Co-op Bank’s near-collapse – supervision under scrutiny

March saw the publication of Mark Zelmer’s ‘Independent review of the prudential supervision of The Co-operative Bank Plc’. Adam Samuel examines its mildly-expressed criticisms and points out an unaddressed conflict in the prudential regulator’s objectives.

Adam SamuelBA LLM DipPFS MCISI FCIArb Certs CII (MP&ER) Barrister and Attorney may be contacted atadamsamuel@aol.com.For links to where you can buy the second edition of ‘Consumer Financial Services Complaints and Compensation’, see www.adamsamuel.com/book.

Zelmer, a former deputy superintendent of the Office of the Superintendent of Financial Institutions in Canada, is clearly a believer that moderate comment is more influential than outright condemnation. His report quietly goes through the Financial Services Authority’s various supervision activities and raises concerns about – rather than says out loud – what most people who followed the demise of the Co-operative Bank think about its regulation of the institution.This and the time-lag between the end of the period under review, 1 May 2008 to 22 November 2013, makes it easier for the Prudential Regulation Authority to say that the world has moved on from the events concerned. It seems startling that Zelmer only started his review in March 2018. More generally, readers of the Sir Christopher Kelly report of April 2014 into the demise of the bank and the Lord Paul Myners report of May that year into the extraordinary governance of the Co-operative movement, might have wished for a more incisive look at the regulator’s role.With very few exceptions, banks generally destroy themselves. This is not the function of regulators. Nevertheless, the public remains somewhat bemused how a bank with a governance structure more suited to running a small grocer shop was allowed to merge with the United Kingdom’s second-largest building society, Britannia, when that institution was in serious trouble. Then, having made a mess of integrating the much larger institution, went on to bid for the ‘Verde’ branches that Lloyds Banking Group was having to sell. A failed takeover bid was never going to destroy the business, at least by itself. It did, however, reveal cracks that ultimately led to a massive re-capitalisation and debt-for-equity swap that valued the original business at zero.The backgroundThe background to the saga, as Zelmer points out, was one of economic stress and regulatory change. Low interest rates made it difficult for banks to recover from the banking crisis and adapt to new capital and liquidity requirements. Following the Northern Rock problem, the FSA had retrained staff, introduced a new supervision model and increased its technical expert numbers to support its enlarged frontline supervisory team. The Co-op was not regarded as having the “same potential systemic impact” as larger institutions, a fact borne out by the limited economic effects of its failure.This disturbed economic climate, in which major banks needed rescuing by the state, created a fear of contagion that in turn threatened the stability of the financial system. It was against this background that the key transaction in the downfall of the Co-operative took place.The Britannia mergerA key moment was the merger of Britannia, the second-largest UK building society and the Co-operative Bank in August 2009. In 2008, the FSA had told Britannia of a range of capital planning problems, a point reflected in the regulator imposing Pillar 2 capital guidance at £424 million – much more than the firm’s original estimate of £120 million. In March 2009, it told the bank of concerns about the institution’s credit risk management shortcomings, the business model’s viability, a failure to tighten lending standards and likely future losses. On 21 November 2009, Fitch downgraded Britannia’s rating.The Co-operative was a poor performing business with a high cost-to-income ratio that needed to expand if it was to develop. Britannia was in trouble. Andrew Bailey – later to run the PRA and now chief executive of the Financial Conduct Authority – considers that Britannia would have failed but for the merger.The regulator hoped that the Co-op’s “funding strength would offset Britannia’s weaknesses”. It encouraged the use of a limited Material Adverse Change provision in the agreement, to reduce execution risk. Nevertheless, it was close to being triggered in July 2009. As Zelmer pointed out, “When one is seeking a safe harbour for a troubled institution, one usually expects the harbour to be larger and more sophisticated than the troubled boat.” He is clear that HM Treasury and the Bank of England wanted the merger to go ahead because of the potential contagion damage to the building society sector if the second-largest institution failed. By approving the merger, Zelmer notes, the FSA was just putting off addressing issues that it knew about. He declines to criticise the regulator for approving the transaction in light of the fear of a Britannia collapse and a general loss of confidence in the building society sector.Zelmer notes that the FSA did not consider the Co-op’s very limited due diligence work on Britannia because this was not standard practice. That is really not good enough. Basically, the FSA was so concerned about the effect of a Britannia failure that it locked the Co-op into the acquisition. Challenge to the Co-op’s hopeless preparatory work would have helped deal with its equally hopeless risk management expertise and governance.The view of the Co-op was that it had limited board skills and poor operational risk management. Nevertheless, the FSA assumed that the Co-op as an existing entity was fit and proper, and continued to be so. “However, there was no apparent consideration of the potential misalignment of the culture or risk management from the high proportion of critical roles that would be filled by senior Britannia personnel,” says Zelmer.Zelmer is adamant that there was not much wrong with the 2009 stress tests done as part of the regulator’s consideration of the merger. He is more critical of the FCA’s use of loan performance indicators that did not take into account potential problems in the future such as refinancing risk – a more pronounced problem for the Co-op than its peers. This is the effect of a deterioration of the underlying assets of the borrowers, such that neither repayment nor the re-financing of the loan at the end of the term will work.The regulator knew of the “vulnerabilities in the merged bank’s balance sheet” and that the bank could need more capital in future. In view of the difficulties that mutual organisations have in raising capital, this might have prompted questions about the appropriateness of the merger.Basically, the slightly crude June 2009 stress test indicated that in difficult 1980s-like conditions, capital would be eroded fairly quickly if management did not do anything about the problem. The tricky part of the calculation concerns the fair value adjustments to the Britannia assets, since they were effectively being bought. In any event, the tests underestimated the payment protection insurance liabilities in the Co-op book. Essentially, though, the FSA had a reasonable idea that by approving the merger, it was taking significant risks. It felt that forcing a capital injection could have ‘spooked’ the bank into abandoning the transaction.The Britannia book was not aggregated by borrower, concealing high levels of concentration around a small number of individual borrowers. This combined with 46 per cent of the loans being in London and the Southeast. The portfolio of subordinated debt was assessed in an inconsistent way and the valuation of the collateral was often not updated. Half of the £415 million book was unsecured. Also, provisioning looked modest, despite some borrowers’ failure to make interest payments in full.Zelmer was concerned about whether stress testing at the time adequately picked up conduct risk impairments and IT write-offs. Such tests are over-simplified simulations that produce a single number with spurious accuracy. Zelmer seems to be suggesting that consideration of a wider range of risks and the incorporation of a broader range of probabilities would be more accurate. The same point could easily be made about product governance stress testing.There was some jigging around with the bank’s capital. Zelmer thought that it was “imprudent” to allow Britannia’s permanent interest-bearing shares (PIBS) to be treated as Tier 1 capital. A further issue was that the tests were based on what the bank was telling the regulator. Operational and mis-selling risks are generally missed.Perpetual subordinated bonds would replace Britannia’s PIBS and be treated as Tier 2 capital after the merger, having been used as Tier 1 for stress testing. Either way, this was “debatable” in view of the fact that many PIBS holders were retail customers.The FSA thus escapes more or less unscathed for its performance over the crucial Britannia merger. The regulator knew that Britannia was a weakened institution and that the Co-op’s management was not properly designed to handle such a transaction. Britannia’s position was a little overstated on its accounts. However, the regulator was under no illusions about what was going on. It needed Britannia to be saved. It was not going to object to the Co-operative Bank doing this on the basis that it would be better if another more professionally-run organisation delivered it.As part of the same concerns, Zelmer asks how the PRA should balance its focus on financial stability damage done by failing institutions against the interests of depositors and creditors. This is the real point about the Britannia merger. It was never clear whether the PRA was protecting the system or the customers or both. One option is simply to impose stricter criteria on firms wishing to carry out significant transactions.Zelmer notes that a merger like the Britannia/Co-op one would nowadays be subject to far more “prudential appraisal”. He is, though, “struck” by how poor the guidance given to supervisors is to assess risks within regulated businesses. Most of the material supplied to staff focuses on “administrative and procedural issues” not on how to assess risks and mitigants used. A merger today would involve greater emphasis on governance, organisational structure and management assessment, operational resilience and operational continuity. Guidance on assessing transactions is urgently needed.However, Zelmer’s main concerns relate to the way in which the FSA regulated the merged institution after the merger.After the mergerHaving sanctioned the creation of a bank with serious concerns about its commercial real estate book and other commercial loans (inherited from Britannia), the FSA runs into much more stinging criticism from Zelmer.Supervisors should have reviewed “aggressively the credit exposures from a valuation perspective as a matter of urgency... to pin down the amount of capital needed and medium term capital raising operations should have been a higher priority for the supervisory strategy”. The supervisory risk assessment letter of May 2010 did not specifically emphasise the bank’s capital issues.The FSA missed post-merger issues with loan loss provisions. It relied too much on the PwC credit book review of 2010 and the bank’s auditors. The regulator should have been reviewing the loan book’s quality as well as valuation and insisted on adequate capital existing to cope with this. The new IFRS 9 accounting standard should promote a more forward-looking view of loan loss provisioning. The existing “incurred loss” approach looks too far backwards. The change towards more predictive work was already in progress at this time. Zelmer feels that the Prudential Regulation Committee and the PRA executive need to ensure that emerging regulatory developments form part of individual firms’ supervision strategies. What was missed were the issues of re-financing, particularly with the Britannia loan book. At the end of the loan term, the collateral was often so damaged that while no payments would be missed, the last payment could not be made to repay the capital. The supervision strategy did not seem to pick this up.Zelmer suggests that FSA supervisors covered a wide range of issues with the bank, which drew attention from the loan impairment and capital adequacy problems.The trouble was that the FSA, instead of doing an in-depth review of the bank’s capital, used Britannia’s former auditors PwC and lost control of the review’s scope. The adequacy of provisions and asset valuations were not on the list to be covered. The 2010 review did not spot any “significant areas of concern”. The bigger issue was that, although the 2009 FSA review had picked up refinancing risk, the regulator preferred to rely instead on PwC and do relatively little.In 2010, the FSA told Co-op to carry a capital planning buffer of £420 million to be drawn down in “extenuating circumstances”. It would – and Zelmer implies should – have been “much larger” if the FSA had not given the bank credit for £1.15 billion in relation to some of the management actions that it would carry out in times of stress. Zelmer “would question” whether management could ever have done these acts “in a timely and orderly fashion”. In September 2010, the FSA lowered the business’s risk rating. This was reversed in the following March.In June 2010, the bank misreported its liquidity position. This generated a section 166 report and further distraction for the regulator. In August 2011, the FSA looked again at the bank’s business model and found very low returns on corporate and business banking because of losses in the back book, low net interest, concentrations in the commercial real estate portfolio and a loan-to-deposit ratio of 233 per cent.Overall, Zelmer criticises the supervision team for failing to focus sufficiently on a few items, such as reviewing the quality and valuation of the loan book, as well as ensuring that adequate capital existed to cover losses. Following the general playbook does not work for firms nearing crisis status. There was also concern that the tone of the regulator’s messages was not always consistent.Accounting problemsZelmer was more critical of the lack of necessary challenge by the FSA of the data supplied by external auditors, the firm or third parties. High staff turnover contributed to this.Like the Kelly report, Zelmer criticises the way in which the FSA failed to question the bank’s change in accounting treatment for its IT platform spending. The Co-op had already started to redo its IT platform well before the Britannia purchase. The bank lent the cost of the project to a sister company, which thus technically owned the project until it was delivered to the bank. This enabled the bank to treat the loans as receivables, rather than consider the payments to develop the system as intangible assets. They have to be ignored for capital adequacy purposes.When the project was written off as useless in 2012-13, there was a sudden significant hit to the bottom line. Zelmer agrees with the regulator that any forbearance at that point would have resulted in a public distortion of the bank’s financial position. Co-op had been quite open about what it was doing all along. There is no evidence that the regulator ever challenged the approach that essentially mis-stated the bank’s capital position. Zelmer was told that supervisors did not have the training to spot an issue related to intangible accounting. He thinks that the accounting treatment might have led to an earlier review of the project and higher provisioning for it. The effect of earlier provisioning would in any event have reduced the impact of the 2013 write-off.Zelmer is concerned that others may not be so forthcoming with unorthodox accounting, creating a risk that the regulator is not focused on the economic reality of such transactions. Greater complexity can itself increase data reporting errors. The PRA needs more formal third-party reviews of key prudential material handed in by banks. The big point here is that the FSA should have been more aware of the economic reality of the bank’s treatment of the project as an intangible and insisted on having the expenses incurred on it properly deducted from the Co-op’s capital when they were incurred. This would have smoothed the hit to the firm’s capital. In an echo of the FSA’s internal audit report on Northern Rock, Zelmer is highly critical about the PRA’s records management. He found it difficult to obtain consistent data across a number of institutions.Heading for a fallThe move to more forward-going assessment of firms in 2012 led to a huge increase in the need for capital. The FSA did not realise the impact that the Co-op Bank would suffer from this. Zelmer recommends that the regulator consider how to help firms adjust to new measures. The real problem was that the FSA did not carry out a proper asset quality review of the loan book until 2013.Zelmer is kinder than the Treasury Select Committee in thinking that the PRA did not need to intervene with the Verde bid. It expressed its concerns about the transaction clearly enough. Unfortunately, Zelmer considers that the Treasury did not appreciate how weak Co-op’s position was until the bid collapsed. The various regulators needed to discuss earlier the various problems with the bid.The regulator did tell the Co-op in June 2011 about the challenges it would face in capital raising, funding, integration as well as governance and risk management. The regulator was warning the Co-op that it did not have what it took to handle the acquisition. It also told Lloyds of this.The FSA put the bank on the watchlist precisely because it was bidding for the banks. Zelmer, though, argues that its placement there was entirely due to the Verde transaction, from which the Co-op had already pulled out in April 2012. In any event, the deal was restructured. However, the regulator did not intervene. At a board meeting in July 2012, two deputy-chairmen, really the only bankers on the board, voted against the transaction. The regulator did not know of this until the end of August.The January 2013 Individual Capital Guidance from the FSA led to a massive increase in the capital planning buffer and a significant rise in Pillar 1 ICG. Zelmer absolves the regulator from responsibility for the Verde fiasco, relying on its warnings to the bank as sufficient. He just feels that the Treasury should have known more about the problems at the Co-op.Both Zelmer and Kelly emphasised the concern that the FSA/PRA was issuing serious warnings to the Co-op board that the directors were not hearing or treating with the same importance. This leads to a point made much of by Kelly but ignored by Zelmer, of how the regulator approved a senior management team consisting of ex-Britannia people and, at the end of the saga, a chairman with no banking experience.Ultimately, the Co-op wasted about £73 million on the Verde bid. It was a bad idea for which Credit Suisse, who suggested it to the bank’s board, should take some of the blame. There was much public  concern about the need to create a further challenger bank. This favoured the Co-op bid. However, yet again, the regulator should probably have stopped a badly-run bank from participating. Zelmer and the PRA make a great deal out of the need for regulators not to run institutions. However, there comes a point when the statutory obligation to protect institutions overrides this.More general problemsLike the internal audit report on Northern Rock, Zelmer criticises the FSA for the way in which supervisors did not file proper records of all firm meetings. He was surprised at how little record there was of engagement between the regulator’s management and the bank. He found it difficult to access information for this report. He was concerned that not all supervisors and senior management used the Risk and Work Manager system to track firms’ outstanding risk mitigation tasks. Regulators need to be able to obtain consistent data across institutions, notably as regards historical fair value adjustments.More generally, Zelmer feels that the regulator needed to allocate supervisory staff and specialists more carefully when risks were crystallising. He found a lack of adequate handover procedures when people leave.His final concern is that if economic conditions remain benign, prudential oversight may fade from view in favour of macroeconomics, a subject closer to the heart of the Bank of England. The curious feature of this comment is that it fits the concern expressed by George Osborne and others that conduct issues were placed ahead of prudential problems in the old FSA.The regulator’s replyThe PRA and Bank of England inevitably replied to Zelmer’s report agreeing with its recommendations. On the balancing of interests, the PRA demonstrates its own confusion by pompously indicating that the regulator’s primary objective is to promote the soundness of firms and to achieve this by focusing on the UK financial system’s stability. It misses Zelmer’s point entirely about handling situations where there is a conflict between the two objectives.Since 2012, the liquidity buffer has required the assets to be unencumbered. The regulator, though, rejected the use of encumbrance limits. The PRA will be participating in the Bank of England’s review on the extent to which encumbrance should be considered in recovery and resolution planning. While accepting the recommendation of more formal third-party reviews, the regulator explains that these were stopped because they were thought to be disproportionate. The Bank of England rejected the audit of regulatory data in 2013.So, what really are the lessons learned?This strangely is the tricky part. The Co-op should never have been seduced into saving the building societies sector by buying Britannia. The bank’s board composition before and after the merger was hopelessly ill-adapted to the management of a financial institution. Yet, the sector has remained intact. Should vast amounts of the Co-operative movement’s capital have been squandered on this endeavour? Nobody forced or even asked the Co-op to bid for Britannia. Its management made this disastrous decision.After the acquisition, the FSA ­– not unreasonably distracted and stretched by other concerns – failed to pin down properly the real financial condition of the institution or the state of its desperately weak board. Nevertheless, numerous other bank failures illustrate that good board composition will not necessarily save an institution. Maintaining Britannia’s leadership in place was probably a condition of the merger. However, the Co-op was in a much stronger position to insist on the appointment of a proper banker CEO. The process by which it installed the Reverend Paul Flowers as chairman, documented in the Kelly report, suggests that it was never able to do such a thing.The regulator expected too much of a bank initially run by a man who had previously come close to ruining two-thirds of it. Zelmer is right to think that the FSA failed to understand the Co-op post-merger business and work with the bank to sort out accounting and capital issues. Even after the 2010 general election when prudential regulation was split off from conduct regulation, there does not seem to have been much of an improvement of focus. This appears to have been reflected in a general sleepwalking attitude of the Co-op board to the same issues. It decided on the Verde bid. This had some attractions, notably the acquisition of a Lloyds Group IT platform to resolve its difficulties in this area. Maybe a benefit of the bid was that it flushed out the imperilled state of the bank’s capital.Perhaps the best and last lesson comes from the PRA’s muddled response to the request for some clarity about its objectives: that it is required to secure bank safety by focusing on the UK financial system’s stability. All bank shareholders and stakeholders need to understand the conflict at the heart of prudential supervision between protecting creditors along with depositors on the one hand, and the financial system on the other. The latter objective has been used here to throw the interests of the former under the proverbial bus.