Description

Long Metro Bank (MTROLN) bonds: 9.5% senior unsecured due 2025 at 75 offer price (16.4% YTM, 350mm outstanding), 5.5% subordinated due 2028 at 38 offer price (22.5% YTM, 250mm outstanding)

 Metro Bank is a £22bn asset UK challenger bank, which is a fancy way of saying that it is relatively small community bank that resembles a UK building society or US thrift: the business model is taking retail deposits and making mortgage loans. While MTRO briefly commanded a growth stock mantle for a few years post IPO, after a series of regulatory mishaps the company has arguably become the most despised bank in European financial analyst circles and the stock trades at option value. The equity has been written up twice on VIC previously in 2018 by jon64 and wrangler, and the comment section of the latter writeup has a very detailed history of what went wrong.

 The main point of this writeup is to argue that regardless of whether or not one thinks Metro will ever generate an acceptable return (8-10%) on its common equity, both the senior and sub bonds are covered at par due to limited asset quality issues and the underlying profitability of the customer relationships. In other words, the bonds are covered by the value of the business to an acquirer in M&A, which is the most likely outcome of a resolution process in the unlikely event that one should occur. Both securities are issued out of the bank’s operating entity (as opposed to a group holding company) and would “travel” in the event of a sale to become obligations of the acquiring firm, a home-run outcome for these securities given the trading levels of debt issued by every other UK bank (1% to 4% yields), who comprise the likely universe of acquirors. The bonds are orphaned securities without published research coverage, index inclusion, or sufficient tradeable market value to warrant the attention of large distressed debt firms. However, given current trading levels the securities offer significant equity-like upside. The bonds trade at distressed pricing because of market concerns around the potential implications of the bank falling below interim MREL regulatory requirements at year end 2020 and the bank’s current lack of pre-tax, pre-provision profitability (PPP). The MREL breach effectively forces the regulators to make a decision in the near term on whether or not to grant the bank additional time to fix itself. Either the bank receives regulatory forbearance on MREL and/or AIRB approval next year to gain time to restructure itself, or the regulators force the sale of the bank in 2021 – the bonds are covered at par either way.

Some assertions to start:

• Banks become insolvent because of losses on assets, typically on bad loans although investments in subprime structured securities in 2008/2009 were a notable additional source of failure in Europe. Regulated depositories almost never fail because of liquidity issues, due solely to their ability to access central bank funding lines as an effective last resort in the event of an emergency. We make this distinction because corporations, including unregulated financial institutions (a.k.a. non-bank lenders), have historically failed on a regular basis due to liquidity issues. Emergency funding allows the banks and regulators to work out an orderly solution, such as an asset sale, rather than a fire sale. Hence credit analysis for depository institutions generally focuses on asset quality rather than common corporate credit topics like debt maturity schedules, liquidity runways, etc.
• Banks have protected creditor classes that are unique to financial institutions, namely retail depositors. Given the panics in the general public that can occur when deposits are perceived to be at risk, after 2008 regulators have placed troubled banks into a “resolution process” which are quietly managed by the regulators themselves rather than undertaking corporate style bankruptcy. A typical bank XYZ which finds itself in resolution generally over-extended itself during an expansion, made loans to borrowers who were unable to repay (generating losses), and now finds itself short of loss absorbing equity capital.
• Resolution tools include the sale of the business to a private purchaser (possibly with the use of a good bank / bad bank split), separating out assets into an asset management vehicle, and bailing-in private sector creditors to prevent government funds being required to recapitalize the institution. M&A is overwhelmingly preferred as an option for resolving the performing parts of the business, as it can be accomplished over a weekend with limited disruption in the public eye. Regulators are sensitive to political concerns which means that the idea of “simply liquidating” a bank with individual retail deposits is a nonstarter.
• The combination of performing loans and zero cost deposits (e.g. checking / transaction / current accounts) has always been profitable due to positive excess spread in investment grade credit. Non-performing loans (e.g. loans that cannot or will not repay) are the source of losses because the “loss given default” is typically a large multiple of the income that had been expected from the associated interest and fees. There is a huge debate about the degree of profitability from non-interest bearing deposits in a zero-forever interest rate world because of low returns on risk-lite assets, but 0% funds are still profitable. As a result, there is nothing inherently unprofitable or bad about Metro’s balance sheet composed of 60% LTV residential mortgages and retail deposits.
• However, Metro Bank is currently unprofitable, and is expected to remain unprofitable in 2021 as well, due to ongoing operational remediation and restructuring costs called “investments” by the company (albeit improving on 2020). The consensus view is that the bank will always be unprofitable, and is therefore spiraling towards an insolvency process that results in impairments on the bonds. Unlike other financial sector resolution examples from history, the consensus view in Metro’s case is that the problem is neither asset quality nor a lack of liquidity but rather that the company has a catatonically stupid business model –it attracts customers with physical retail branches – incapable of generating profits, a self-evident truth available to everyone but management (and, apparently, us). 
• It is by now obvious that the deposit-led growth strategy employed by prior management does not generate acceptable returns on equity in a zero-forever interest rate world. However, the problem that Metro faces today is not that the fact that it pays ~26mm per year in branch lease costs on a revenue base of ~375mm, but rather that it lacks minimum scale to cover back office expenses. Prior management held a tour-de-force in how not to run a start-up business with regards to cost management, and the bank todays finds itself in the position of a turboprop engine coupled to a 787 airframe. 
• A number of analysts confidently propose that no sane bank would acquire Metro on the basis of branches being an outdated distributed model. It is true that Metro historically invested a lot of money into its branches and capitalized that spend on the balance sheet as real estate value. As a result, Metro has struggled to earn a decent return on the historical sunk costs, but this is also largely irrelevant as the market has long priced Metro’s equity at a small fraction of book value. However, the reality is that the running costs of the branches are very low relative to the income generated and the potential 40%-50% operating cost savings that could be realized by an acquiror are very attractive financially. Frankly, the rental income from safety deposit boxes alone in the branches covers more than half of branch lease costs (Metro is the only high street bank to offer them). The issue has been that the back office operating expense burden is so large relative to the quantum of revenue generated at 0% interest rates. As a standalone company, Metro needs time to grow into its back office for management to try to realize the option value of the equity. If the company is sold the profitability would be dramatically higher after cost synergies but the majority of the upside would likely accrue to the acquiror and not current equity holders.
• There are only two outcomes for Metro Bank at this point: either new management will be successful in selling additional financial products to its existing customer base to drive more revenue off the existing back office cost structure (their so-called “J-curve” turnaround), or management will need to sell the business to a competing financial institution who will close down the duplicative back office overhead. The debate about Metro surviving on its own or not is primarily about how much value can be realized by public equity holders, and regulators will make the final call on that topic. If the turnaround is successful the bonds should re-rate to par over time and if the bank is sold to a larger peer the bonds should re-rate to par much more quickly as the bonds are issued at the operating entity and will become obligations of the acquiror. 

Much of Metro’s current predicament is both highly unusual and the product of its recent history. Hence, we include a brief overview to explain how the company got to where it is and what the likely path forward is.

Metro Bank: pipe-dream business plan or creating a customer-service-focus niche in notoriously stodgy UK banking?

The original idea: the founder of Metro Bank was Vernon Hill, a controversial but successful American financier who founded and sold Commerce Bank in the US for billions in 2007. The Commerce model was “deposit-led”, which meant making bank branches as appealing as possible to retail and small-business customers in a given geography (NY/NJ) to capture their business in zero-cost transaction accounts. In order to attract retail deposits he built branches that were overtly focused on being warm and customer friendly rather than the typical cold, institutional, security-driven branch experience seen elsewhere. The insight was that the extra costs at the branch level of being open 7 days a week with longer hours of operation, having change-counting machines, dog friendly environments, etc were relatively small compared to the attractiveness of those features in the mind of an average consumer. After taking receipt of checking accounts paying 0%, Commerce would then invest the funds in the lowest risk asset classes possible to earn a return based primarily on the deposit margin with a little extra pickup on duration and/or secured credit spread risk (e.g. mortgages). The key to this model was that Commerce didn’t need 200 loan officers and a huge credit risk department to invest the balance sheet this way, only a couple of fixed income portfolio managers with a Bloomberg. The lack of asset-side operating expenses more than compensated for slightly higher branch costs, resulting in good efficiency ratios / operating margins. US banks like PB, FFIN, and WABC run a similar strategy of low-cost deposits into bonds today.

Metro 1.0 from 2010 – The original Metro featured the same business plan as Commerce Bank but located in the UK where popular distrust of the main “high street” / money center banks was even higher than in the US stemming from widespread UK government bailouts in the GFC. Metro’s self-chosen comparisons with Apple (“omni-channel retailing excellence”) were difficult to stomach but they actually got a number of things right about attracting consumers and local small businesses specifically. Mainline UK bank analysts hated the strategy from Day 1 as it was predicated on opening new branches (which the mainline banks were shrinking after 200 years of expansion) and investing in customer service (which they were also shrinking in an effort to cut costs). Yet the reality is that Metro rapidly grew its customer base because it tapped into an unmet segment of the retail market that wanted high touch service with a sense of “physicality” to the concept of safety in their savings: Metro was the only bank to offer old school safety deposit boxes on the market, and rental revenues from those units covered 80% of base store lease expenses even in downtown London on 2012 vintage retail rent levels.  Metro’s main flaw was they completely forgot to control costs and maintain operational discipline in their wildly expansive growth plans. Management seems to have been so consumed by some long lost sense of Manifest Destiny regarding their inevitable success in “disrupting” the mainline banks that they lost the financial discipline that made earlier ventures like Commerce a success. The placement of the IT and accounting department into a flagship City of London office building two blocks away from Goldman Sachs (the “old bailey lease”) was probably the pinnacle of hubris but also indicative of their approach to running the firm.

The Fall:  Metro IPO-ed in 2016 and the bank’s growth was quickly rewarded with the highest multiple in European financials. Yet, this growth required a steady stream of new equity capital and Metro’s fundraising efforts with public investors in 2017 and 2018 were always linked to aggressive growth targets and high returns to come in 2020 and beyond. Unfortunately for management, the operating environment for UK banks became increasingly competitive over that time frame. The UK’s “ring fencing” regulatory structure trapped excess liquidity in the domestic market, causing the main 6 banks to aggressively compete on price for new mortgage lending. Rather than walk back previously communicated targets, management instead attempted to out-grow declining unit economics in the mortgage market via an expansion into commercial relationships (merchant acquiring, treasury management, low LTV commercial real estate lending, etc). Commercial loans are significantly more complicated to originate and service than basic residential loans. So, while Metro did not over-extend itself on credit quality in its commercial expansion, it did on the operational costs side of the business. Given continual budget constraints and the need to simultaneously grow quickly, Metro seems to have to made a number of operational choices favoring short-term fixes (hiring lots of people for manual processes, particularly on the loan processing/operations side) rather than longer-term efficiency such as implementing software to automate those processes. As a result, back office operating expenses ballooned to nearly ~200mm by the end 2018 (55% of the total).

Finally, at some point during 2018, fateful choices were made which resulted in the misstatement of risk densities on commercial real estate loans. Risk density is simply the assigned risk weighting (0% to 100%) per £ of principal value that a financial asset holds for regulatory but not IFRS purposes and is an internal designation. Government bonds are typically 0%, the most risky loans are 100%. The core premise is that not all loans are equally risky from a probability of default, and hence from a regulatory perspective greater leverage should be allowed on a low LTV prime residential mortgage on the one end versus a subprime credit card on the other. Hence, in the UK, a low LTV mortgage on residential property may only carry a risk density of 15% while a student credit card (or leveraged loan in the commercial context) would carry a 100% risk density. The core issue with a topic such as risk density is that the regulator is fundamentally beholden to banks to report truthful information on loans as part of the supervisory process; it is a matter of trust between the regulators and the banks that risk characteristics on loans are presented accurately.

It appears Metro either intentionally or mistakenly applied certain risk models to a number of commercial real estate loans that resulted in their having a 50% risk weight rather than the appropriate 100% risk weight, with the result of overstating capital ratios reported to regulators. It also appears Metro intentionally or mistakenly classified certain buy-to-let investment properties as “retail” rather than “commercial” with the same effect – although in Metro’s defense some of the distinctions are a bit technical (does a single individual who owns and rents out 4 detached houses as investment property for their primary livelihood qualify as a ‘retail’ borrower or as a ‘small business’ commercial borrower?). Did Metro’s management bend the truth to report slightly higher capital ratios so that equity market perceptions on growth capital needs would be lower or did they just make an honest mistake? We’ll never know the full story.

Regardless, the risk density misstatement represented a major and fundamental violation of trust with the regulators, who promptly extracted their pound(s) of flesh via enforcement proceedings, fines, and conduct investigations. The combination of regulatory investigation and accounting misstatements triggered a small-scale run on the bank in early 2019, with a decent portion of the newly acquired commercial customer base leaving the bank in the span of a few months (around 50% of total commercial deposits). Metro was forced to undertake an emergency rights offering in 2019, sell a profitable loan portfolio, sell a profitable securities portfolio, and raise high cost 12-18-month fixed term deposits to stabilize the funding base. Regulators then ousted the entire management team (including the founder) affiliated with the financial restatement and brought in a new executive with a history in restructuring and existing relationships with UK regulators.

The turnaround / How is this J-curve thing supposed to work?

When the new CEO Dan Frumkin took control in late 2019, his first task was primarily to restore relationships with customers and regulators, then figure out a way to earn acceptable shareholder returns. Metro Bank historically invested a massive amount of capital into the branch network, and impairments on the capitalized value of real estate & leasehold improvements in particular are the main risk to the book value of equity. The capitalized leasehold improvements are sunk costs, which is what management is referring to when it says its costs are fixed – Metro has to both “sweat” the existing customer relationships harder (more revenue per customer) and increase the amount of customer relationships per branch in order get a better return on the existing equity. The running / operating costs of the branches are actually relatively low, but the capitalized investment cost on the branches (and the equity attributed to them) are high and fixed.

To increase revenue from existing customer relationships Metro has to increase the number of products sold to that customer. Customer profitability for any bank anywhere is 100% correlated with the amount of product cross-selling. In 2020 to date Metro acquired a consumer credit scorecard for unsecured lending, entered into the specialist mortgage market, and developed multiple micro/small SME products like receipt management, cash management, tax management and accounting. Building out these product capabilities costs money, which is a large portion of the £250-300mm cumulative “investment spending” that Metro guided to in their February 2020 investor presentation. Spending that money also generates operating expenses on the income statement, which results in accounting losses and capital ratio decreases during the first two years of the five-year plan, hence the “J” curve shape of the turnaround. The problem with the J-curve is that COVID dramatically altered the operating environment in 2020. COVID has not only reduced customer activity (which lowers loan originations and fees) but also greatly increased the probability of higher ultimate loan losses / charge-offs. As a result, the J-curve is deeper than originally planned and has placed Metro in the position of being in violation of interim MREL regulatory capital ratios.  MREL is short for “Minimum Requirement for own funds and Eligible Liabilities” and is a new regulatory standard in Europe at the senior unsecured level. Please see the regulatory capital notes in the appendix for background on the regime.

What happens from here?

Regulators could grant AIRB approval: buying the company time to execute the turnaround, which either works (in which case the bonds gradually rerate) or doesn’t (in which case the company will be put up for sale).

AIRB means “Advanced Internal Ratings Based” approach to risk weightings for regulatory capital purposes, as distinct from the standardized approach. Capital ratios are the ratio of regulatory capital / risk weighted assets (RWA). For Metro, RWA are driven by credit risk, which means the risk of default on loans outstanding. For a UK residential mortgage under the standardized approach, the RWA on £100 in loans outstanding is £35, for a risk density of 35%. Given that the risk density is the same for any UK mortgage loan, the approach implicitly assumes that all residential mortgages in the portfolio have the same credit riskiness. This is obviously not true in practice, as differences in LTV, payment/income ratio, customer payment history, etc mean that some mortgages are much less risky than others. AIRB is a data statistics exercise where banks “prove” to the regulator that their internal credit rating models can accurately capture the riskiness of a loan, based on how the credit ratings correlated in history with actual, realized defaults. This allows banks on AIRB models to use lower risk densities – about 15% on average for the banks with whom Metro competes. AIRB also enables Metro’s competitors to earn a higher return on equity because they hold less regulatory capital per £ of loan outstanding.

A bank has to get regulatory approval to use AIRB, and Metro has been working on its application for 4+ years at this point. The last public commentary from the company to investors was not to expect model approval “before 2021”. AIRB model approvals are granted on a portfolio by portfolio basis, and Metro has applied for approval on the residential mortgage portfolio – in other words not the commercial or buy-to-let loans which were the source of controversy under prior management. The bank has reportedly spent as much as £50mm on consultants and data scientists over time to check every box on the application.  Therefore, when faced with a shortfall on MREL capital ratios due to COVID-related disruptions, the path of least resistance for the regulators would be to grant Metro model approval which would overnight create ~250bps of capital ratio and buy the bank time to finish the J-curve turnaround. This path is not only the easiest one for the regulators to take (approve something that they were likely to do at some point anyway), but it is also the most equitable, as Metro is currently at a distinct disadvantage versus its competitors due to the lack of AIRB regulatory approval to date. The 250bps of additional capital are more than sufficient to bridge across the bottom of the J-curve until the investment spending is done and the bank regains profitability, allowing the bonds to rerate over time.

Regulators could deny AIRB approval: on a temporary or a permanent basis, either way a de facto order to commence a sale of the business.

Regulators could deny AIRB but say “just wait another 12 months”. In this case Metro would likely increase both its CET1 and MREL ratios by selling high risk density assets like commercial term loans (~100% risk density). While this would clearly increase losses on the income statement in 2021 by removing interest income, the impact could be mitigated somewhat by originating more 90% LTV mortgages with 35% risk densities at a similar coupon. Metro would likely also slow the pace of “investment” to bolster income statement profitability in 2021/2022 at the cost of further pushing out its ROTCE target timeline. However, a delay on AIRB at this point after years of applications and tens of millions spent implies that regulators effectively would be telling management to wrap it up and pursue a sale. To quote Lindsey Buckingham “If you don’t love me now, you will never love me again.”  Metro could try to tough it out for a few years but without AIRB it will never earn a decent RoE given competitive disadvantages. Hence, we think management would choose to sell to a competing bank that would simply use its own model with lower risk weights on the same assets. Per the analysis in the table below, we believe the bonds are covered at par in a sale.

Regulators could also deny AIRB but do so in a market environment that is so bad that selling commercial term loans would be capital destructive rather than enhancing given huge discounts to par on sale. In the table below we model the customer facing P&L including the costs of both the subordinated and senior unsecured bonds, and argue that not only are both covered at par but that residual value exists for the equity. Given that this would be a distressed asset sale, we assume that the buyer requires that the seller pay for all back-office restructuring costs, writes down all capitalized software investment to zero, writes down investments in the branches, and discounts the loan portfolio sufficiently to cover 5 years of cumulative loss under the Bank of England’s stressed loss scenario (similar in magnitude to 2008-2010 realized losses for NWG/LLOY). The conclusion is that the business is sufficiently profitable post cost synergies and the 5-year stressed loan loss content is sufficiently low that even a dirt-cheap sale price does not impair the bonds. To note, the bonds cannot be impaired outside of a formal regulatory-driven resolution process, and the principle of “no creditor worse off” signifies that regulators would pursue a sale well before employing other tools like write-downs / bail-in that would be subject to legal challenge.

A less likely possibility is that regulators do not grant AIRB approval but decide to just push MREL compliance out into the future for either Metro specifically or for the whole industry. As an example, the regulators just granted another mortgage-focused challenger bank (One Savings Bank) until July 2025 to reach full MREL implementation with an interim requirement of 18% RWA by July 2023 - targets that Metro could reach without any remedial action. MREL calculations are determined on a bank-by-bank basis rather than using a single rule for the industry, and there is no particular reason why a longer implementation timeline would not available to Metro considering the efforts the bank has already made to satisfy regulators (375mm forced equity raise in May 2019, 350mm forced MREL raise in October 2019 to prove market access, and then replacing all management including CEO and Chairman). Furthermore, the entire MREL regime is under review with an update expected in December. The threshold for MREL applicability in the UK is a balance sheet greater than £15bn in size. That threshold is an outlier in global banking regulation as almost every other country specifies bail-in regulation for institutions with at least €100bn in assets – including every other country in Europe. Community banks are obviously opposed to MREL and their argument goes that subjecting small banks to the same regulatory burdens as global systemically important financial institutions inhibits competition since the smaller banks don’t have the operating scale to absorb the MREL costs – which is 100% accurate but possibly irrelevant. The Bank of England, in theory, wants to promote competition from so-called challenger banks and the current MREL thresholds directly inhibit that. However, we believe that AIRB is the more pressing issue for Metro’s long-term economics, so if regulators waive MREL requirements but do not grant AIRB, then Metro still should pursue a sale to maximize value for the equity holders.

Summary:

Metro Bank today is essentially two parts – a small profitable challenger bank focused on the residential mortgage and SME market hitched to a very large back-office organization currently sized to handle a business multiples of Metro’s current scale. As a standalone entity, Metro’s P&L suffers from its subscale status and legacy of poor planning on infrastructure costs. As a pure challenger brand targeting a distinct niche of the market, Metro would be quite profitable to an in-market acquiror with existing back office infrastructure in place. Hence the topic of Metro’s profitability is very much a state-contingent discussion: on the one hand as a standalone entity many of the operating costs are fixed in the short term as the business cannot operate without back office infrastructure like IT and loan processing. On the other hand, as a sale candidate, almost all back-office costs are duplicative.  While it is true that the “core bank” profitability has been reduced by the events of 2019 and COVID, it has not been permanently impaired.  The high cost fixed term deposits raised in 2019 will have run off by the end of 2020, increasing net interest income.  Metro’s core retail and SME deposit growth has actually continued in 2019-2020, and fee activity should recover with economic activity. Finally, mortgage origination spreads have increased significantly in 2020 such that the unit economics of Metro’s core business are improving now after 4 years of decline. Our best guess is that regulators grant the bank time, the turnaround eventually works, and the bonds rerate from current distressed levels – but if it doesn’t then we think the outcome is a sale of the business, which is frankly a much better outcome from a bondholder perspective.

Above is our attempt to quantify economics for a bank acquiring Metro in a distressed sale scenario. We assume that the acquiror builds in discounts on loans equivalent to the Bank of England stressed loss scenario, which also approximate industry losses seen in the 2008 GFC. It is important to note that we think Metro’s actual loan losses in the next several years will be a fraction of this amount. We also assume the acquiror writes all software investment to zero and impairs branch investment by 50%.  Finally, we assume in the purchase price that the sellers bear the full restructuring charge for the bank’s back office. Even then, there is still residual value for the equity without impairing either bond. Since Metro’s customer proposition is a high-touch, omni-channel experience, we do not assume lease break costs because shutting the branches would result in customer loss. Also, given the profitability of the branches it is unnecessary. This acquisition would be very accretive on an earnings basis to any acquiring UK bank but capital ratio dilutive (given the amount of write-downs assumed up front) but as major UK banks have excess capital positions with limited investment opportunities, we don’t see this being a problem

Stressed loan loss detail:

Metro is very unlikely to realize the stressed loan losses in the table above. The stress tests on which the losses are based assume 33% drop in UK home prices, which seems unlikely as home prices during COVID have actually risen. It turns out negative real interest rates globally have turned housing into a de facto store of value for consumers in a way that wasn’t present during 2008/2009. On commercial loans, Metro is protected by multiple sources of repayment capacity: the commercial real estate on which the business operates, the working capital of the business, the cash flow generated by the business (if any), and a personal guarantee from the high net worth owner of the business. Some or all of the above may prove to have diminished in value but loss given default is likely to be better than the credit exposure underlying the historical industry statistics.

Common objections:

The regulator has bail-in power to write-down the bonds, so they will just do that. It is true that banking resolution law in the UK permits the regulator to unilaterally write down the value of capital instruments such as the LT2 and the MREL bonds, thereby increasing equity in the bank. We would first note that the relevant regulatory breach is at the MREL / total capital level, so writing down existing MREL bonds to generate more equity does not solve the issue of noncompliance with MREL ratio – it simply shifts the composition of the existing capital stack to more equity without increasing the total. We would also note that the “no creditor worse off” legal principle means that a unilateral write down by the regulator would be subjected to legal challenge, as it is overwhelmingly likely that creditors would be better off in a regulatory proceeding involving a sale of the business. As a result, the regulator would first try to sell the business as opposed to starting with bail-in / write down powers.  

Metro Bank will never make money:  The argument goes that Metro will be granted forbearance on MREL, but then proceed to lose so much money over the course of 2021 and 2022 that it will be forced to equitize the LT2 bonds in order to preserve the CET1 ratio. Management’s internal CET1 minimum is 12% and the regulatory minimum is 9.6% versus the last reported statistic of 14.5%. If Metro does get to an uncomfortable position with the CET1 ratio, management can either sell high risk density assets, pursue a voluntary liability management exercise (LME) on the LT2, or try to sell the company. The LT2 cannot be impaired or “turned off” outside of an insolvency process, so management cannot force holders to participate in any voluntary exercise. Instead, they could tender for the bonds with new equity at a significant premium to bond trading prices to incentivize arb accounts to participate (buy the bonds, short the equity to lock in the tender premium) – which would be wildly dilutive to equity holders. Our guess is that Metro would first attempt to sell commercial loans with high RWA and then try to sell the company to maximize equity value before diluting shareholders with an equity LME.

We also think it’s important to consider exactly why the company is expected to lose money – the £250-300mm of “investments” through 2024. It is true that the majority of investments made in 2020 were nondiscretionary and noneconomic in nature: customer remediation charges, regulatory remediation charges, expanding anti-fraud and anti-money laundering capabilities, and consulting spending on the AIRB project, all of which could be considered the price required to achieve AIRB approval which obviously may or may not happen. However, other investments such as exiting the old Bailey lease (3-4 year payback), purchasing a consumer lending scorecard via RateSetter, and building out fee generating services for the SME market will pay off in the future via higher income and lower costs. As a result of the efforts made to date in 2020, we believe the bank has a lot more flexibility in the timing of its investment spend going forward. The RateSetter acquisition alone obviated the need to spend tens of millions on internal development of an unsecured lending consumer scorecard that would have otherwise been incurred next year. We find it unlikely that management will sleepwalk into a CET1 ratio problem as a result of, for example, hiring IT consultants to build new software in 2021 instead of 2022.

This situation is just like Co-Op bank, and the LT2 were equitized in that example: A recent precedent in distressed UK banks is the Co-Op bank, a non-listed mortgage lender that was formerly part of the Co-Op group, whose primary business is supermarkets. In early 2017 COOP reported an enormous loss for 2016 and guided that it would be in breach of required common equity ratios (Pillar 1 + Pillar 2A) not only in 2017 but also for four years following. The scale of the loss and low ongoing profitability of the bank caused their LT2 bonds to drop to 50% of par. The regulators immediately demanded that the parent inject additional equity capital to cure the breach or else pursue a sale of the business. The COOP management estimated that 700mm of additional equity would be required vs then-current equity valuation of ~56mm. The bank pursued a sale process, found no bidders, and the LT2 bondholders ended up owning the company as part of back-stopped rights offering led by Goldentree and Silver Point. The reason that the bank was not saleable was that it had numerous issues which no other UK bank wanted to take on: a large unfunded pension liability which was increasing due to falling gilt rates, a large PPI claim liability that was difficult to cap, and the lack of a modern IT system as a result of having underinvested in systems for the previous 100 years. Importantly, Metro suffers from none of these issues due to its having been founded in 2010 – it never had a defined benefit pension, it never sold PPI, and it has arguably among the most modern IT infrastructure of any major UK bank (it doesn’t have mainframes running COBOL somewhere in the system). As a result, we do not think that the Co-Op Bank analog is a good argument for why Metro couldn’t be sold.

Stylized regulatory capital stack & Metro’s specific capital stack:

The 5.5% bonds are in the Tier 2 bucket, while the 9.5% MREL bonds are in the senior non-preferred bucket.

Above is Metro’s regulatory capital structure as of 6/30/20 as well as a stylized bank capital structure with priority ranking in resolution (the top is first to absorb losses). The most serious ratio is the required common equity tier 1 ratio, as common equity is loss absorbing on its own (e.g. without regulator intervention or a legal process). If a bank loses money in a given year, common equity is reduced in retained earnings without the regulator doing anything – equity is considered “going concern” loss absorbing capital.  Pillar 1 capital minimums are the bare minimums and in practice banks never approach them. Pillar 2 is part of the supervisory process where the regulator makes judgements about a given bank’s risks above and beyond the minimum capital requirements as part of ICAAP (internal capital adequacy) and specifically annual stress testing exercises. Hence the level of Pillar 2 capital required can vary by bank. Common equity and equity-like instruments such as AT1 bonds (CoCos) satisfy the overwhelming majority of Pillar 1 and Pillar 2 requirements plus any additional regulatory buffers like capital conservation buffers, counter-cyclical buffers, domestic SIFI buffers, etc.

Senior to common equity and AT1 bonds are Tier 2 bonds and non-preferred senior unsecured bonds (NPS), which are considered “gone concern” capital. Metro’s 5.5% issue is a Tier 2 bond while the 9.5% bond is NPS that qualifies for MREL. Gone concern capital signifies that the bonds are designed to bear loss in the event of a formal insolvency process in order to protect depositors, but not outside of such an event. To my knowledge, shortfalls in gone concern capital ratios have rarely, if ever, have served as a source of regulatory intervention historically because a deficiency in common equity ratios almost always occur first. The distinction between NPS and senior preferred is basically to designate NPS as specifically loss absorbing for purposes of meeting MREL requirements while senior preferred is not.  

MREL rules were formulated in 2016 with the memories of the 2008/2009 banking crisis explicitly in mind. During the GFC, governments recapitalized banks at the common equity and preferred equity levels given the belief post Lehman that allowing “hard defaults” which impaired senior unsecured creditors would trigger a cascading wave of bank defaults due to the interconnectedness of the institutions via derivatives. The core issue was that derivative counterparties and commercial depositors were legally pari passu with senior unsecured bondholders. Equal seniority under law made it difficult for governments to legally haircut bondholders in bankruptcy while leaving other creditors intact whom they wished to protect - specifically derivative counterparties like other banks as well as corporations, states, and local governments. Hence government support for failing institutions occurred at the common and preferred equity levels with the consequence of “bailing out” senior bondholders who otherwise would have been facing losses.  The MREL regime was therefore designed to create a class of specifically loss-absorbing senior unsecured creditors that avoids this problem in the future.  MREL levels globally have converged to around 20-30% of risk weighted assets, which regulators believe, rightly or wrongly, is sufficiently high to prevent national governments from ever having to recapitalize a financial institution again. The reality is that the complicated system of pillar 1, pillar 2, and add-on buffers currently in place requires banks to retain a level of equity capital that is multiples higher than the levels witnessed prior to 2008. So, while senior unsecured MREL bonds are now explicitly loss absorbing by law (whereas they were not beforehand) the actual risk of default at the MREL level is arguably the lowest it has ever been. 

Catalyst

Sale of the business