in the fourth column. The CRS results indicate that the efficient frontier is dominated by the commercial banks. There is indeed a significantly large efficiency gap between the two sets of banks: the savings banks’ average efficiency score is less than 2/3 of the commercial banks’ average. About three-quarters of savings banks (33 out of 43), including one of the largest, mark an efficiency score below the overall sample average. All the largest savings banks, but also two commercial banks, seem to operate under decreasing returns to scale. In the VRS model, savings banks’ performance improves marginally: two entities, the size of which in terms of market share is small, are accorded efficient status and the average efficiency score of the savings bank group is about 70 percent the average efficiency score of commercial banks. Nonetheless, 31 savings banks mark a below-sample-average efficiency score.

Table 2.

Spanish Banks: Pre-M&A Efficiency Scores

article image
article image
Figure 3.
Figure 3.

Spanish Banks: Efficiency Scores 1/

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A004

1/ Size of the bubble indicates a bank’s market share.

17. The CRS score is called the (global) technical efficiency (TE) since it measures efficiency without taking into account scale effects, while the VRS score expresses the (local) pure technical efficiency (PTE) under variable-return-to-scale circumstances. It is possible to decompose the TE score into two components, one due to scale inefficiency and one due to “pure” technical efficiency. The ratio between the CRS and the VRS score provides a measure of scale efficiency:

(1)SE=θCRSθVRS

therefore, rearranging the terms:

(2)θCRS=θVRS×SE

That is, the CRS technical efficiency measure is decomposed into “pure” technical efficiency and scale efficiency.

18. This is graphically represented in Figure 4, where the CRS and VRS efficiency score are reported on the horizontal and vertical axis, respectively; while the size of the bubble represents a bank’s market share. In the sample considered, four commercial banks are estimated to operate at the most productive scale size since they are fully efficient both under the CRS and the VRS model (their respective bubbles lie on the (1,1) corner); another large commercial bank marginally underperforms this group of institutions. Two savings banks are “locally” but not “globally” efficient (full VRS efficiency but low CRS score) due to their scale inefficiency (while one institution is estimated to operate at decreasing returns to scale, the other one seems characterized by increasing returns to scale). The majority of savings banks lie along (or very close to) a 45 degree ray since they exhibit similar efficiency scores under the CRS and the VRS model. In other terms, their respective scale score is equal (or very close) to 1.14 Therefore, their low total efficiency (as measured by the CRS model) seems to be caused by inefficient operations rather than scale inefficiency. On the other hand, for a group of savings banks that lie above the 45 degree ray, scale inefficiencies contribute to explain their relatively low total efficiency score. This group comprises almost all the largest savings banks, whose scale inefficiency is due to decreasing returns to scale. In the case of the remaining savings banks of the group, which have small market shares, the existence of (unexploited) increasing returns to scale explains their scale inefficiency.

Figure 4.
Figure 4.

Spanish Banks: Pre-M&A Efficiency Scores

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A004

19. As mentioned above, mergers can be a powerful instrument to restructure credit institutions. The Inefficient Management Hypothesis suggests that inefficiently managed “target” banks provide a potential for wealth gains for “bidder” banks if the consolidated banks are transformed into well-managed banks.15 Although in the case of the consolidation process of the Spanish savings banks is difficult to “identify” what institution is the target and what institution is the bidder, Table 2 shows that in the grouping it is hard to find an outstanding “leader” in terms of efficiency (though there are clear cases of leadership in terms of size).

20. The Low Efficiency Hypothesis may fit the Spanish case better. According to this theory, the merger works as a “wake-up call” for the target bank’s management, which could use the merger as an opportunity to implement substantial corporate restructuring and to improve the efficiency of the consolidated bank, even though either or both the target and the bidder bank do not compare favorably with their industry peers.

21. To evaluate whether the new configuration of the Spanish banking sector that emerged by this wave of M&A can potentially help improve efficiency in the system, a test based on “virtual mergers” has been carried out. Although the grouping of institutions is the actual one, the mergers are “virtual” since are based on pre-merging information. As shown in Table 2, in most case, the merging institutions show a blend of decreasing and increasing returns to scale. The test will allow whether in the new environment overall efficiency would improve, whether scale factors continue to play a role, and whether the new institutions exhibit increasing or decreasing returns to scale.

22. To this end, the new “virtual” institutions have been constructed based on the assumption that each participating savings bank is locally efficient, that is it lies on the efficient frontier determined by the VRS model. In other words, the inputs and outputs of the new “virtual” institutions are given by the sum of the inputs and outputs of merging institutions should they lie onto the efficient frontier (the computer program provides the input and output that would correspond to such a situation).16 This sort of projection of the merging entities onto the efficiency frontier (calculated under the VRS model) would mimic the impact of the ongoing restructuring and reorganization process. To illustrate the process, Figure 5 shows it in the case of single input and single output.

Figure 5.
Figure 5.

Virtual Merger Representation

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A004

23. In this simple example, S1 and S2 are two locally inefficient savings banks since they are positioned within the efficient frontier. The restructuring process brings them to position onto the efficient frontier (point A and B, respectively). S3 is the new “virtual” institution that is created by the merger of the two savings banks.

24. In this new “virtual” environment, however, a new efficient frontier would prevail and the DEA exercise is hence run again on the new set of institutions. The results of this exercise are reported in Table 3 and illustrated in Figure 6.

Table 3.

Spanish Banks: Post M&A Efficiency Scores

article image
Figure 6.
Figure 6.

Spanish Banks: Post-M&A Efficiency Score 1/

Citation: IMF Staff Country Reports 2011, 216; 10.5089/9781462340545.002.A004

1/ Size of the bubble indicates a bank’s market Share.

25. In this new hypothetical setting, the efficiency frontier remains dominated by commercial banks but almost all the new “virtual” institutions show a sizeable increase in their technical efficiency, as indicated by CRS efficiency scores much closer to 1. The average CRS efficiency score for savings banks is much closer to the commercial banks’ average. However, two groups formed by relatively small savings banks underperform the other mergers in terms of efficiency score while presenting a scale score equal or very close to 1. This result indicates that these two mergers, which involve savings banks from within the same region, may be ex ante expected, ceteris paribus, to continue to perform less efficiently due to their inefficient operations rather than scale inefficiency. Furthermore, a number of the new entities continue to operate at decreasing returns to scale and hence they have the possibility to improve their efficiency by scaling down their activities.17

D. Conclusions and Challenges Ahead

26. This study suggests that before the recent consolidation process, the savings bank sector had accumulated significant inefficiency—both of scale and operation—that the crisis has revealed. Although the results have to be considered with caution, they suggest that the road to achieving efficiency gains may be challenging as the new groups do not seem to include a “leader” in terms of operational efficiency, though many do have a leader in terms of size. Nevertheless, the merger process could still enhance efficiency by providing a “wake-up call” to improve the management of the merged bank and by reducing the fragmentation of the sector. Modeling the “best case” scenario of the mergers, in which all the banks involved in a merger are (purely) technical efficient, a substantial improvement in overall efficiency can be obtained. However, some of the mergers among small institutions do not seem best configured, ex ante, to deliver substantial efficiency gains. This underscores the importance of achieving these efficiency gains, as planned, through substantial restructuring, reorganizing, and downsizing that could also prompt a second round of integration.

Appendix

The following provides a short description of the DEA methodology.18 Assume that there are k inputs and m outputs for each of the n banks. For the i-th bank these are represented by the vectors xi and yi, respectively. The k x n input matrix, X, and the m x n output matrix, Y, represent the data of all n banks. It is also assumed that banks are operating with constant returns to scale (CRS). For each bank, the purpose is to obtain a measure of the ratio of all outputs over all inputs, such as u’yi/v’xi, where u is an m x 1 vector of output weights and v is k x 1 vector of input weights (superscript’ indicates transpose).

To select the optimal weights, the following mathematical programming problem has to be solved:

maxu,vuyi/vxi(3)s.tuyj/vxj1u,v0j=1,2,….,n

To avoid infinite solutions to the above problem, the constraint v’xi = 1 is imposed, which leads to:

maxμ,vμyi(4)s.tvxi=1μyjvxj0μ,v0j=1,2,….,n

where the notation of the weights has changed from u and v to u. and v, respectively, in order to reflect the transformation.

Using the duality in linear programming, an equivalent envelopment form of the above problem can be derived:

minθ,λθ(5)s.t.yi+Yλ0θxiXλ0λ0

Where θ is a scalar and λ is a n x 1 vector of constraints. The value of 9 is the efficiency score for the i-th bank, which ranges between 0 and 1. Therefore the problem has to be solved n times, one for each bank, in order to have the full picture.

However, the CRS assumption is rather restrictive. A number of factors, including imperfect market competition, may cause a bank to be not operating at optimal scale, i.e. along the flat portion of the long-run average cost curve. To allow variable returns to scale (VRS), it is necessary to add to the problem in equation (4) the convexity constraint:

Iλ=1(6)

where I is n x 1 vector of ones.

The difference between the efficiency scores calculated under the VRS and the CRS assumptions provides an indicator of scale inefficiency. In other words, the difference between the two efficiency scores indicates the additional gain in efficiency that could be achieved if banks were operating at the long-run equilibrium CRS.

References

  • Ayadi, R. and G. Pujals (2005) “Banking Mergers and Acquisitions in the EU: Overview, Assessments and Prospects,” SUERF - The European Money and Finance Forum.

    • Search Google Scholar
    • Export Citation
  • Avkiran, N. (2005) “Developing foreign bank efficiency models for DEA grounded in finance theory,” Socio-Economic Planning Science, 40, pp. 275296.

    • Search Google Scholar
    • Export Citation
  • Banker, R., H. Chang, and S. Lee (2010) “Differential impact of Korean banking system reforms on bank productivity,” Journal of Banking & Finance, 34, pp. 14501460.

    • Search Google Scholar
    • Export Citation
  • Casu, B., C. Girardone, and P. Molyneux (2004) “Productivity change in European banking: A comparison of parametric and non-parametric approaches,” Journal of Banking & Finance, 28, 25212540.

    • Search Google Scholar
    • Export Citation
  • Coelli, T. (1996) “A Guide to DEAP Version 2.1: A Data Envelopment Analysis (Computer) Program,” CEPA WP 96/08.

  • Cook, W. and J. Zhu (2005) Modeling Performance Measurement: Applications and Implementation Issues in DEA, Springer.

  • Cooper W., L. Seiford, and K. Tone (2007) Data Envelopment Analysis, Springer, Second edition.

  • Group of Ten (2001) “Report on Consolidation in the Financial Sector,” January.

  • Holod, D. and H. Lewis (2011) “Resolving the deposit dilemma: A new DEA bank efficiency model,” Journal of Banking & Finance, forthcoming.

    • Search Google Scholar
    • Export Citation
  • International Monetary Fund (2010) “Spain,” Country Report No.10/254, July 30.

  • Kohers, T., M. Huang, and N. Kohers (2000) “Market perception of efficiency in bank holding company mergers: the role of the DEA and SFA models in capturing merger potential,” Review of Financial Economics, 9, pp. 101120.

    • Search Google Scholar
    • Export Citation
  • Röller L., J. Stenneck, and F. Verboven (2000) “Efficiency Gains from Mergers: Research Area Market Process and Corporate Development, August.

    • Search Google Scholar
    • Export Citation
  • Sathye, M. (2001) “X-efficiency in Australian banking: An empirical investigation,” Journal of Banking & Finance, pp. 613630.

1

Prepared by A. Giustiniani. I am grateful to Kevin Ross for his comments and technical help. All the flaws and errors remain mine.

2

A SIng savings banks pool resources (e.g., capital, liquidity, risk management) with a central entity while maintaining some practical some practical and legal independence, also called “cold-merger,” is a sort of joint-venture in which participating. In the note, especially in the empirical section, the term “merger” will be used indistinctively.

4

IMF (2010).

5

Recently one of the original SIP (Banco Base) broke up. While one of the participating savings banks (Caja del Mediterráneo) is currently seeking a new partnership, the other savings banks decided to form a new SIP (Effibank). Moreover, the three Basque savings banks (Kutxa, BBK, and Vital) are negotiating a possible merger.

6

The FROB has been authorized to acquire stakes in banks’ share capital for a limited period of time (no longer than 5 years) to strengthen their own funds. The beneficiary institutions have to implement a recapitalization plan, approved by the BdE. In case of a savings bank or an SIP, the lending activity has to be transferred to a bank by the mechanisms stipulated by the law (indirect exercise of financial activity or conversion into a foundation owing a bank).

7

In spinning-off their banking business, two institutions so far (BFA-Bankia and Caixa) have segregated their impaired real-estate assets in a separate company (either credit institution or other financial entity) together with other profitable assets to compensate for the low income stream of the former group of assets.

11

Given the context, perhaps the number of branches and employees would have been more direct variables to consider. Unfortunately, those data were not available for all credit institutions in the considered period. Other specifications of the model comprising flow and balance sheet variables have been tested without significant improvements.

12

The DEA program by Cook and Zhu (2005) has been used in this study.

13

The analysis does not take into account the recent breakdown of the SIP at the basis of Banco Base as well as the potential merger between Basque savings banks.

14

Geometrically, the scale score would be represented by the cotangent of the angle formed by the ray joining the axes origin with the bank-data-point.

15

T. Kohers, M. Huang, and N. Kohers (2000).

17

As for the other results, one savings bank, BBK, is now fully efficient under the CRS and the VRS model; while the other two savings banks from the Basque region (Vital and Kutxa) could improve their efficiency by scaling up their activities. Ibercaja and Pollensa are not locally efficient any longer and this would explain the drop in their technical efficiency, since the scale factor is equal to 1.

Spain: Selected Issues
Author: International Monetary Fund