Discretion in the form of intervention can remove it from the control of the state.
Suppose a thug puts a gun to your head and demands your wallet. Because you have sentimental photos in your wallet, you offer instead the keys to your car. You would rather lose the car than the wallet. Can the thug claim in court that you voluntarily gave him your car, therefore he did not steal it? Most judges would agree with you that you did not willingly give your car away.
However, on 19 March 2019, the General Court ruled that a recapitalisation of the Italian bank Tercas by a private banking consortium was not State aid because the members of the association decided to inject money in the bank to save it rather than let it go bust and pay to depositors the legally mandated amount of up to EUR 100,000 per depositor. Was the recapitalisation, as the lesser of two evils, attributable to the consortium or the state?
The General Court ruled in joined cases T-98/16, T-196/16 and T-198/16, brought by Italy, Banca Popolare di Bari, formerly Tercas Cassa di risparmio della provincia di Teramo [Banca Tercas], Fondo Interbancario di tutela dei Depositi, and the Bank of Italy, against the European Commission.1 All the applicants requested the annulment of Commission Decision 2016/1208 which had found that State aid for Banca Tercas was illegal and incompatible with the internal market.
The Italian private consortium Fondo interbancario di tutela dei depositi [FITD] injected capital in Banca Tercas [Tercas] after the bank had been put under special administration. The injection was authorized by the Central Bank of Italy in July 2014.
Tercas was a privately-owned bank which was acquired by BPB, a holding company of a privately-owned banking group. FITD is a private law inter-bank consortium that was formed on a voluntary basis in 1987. This consortium is mutual in nature and has been set up to pursue the common interests of its members. The purpose of FITD is to ensure depositors of its members. In 1996 FITD was recognized by the Bank of Italy as one of the deposit guarantee schemes authorised to operate in Italy. According to its articles of association, FITD intervenes to protect deposits up to EUR 100,000 [this is the so-called compulsory intervention], but also on a voluntary basis, if the intervention makes it possible to reduce the cost of the guarantee of deposits borne by its members [the so-called voluntary intervention].
More specifically, when FITD can expect a lower cost, instead of reimbursing deposits in the event of compulsory administrative liquidation, it may intervene in the form of asset and liability transfer operations to support the member concerned. In other words, instead of allowing a member in financial difficulty to go bust and then pay off depositors, FITD may bail out the member and keep it operational, if the bail out is less costly.
On 30 April 2012, the Italian authorities placed Tercas under extraordinary administration and appointed a commissioner to manage Tercas. After several rounds of negotiations between FITD and Tercas, eventually on 7 July 2014, the Bank of Italy authorized an intervention by FITD in favour of Tercas. This intervention provided for three measures which formed the subject-matter of the decision of the Commission: A capital injection of EUR 265 million; a guarantee of EUR 35 million; and another guarantee of EUR 30 million. FITD had decided that this form of intervention was less costly than a simple payment of up to EUR 100,000 per depositor. At the time of the intervention, Tercas held about EUR 2.7 billion in deposits. To ut this intervention in perspective, at about the same time the average amount of savings in Italian banks was EUR 25,000 per depositor. If Tercas had more than 11,000 depositors, it would have cost more than EUR 265 million to FITD to compensate them.
In December 2015, the Commission concluded after an investigation that the intervention constituted incompatible State aid and ordered its recovery because the aid did not comply with the relevant guidelines on restructuring of financial institutions.
The main argument of the applicants was that the resources used by FITD were not state resources. Therefore, its intervention in favour of Tercas could not be classified as State aid.
The General Court first recalled the meaning of the concept of State aid and in particular that in order for benefits to be qualified as “aid”, they must be granted directly or indirectly by means of state resources and be attributable to the state. It is not necessary to distinguish between aid that is granted directly by the state from that which is granted through a public body or private, designated or established by the state. [paragraphs 62-64]
The Court stressed the importance of imputability of aid decisions to the state. This is because even if the state is able to control a public enterprise and exercise a dominant influence over its operations, the actual exercise of that control in a concrete case cannot be automatically presumed. [paragraph 68]
In a situation where the aid is provided by a private entity, the Court warned that it is not possible to presume that the state is in a position to control this entity and to exert a dominant influence over its operations even if there are share capital links between that entity and the state. [paragraph 69]
The Court decided to examine first whether the decision of FITD could be imputed to the state and then whether the resources used came under the control of the state. However, the Court began its analysis by looking into the “origin” of the resources, without explaining how the concept of origin differed from the concept of imputability, especially given the fact that the Commission itself had stressed that their origin was irrelevant.
The state origin of the measures
The Commission had considered that the determining factor was not the direct origin of the resources [paragraph 112 of the Commission decision], but the degree of intervention of public authorities in the definition of the measures and their methods of financing. The Court summarised the elements taken into account by the Commission:
- Italian authorities had entrusted FITD with a “public mandate” relating to the protection of depositors. The Commission thought it was not relevant that FITD was established as a private-law consortium.
- Italian authorities were able to influence all stages of the implementation of the measures in question. The Bank of Italy had the power to authorise the intervention of deposit guarantee schemes and representatives of Italian public authorities were present in all the decision-making meetings.
- Italian banks were obliged to join FITD and contribute to the interventions which were decided by its management. The intervention in question was attributable to FITD.
The Court noted without disagreeing the reasoning of the Commission and the facts it took into account. [paragraphs 71-82]
Do you know we also publish a journal on State aid?
The European State Aid Law Quarterly is available online and in print, and our subscribers benefit from a reduced price for our events.
On the imputability of the measures in question to the state
The General Court observed, first, that in order to assess the imputability of a measure to the state, it is important to examine whether public authorities were involved in the adoption of that measure. [paragraph 83] As established in the landmark Stardust Marine case [C-482/99], imputability can be inferred from a number of indicators each of which may not be decisive on its own.
The Court cautioned, however, that given the autonomy of private entities, it is not enough that the Commission proves that the absence of influence by the state is unlikely. [paragraph 89] Here the Court uses a double negative expression that is difficult to understand its meaning.
I suppose the expression “the absence of state influence is unlikely” implies that the presence of influence is more probable than the simpler expression “the presence of influence is likely”. The use of the more complex expression by the Court suggests that the Court intended to raise the standard of proof and make it harder for the Commission to prove the actual involvement of the state. [For example, it is easier to defend the statement that “we cannot exclude the possibility that climate change causes the present weather patterns” than to prove the statement that “climate change is likely to cause the present weather patterns”. The former it is sufficient to have mere doubt, while the latter requires substantial evidence.]
Indeed, the Court went on to stress that, in view of the fact that the measure in question was provided by a private entity, the Commission was all the more obliged to substantiate the evidence that allowed it to conclude that there was sufficient proof that the measure had been adopted under the effective influence or control of public authorities. [paragraph 90] According to the Court, only a significant degree of intervention makes it possible to conclude that the measures at issue are attributable to the Italian state. [paragraph 91]
The Court noted the claim of the applicants that the intervention in favour of Tercas was decided by the management of FITD with the unanimous agreement of all its members, after a simple request from the administrator of Tercas. According to them, no public authority could have required FITD to proceed with the intervention if it had not been in the interest of its members of the consortium. [paragraph 92] Then the Court examined the evidence on which the Commission relied in its decision.
On the scope of the public mandate entrusted to FITD
The Court observed that the support measures of FITD were mainly aimed at pursuing the private interests of its members and to avoid the more severe financial consequences of a refund of deposits in the event of compulsory administrative liquidation and the repayment of deposits. FITD would choose the less onerous means whenever there were prospects for recovery of the troubled bank benefiting from the intervention. [paragraphs 96-97]
The fact that FITD chose to inject capital in Tercas and voluntarily assumed a certain risk of losing its money was not surprising at all. The alternative was a sure loss of up to EUR 100,000 per depositor. FITD would not have chosen to bail out Tercas if it did not have the obligation to reimburse depositors. The Court ignored the significance of this factor. It focused on the form of intervention rather than the fact that FITD had to intervene. For the Italian state, the form was irrelevant as long as deposits up to EUR 100,000 were safe.
A numerical example can shed light on this important point. FITD had two options: not to intervene and, say, lose 100 with certainty by paying depositors or intervene with, say, a capital injection of only 80 and have only a 70% chance of making Tercas viable again. However, if the intervention failed, FITD would still pay the 100 to depositors. Therefore, the monetary values of the options of FITD were:
No intervention = Sure loss of 100
Intervention = Return in case of success – loss in case of failure = (80 x 0.7) – ((100 + 80) x 0.3) = 56 – 54 = 2
A chance of getting back 2 is better than a certain loss of 100. Or, to put it differently, a loss of 78 [= 80 – 2] is better than a loss of 100. But no rational investor who did not have the obligation to compensate depositors would voluntarily invest 80 in a failing bank even if there was a pretty good chance of making it viable again. The expect return of the capital injection itself was a loss of 24 [= 80 – 56]. FITD would not have intervened unless it was obliged by the state. Once more, the form of the intervention seems to be irrelevant as long as depositors were saved.
Nonetheless, the Court went on to state that for FITD members, support interventions were also intended to avoid negative consequences for them and for the banking sector as a whole, particularly in terms of reputation and the risk of depositors panicking by forced administrative liquidation. [paragraph 98]
This is an interesting, but unsubstantiated argument. Had any of the parties in the case made such a statement, the Court would have surely struck it down for being an unproven assertion. Just because it is in the collective interest of the banking sector to avoid runs on banks, does it necessarily follow that banks would always support each other or that that was the actual reason for which FITD intervened in Tercas?
I suppose that the Court referred to the private interests of the members of FITD in order to demonstrate that they acted voluntarily. Indeed, they chose the less costly option voluntarily. But that was because they were required by law to intervene to protect deposits up to EUR 100,000. Nothing in the judgment suggests that in the absence of that requirement, FITD would have intervened at all.
Then came the first critical appraisal of the Commission’s decision by the Court. It found that, contrary to the Commission’s contention, the support measures did not implement any public mandate imposed by Italian law. [paragraph 100]
The Court referred to the Italian Banking Act, which provided that deposit guarantee systems had to make payments in the event of compulsory administrative liquidation of licensed banks and inferred that the public mandate imposed on the FITD consisted solely of reimbursing depositors only up to EUR 100,000 when a bank was compulsorily liquidated. Moreover, it considered that the Italian law in no way obliged the FITD to carry out other support measures. [paragraphs 101-102]
The Court concluded that the possibility for FITD to perform support interventions did not emanate from any legal obligation, but only from autonomous decisions of the member banks. Therefore, contrary to what the Commission claimed, the support measures had a purpose other than the repayment of deposits and did not constitute implementation of a public mandate. [paragraphs 104-106 ]
In essence, the General Court focuses on the form of intervention chosen by FITD and ignores that FITD had to intervene in one form or another.
On the autonomy of FITD when deciding on the intervention
Then the General Court examined the decision-making autonomy of FITD. With respect to the authorisation by the Bank of Italy of the intervention of FITD in favour of Tercas, the Court did not consider it as an indication that the measure in question could be attributed to the state. First, the authorisation was given only after the conformity of the measure with the relevant banking regulations was verified. Second, the Bank of Italy did not require FITD to intervene to support banks in difficulty. Third, the representatives of the Bank of Italy attended meetings of the management of FITD as observers with no voting or advisory role. Moreover, the Commission, according to the Court, adduced no evidence to support its assertion that the presence of the representatives influenced the decision of FITD or coordinated the discussion between the members of FITD. The power of the Italian supervisory authorities to place a bank in extraordinary administration does not enable them to compel FITD to support the bank. [paragraphs 113-131] When the Court refers to “support”, it means the intervention of FITD in the form of capital injection.
The General Court concluded that the Commission had erred in claiming that it had demonstrated that the Italian authorities had exercised a substantial public control in the definition of the FITD intervention in favour of Tercas. Therefore, the Commission failed to prove that the measure in question was imputable to the state. [paragraph 132]
I think it is irrelevant that the Italian authorities did not exert influence on the decision of FITD as long as FITD was compelled by law to cover deposits up to EUR 100,000 in case Tercas would fail. It was in the interest of FITD to prevent failure if it would cost it less.
On the financing of the intervention by means of state resources
The General Court, first, explained that even when aid is funded by resources which are not permanently in the possession of the state, the fact that they remain under constant public control and therefore at the disposal of the state is sufficient for them to be classified as state resources.
Then it recalled that in order to conclude that FITD’s intervention in favour of Tercas was financed out of state resources, the Commission took into account the following: the fact that the FITD had a public mandate; the control of public authorities over the resources used by the FITD to finance the intervention and the fact that the contributions used by the FITD to finance the intervention were mandatory. [paragraph 139]
With respect to the public mandate of FITD, the General Court pointed out that the mandate imposed on deposit guarantee schemes in Italy only required the repayment of covered deposits in the event of default of a credit institution. The public mandate did not require that those schemes also had to intervene before such a failure. [paragraph 141]
With respect to the control exercised by public authorities over the resources used by FITD to finance the intervention, the Court reiterated that the intervention in question was not dictated by public authorities. The Bank of Italy only exercised its supervisory powers to ensure the stability of the financial system. [paragraphs 142-149]
With respect to the mandatory nature of contributions to FITD, the Court first noted that the funds used for the intervention of FITD in favour of Tercas were private resources which had been contributed by the member banks. [paragraph 151] The amount of contributions was fixed in proportion to the assets of each member.
The Court further noted that the obligation of the members of FITD to contribute to the intervention decided by the latter was not mandated by the legal obligation to pay cover deposits but emanated from an autonomous decision of FITD. [paragraph 154] The intervention of FITD in favour of Tercas was in the interest of all its members, since they could have to spend more money in case of reimbursement of depositors. [paragraph 156]
The Court concluded that the compulsory nature of the contributions of the members of the FITD to the intervention originated in a decision doubly accepted by those members, not only by their decision to join FITD whose articles of association provide for the possibility of supporting banks through capital injection, but also by their decision to endorse the intervention which was in line with their interests. [paragraph 159]
On the basis of the above reasoning, the Court found that the Commission did not sufficiently establish that the resources in question were controlled by the Italian authorities so that they would be at the disposal of those authorities. [paragraph 161]
It then proceeded to annul Commission decision 2016/1208.
The General Court focused exclusively on the fact that FITD could choose a less costly option than just to pay out depositors. Apart from asserting in general terms without any proof that it was in the interest of all members to protect financial stability [paragraph 98 of the judgment], the Court never asked the fundamental question whether the form of FITD’s intervention made any difference to the effect of the intervention and to the achievement of the public policy objective of protecting covered deposits. FITD would not have chosen to bail out Tercas if it did not have to compensate depositors. Covered depositors and the state were indifferent between reimbursement of covered deposits and the safe keeping of those deposits in a freshly recapitalised bank. The General Court did not see the gun of the state pointing to the head of FITD. It saw only the supposedly voluntary injection of capital that indeed was not explicitly demanded by the state.
1 The full text of the judgment, in languages other than English, can be accessed at: http://curia.europa.eu/juris/fiche.jsf?id=T%3B98%3B16%3BRD%3B1%3BP%3B1%3BT2016%2F0098%2FJ&oqp=&for=&mat=CONC.AIDE%252Cor&lgrec=en&jge=&td=%3BALL&jur=C%2CT%2CF&etat=clot&dates=%2524type%253Dpro%2524mode%253D1M%2524from%253D2019.03.01%2524to%253D2019.04.01&pcs=Oor&lg=&pro=&nat=or&cit=none%252CC%252CCJ%252CR%252C2008E%252C%252C%252C%252C%252C%252C%252C%252C%252C%252Ctrue%252Cfalse%252Cfalse&language=en&avg=&cid=7995137.