Introduction

In the summer of 2007 the "subprime mortgages" crisis started. These were a financial product whose risks were spread among various agents and markets through securitization and credit derivatives. At first, this situation did not affect the Spanish authorities directly, but it was one of the triggers of a chain of effects in global economies and financial markets that soon, among other factors, would affect Spain.

Its immediate consequence was the deterioration of liquidity and funding, which drove several authorities to take support measures which had varying effects both over time and in inland areas of application.

In the case of Spain the passage from the crisis of the financial economy to the real economy took less than a year from the appearance of the first symptoms in North America. The Spanish economy fell into recession in late 2008. The fall in production and rising unemployment pushed up default rates of banks and savings banks and, consequently, the need for higher provisions of insolvency. Furthermore, there was an increase in financial costs brought about by the increasing reliance on wholesale funding and also by sudden difficulties in these markets. In addition, the slowdown in the housing market affected many entities that had a high concentration in the field of real estate development and construction (^{Aríztegui, 2011}).

With this background, it was already clear that there was an urgent need to undertake a restructuring of the sector and to support this process in a way that allowed the solvency to be strengthened of those institutions that decide to undertake plans to improve their efficiency through mergers or other integration transactions.

Clearly by late 2011, and throughout 2012 and 2013, we were witnessing an intense reorganization process of territorial financial institutions (savings banks) which, forced by a series of regulatory rules, sought to strengthen the system and thus "solve the problems of liquidity, solvency and default of these entities" (^{Palomo, Sanchis-Palacio, & Gutiérrez-Fernández, 2011}).

Consequently, savings banks^a are in particular the main figures in the restructuring process. Their legislation reform involved significant changes in its legal form, with several variants. Thus, the adjustment to the declining demand gave rise to a situation of excess capacity, leading to an urgent need to resize the sector (^{González Sánchez & García Muiña, 2011}). All the savings banks, except Caixa Ontinyent and Pollensa, transferred their assets and liabilities to another entity regulated as commercial bank with the structure of a public limited company, in which those saving banks which have not been nationalized or taken over are major shareholders.

These circumstances are of sufficient importance to merit research on whether these restructuring processes are improving the situation of the sector. We must point out that these restructuring processes are not unique of the Spanish Banking system. As we deep in the previous literature, we observe how these processes were carried out in other countries, even when the intensity of the crisis was not as worldwide as now. In this sense, we noted that several studies addressed the impact of the reform of the financial systems in the sector from an international point of view. An example of this can be seen in Table 1, although some of the studies are not strictly comparable in terms of full homogeneity, not only for the political systems in the respective countries but also for the type of problems, as they provide information on the intended purpose of restructuring examples in the banking sector at some level.

After this review, we observed how after the restructuring process there are two variables, efficiency and solvency, which in most cases improve considerably.

This paper aims to be the first in a series of observations over the next few years (until 2016 at least) and aims to ascertain, first which financial entities had higher efficiency and solvency values. It also seeks to discover whether the concentration decisions of the Spanish banking sector followed the guidelines according to the values obtained by these two variables and according to previous experiences in other contexts. After that first goal, we thought it important to analyze the situation after the first integrations and to provide the methodology, with the values of the ratios indicated, for analysis in future years.

Therefore, for proper monitoring and analysis, this paper is structured in five sections. After the introduction, the second part provides a review of the development of those standards affecting the banking sector during 2009-2013. The third section describes the sample and presents the methodology and variables used in the analysis. Section 4 presents the findings and Section 5 analyzes the entities that were involved in any restructuring process. The paper ends with the main conclusions and lines for future research.

Literature review

At a European level, several studies analyze the significant impact degree that the Economic and Monetary Union (EMU) has on the competence of the banking sector in those countries which adopted the single currency (^{Bikker & Groeneveld, 2000}; emsp ^{De Bandt & Davis, 2000}); whether bank concentration in a national market has a negative impact on the financial solvency of European banks (^{Uhde & Heimeshoff, 2009}) or the determinants of bank profitability in several European countries, and also that a higher degree of concentration leads to a higher profitability (^{Molyneux & Thornton, 1992}). In this sense, international comparisons of banking efficiency can be used to identify, among others, issues of international consolidation of the banking industry. ^{González (2009)} analyzes the influence of bank efficiency and policy variables on the market structure. Their results are consistent with the efficiency-structure hypothesis, which states that the most efficient banks have, on average, higher levels of domestic market that enhance market concentration.

Studies have demonstrated appropriate values of efficiency and solvency of the Spanish banking sector, mainly in late 2008 (^{Álvarez, 2008}; emsp ^{Maudos, 2009b}). In this sense, although this efficiency had fallen from 2007, in 2008 it stood at 47%. As regards degree of solvency, in December 2008, with a value of 11.3%, Spain was at levels close to the European average (11.9% in the EU-15), well above the minimum required 8%. (^{Maudos, 2009a}). However, in 2010, efficiency deteriorated. Bigger banks were more efficient than the industry average (^{Maudos, 2011}). As for the differences between banks and savings banks, previous studies show how the savings banks are less efficient and solvent than banks. ^{Fernández de Guevara (2007)} analyzed the period 1992-2004 and ^{Carbó and Maudos (2010)} to December 2009. ^{Climent (2013)} analyzes the determinants of solvency of the Spanish financial institutions in the period 2006-2009, concluding that banks achieve better solvency than savings banks.

Other studies show how in banks and savings banks there is a significant positive correlation between measures of size, efficiency and financial profitability (^{Marín, Gómez, & Gómez, 2008}) or ^{Bernad Fuentelsaz, & Gómez (2009)}, who claim that although it seems that mergers and acquisitions have sometimes been an effective way to deal with increased competition deriving from the new competitive environment, they do not always manage to improve the results of the entities involved. Moreover, other studies like ^{Fernández, Fonseca, and González (2006)} find that banks take more risks than savings banks, and ^{Fonseca and González (2005)} state that as the political presence increases on savings banks, so does their risk, as they may take riskier investments in the pursuit of political ends. The possibility of transferring funds in the commercial relationship between savings banks and regional governments can also favor risky investments, especially in saving banks with higher levels of public participation.

Elsewhere, the latest stress test published (September 2012) for a number of entities, shows this initial trend of solvency strength behind decisions made.

Regulatory developments 2009-2013

Apart from the legal aspects of the savings banks and their need to transform and update as a consequence of the necessary adaptation to European directives, (including the initial definition of the Fund for Restructuring and Banking management), we can state that it was not until late 2011, early 2012, when the standards, were issued (and measures of economic and financial policy), were clearly determined in Spain regarding restructuring and, especially, decreasing the debt of the financial system.

In this sense, we briefly show in Figure 1 synopsis and content of the standards and their evolution from 2009-2013.

Fig. 1 Standard evolution 2009-2013. Source: own elaboration. 

As a result of all this legislation, several steps were taken to restructure the sector, as is shown in Table 2.

Without doubt, there will still be changes made in the last column, so it is important to note the content and results of the so-called stress test. The last one made in Spain points to a "clear exit point" from what we have commented on above in terms of future changes.

The stress tests have sought to assess the resilience of the banking system, and banks' ability to absorb possible shocks related to their credit risks and the market, including the risk arising from investments in sovereign debt. Two macroeconomic scenarios were mainly considered: a reference and an adverse one. The latter will determine the potential need for additional capital.

In Europe, the European Banking Authority (EBA) published the stress tests of the European Union for the years 2010 and 2011 and announced that it would not conduct new stress tests for 2012, with the next stress test for European banks to be in 2014. Nevertheless in Spain, in a clear attempt to be transparent, a series of additional stress test, in addition to the European ones, were made in 2012.

A summary of the stress tests carried out by the Spanish banking sector is presented in Table 3. We show the estimated capital needs for the Spanish banking sector under an adverse scenario.

It is important to note that changes in the methodology, along with others in how results are presented, makes it difficult to compare results.

Sample, variables and method

Method

The analysis seeks to answer the question of whether the restructuring process of the banking sector followed the guidelines of efficiency and solvency of the sector. We use a multivariate model that allows us to verify the predictive capability of our previously defined explanatory variable, RESTRit, where each entity is i and year is t:

RESTR: Before or after the restructuring

B_S: Bank or Savings Bank

TEST: Additional capital requirement under the stress test

TIER1: Percentage of Tier1

SIZE: Logarithm of total assets

With this model we analyze the capacity of the explanatory variable to predict two versions of the dependent variable, Xit, the efficiency (EFFICit), defined as the ratio between operating expenses and gross income, and solvency (SOLVit ), defined as the ratio between equity and liabilities.

With the above model, we analyze the regression of two versions, depending on whether the dependent variable is efficiency or solvency.

Results

Table 5 shows the mean, standard deviation and percentiles 25.50 (median) and 75 of the analysed variables. In order to go deeper into the relationship between the variables, we show in Table 6 their correlation coefficients. It is observed how these coefficients could be described as fair values and, therefore, we can state that multicollinearity is not a problem in our models, since the variance inflation factor (VIF) used to ensure that our results are not biased by the correlation between the explanatory variables (^{Marquardt, 1970}) is, in all cases, within acceptable limits (^{Guajarati, 1997}).

In order to clarify the explanatory model, we show the Granger causality test.

Efficiency:

Variable definition:

Unrestricted model (U):

Restricted model (R):

The test is performed by comparing the F statistic, given by:

Fort the analyzed case, we have:

On the other hand, the SSR are shown in the following tables,

Then, by replacing the respective values in equation (3) we obtain:

Therefore, we conclude that restructuring has an statistically significant effect on efficiency according to Granger.

Solvency:

Variable definition:

Then, by replacing the respective values in equation (3) we obtain:

Therefore, we conclude that restructuring has no effect on solvency according to Granger.

Table 7 provides the mean value of the main variables analyzed. It is calculated according to the RESTR variable. The value of the Mann Whitney and the associated significance level are also given.

The results show how the two variables present better values on simple average before the restructuring process than later, which is logical at this first stage. These differences are statistically significant for the EFFIC variable.

Table 8 shows the evolution of the two variables for the Spanish banking sector during the period 2008-2012. We note that until 2011, the two variables get worse, with efficiency being hit harder, rising from 50.01 in 2009 to 64.28 in 2011. However in 2012 there is a change of trend and both variables begin to improve. In both cases the financial decisions of integration taken could be the explanation.

Tables 9 and ¹⁰ show the comparison, for each year, of efficiency and solvency for both Banks and Saving Banks. We must take into account that the restructuring process of the banking sector led to the grouping of savings banks in SIP that finally ended in the creation of banks. We must point out that those banks created through SIP among saving banks, were still included in the group of savings banks when presenting the data. This is because they originally came from savings banks and in the analysis period they are still in an initial process of restructuring.

First, we note that the efficiency of banks is better than the savings banks. These differences are statistically significant for the years 2008, 2009 and 2011. Thus, our first research question, about efficiency, confirms as valid those financial policy decisions proposed as more efficient from the economic and financial point of view. That is, the legal transformation of the savings banks, their absorption by the banks and/or their integration to become banks.

Similarly, for solvency, it is observed how the solvency of banks is better than savings banks. These differences are statistically significant for the years 2009, 2011 y 2012. Therefore, the above comments for efficiency would be the same for solvency.

Analysis

Table 11 shows the results of regression models 1 and 2. The results obtained are entirely consistent with what was indicated in the descriptive analysis. For Model 1, where the dependent variable is efficiency, RESTR variable reflects a positive relationship in the coefficient and a statistically significant value. This positive relationship in the coefficients is just confirming that efficiency in the banking sector was, on average, better before the restructuring than later. This would be explained by the fact that the restructuring process which would bear fruit in subsequent years began in 2011. This shows that those with better efficiency and solvency are the ones which must offset the lower values of the rest, which explains the drop in average after the initial restructuring.

Meanwhile, variables SIZE and TIER reflect a negative relationship in the ratio and return statistically significant values. This negative relationship indicates that larger institutions are more efficient than smaller ones. Also, entities with the highest percentage of Tier1 ratio have higher efficiency than those with the lowest percentage of Tier1.

As for model 2, where the dependent variable is solvency, results show that B_S and TIER variables reflect a positive relationship in the ratio and have statistically significant values. These positive relationships indicate that banks have better solvency ratio than savings banks and that those which show the highest percentage of Tier1 have higher solvency ratios.

Conclusions

As a result of the general economic situation, the Spanish financial system was forced to undertake a restructuring process, which finally began in late 2011 - annual accounts presented in June / July 2012 - and during the years 2012-2013. The existence and disposal of the annual accounts after the application of 2011 and 2012 provides consistency to our study.

This paper, in its main research question and from the public inancial data available, confirms that:

We confirm that these processes have not been yet fruitful in terms of the values obtained by the Spanish financial institutions in their ratios of efficiency and solvency but no doubt they will in the future, especially considering the legislative decisions taken in late 2011 and throughout 2012, and financial information containing the first financial statements published after restructuring.

Indeed, from the second analysis made, which is justified because of the data-set that it provides, which means we can carry through this investigation in the following years (2013-2016), we can conclude that:

In short, we can say that these results justify the need for further steady implementation and control of the reform processes in the Spanish banking sector. The experience of previous restructuring processes shows that, although they certainly do require time, it is probably harmful to extend them too much, as this involves uncertainty in the market and the main agents related to these institutions. Therefore, the decisions taken on economic and financial policy which establish deadlines and coercive criteria seem to be correct, because previous literature confirms that restructuring should be done in a decisive and quick way, with regulatory intervention and keeping investors and several other agents in the process informed.

Regarding the future research lines, we understand the need to supervise the future development of both main variables analyzed (profitability and solvency). We would also like to check if, as we understand that it will happen, the evolution of both variables will still be positive and will increase when the restructuring processes come to an end.