ABN AMRO Financials Outlook 2018 – Fundamentals to trump the QE slowdown

by: Tom Kinmonth

  • The positives far outweigh the negatives for European bank credit in 2018
  • ECB will reduce overall asset purchases, but covered and corporate bonds to be maintained at current levels and continue for the whole year in our view
  • Meanwhile, bank fundamentals are strong and improving with larger capital…
  • …and upcoming MREL regulation should further bolster balance sheets
  • Finally, the economic recovery is returning, which will support profit growth
  • Against this background, we have the following key conviction views:
  • AT1s: Overweight short-dated AT1 paper, the jewel in the bank debt crown…
  • … although the long-end AT1s are not priced for extension risk
  • Tier 2: Underweight, especially mid-to-long duration names
  • Senior Non-Preferred: Overweight, especially on strong GSIB names
  • Traditional Senior: Underweight, but still interesting versus other asset classes

DISCLAIMER: This report has not been prepared in accordance with the legal requirements designed to promote the independence of investment research, and that it is not subject to any prohibition on dealing ahead. This report is marketing communication and not investment research and is intended for professional and eligible clients only.

ABN-AMRO-Financials-Outlook-2018-Fundamentals-to-trump-the-QE-slowdown.pdf (369 KB)

After an absolutely stunning year in terms of performance for European bank debt, we take a look at what the future holds for 2018. We tread carefully into our selections for 2018 and aim to strike a balance between the strong improvement in fundamentals across both the banking sector and economy versus the gradual easing of the ECB’s monetary stimulus.

The fundamentals of banks have improved, a trend set to continue

Starting with the fundamentals, banks are entering a New Banking World. The banking landscape has changed tremendously over the last four years, and bank balance sheet improvements will continue for the next few years. Almost every measure across bank balance sheets has improved.

The strongly enhanced levels of equity capital and the structural de-risking of bank balance sheets have generated strong buffers to protect bank debt. Crucially, the improvement observed in balance sheets is not over, and the regulatory focus will still enhance fundamentals to the benefit of the banking sector. For example, the goal to create a European deposit guarantee scheme (EDIS) should decrease systemic and idiosyncratic risk by the reduction of Non-Performing Loans (NPLs) and sovereign holdings, as we discussed here.

Indeed, regulation has already had a large role to play in the story, and has had a profound impact on the bank fundamentals. To name just a few leaps that have been made; the application of the bail-in legislation (BRRD), a pan-eurozone supervision entity was formed (the SSM), and the creation of a new debt class in the bank creditor hierarchy.

The macro landscape arouses the gradual winding down of stimulus

What is more, a strengthening economy will also be constructive for the banking narrative in 2018. Finally, the bank fundamentals will be supported by a return of strong economic growth in Europe. The eurozone economy has been growing above the trend rate in the past four and a half years and growth is expected to remain above the trend in 2018-2019. Further details can be found in our eurozone 2018 outlook, here.

Overall in Europe, there still exists significant slack in the labour market, however, wage growth will probably pick up modestly in the course of 2018. Combined with the positive growth outlook, this should give the ECB enough confidence to reduce its stimulus very slowly following the end of its current asset purchase programme in September 2018. Indeed, we expect an additional period of purchases for an extra six months past September 2018, three further months at EUR 20bn followed by a final three months at EUR 10bn. The ECB’s slowdown of QE should also be borne mostly by the public sector program (PSPP), while the corporate bond and covered bond purchases will be maintained closer to current levels. In regard to funding costs in 2017, the ECB has been very careful not to lift the foot off the gas too quickly, a trend we expect to continue. We foresee that ECB policy interest rate hikes may not come until the second half of 2019.

All-time lows for bank spreads, the creation of compression and flat curves

The cocktail of improved fundamentals combined with the monetary purchase programs (the Covered Bond Purchase Program (CBPP3) and the purchase of non-financial debt under the Corporate Sector Purchase Programme (CSPP)), has dragged down bank funding costs to be at all-time lows across all ranks of bank debt. We have seen the result create two traits in the market; a compression of spreads (the lowest difference between the ranks of debt) and a very flat yield curve.

Regulation to dictate the future funding mix

A further reason to be positive on certain ranks of the banking sector is the likely shape of minimum requirement for own funds and eligible liabilities (MREL) regulation. We judge that some market participants differ with our opinion on bail-in, and in particular the Tier 2 class.  So we take some time here to make our case.

There are two key parts to a bank rescue; first a loss absorption and then a recapitalisation. The CET1, AT1 and likely Tier 2 can provide the loss absorption, then the HoldCo Senior and Senior Non-Preferred debt can provide the recapitalisation. Our opinion is that regulators will push for enough capital, supplied via HoldCo Senior and Senior Non-Preferred debt to perform a complete recapitalisation of a bank.

The diverging paths do mean though that the European authorities have a problem. If MREL (which is just for European banks) sets a higher requirement than global standards then it could be costly and/or uncompetitive for the European banks. However, on the other side of the coin, regulators and member states will be acutely aware that the traditional senior debt would be a total nightmare, if not impossible, to bail-in. In the four European bank failures in the last twelve months the traditional senior debt class has not been touched, despite the amount raised from investors for the bail-in being nowhere near to regulatory desires. Legally it is very difficult to bail-in the rank of debt. Furthermore, as less traditional senior debt becomes available for bail-in, the regulators will be even less likely to go through the minefield of trying to find a way to bail it in.

Therefore, any capital requirements that allow traditional senior to be used for compliance is more just for show, and not of any benefit. In the end, a bail-in and full recovery of a bank is not fully possible without a fully subordinated requirement, which would mean MREL is set at a high level (26-28% of risk-weighted assets). If this was to be the case, this would boost the credentials for HoldCo Senior  and Senior Non-Preferred rank of debt.

The Quandary for the Authorities

The Desire for Regulators

The overall calibration of the MREL Pillar 2 requirement should allow for an appropriate loss absorption capacity (equivalent to the prudential capital requirements including the capital guidance) and for recapitalisation of a bank up to the level that would allow the bank to comply with prudential authorisation requirements and to have a sufficient market confidence buffer (MCB), in accordance with measures included in the resolution plan

 The Subordination Dilemma

The key question is whether subordination should be mandatory at higher calibration levels than the MREL Pillar 1 and for a broader subset of banks than G-SIBs. Some Member States call for the entire Pillar 2 MREL requirement to be met with subordinated instruments for all banks. They also call for MREL equivalent to 8% to be met with subordinated instruments for all banks. Other Member States prefer to limit mandatory subordination to the Pillar 1 MREL requirement alone. Moreover, some Member States prefer that resolution authorities should not be allowed to require subordination for the MREL guidance. As a compromise, the Presidency suggests that mandatory subordination should apply only to the Pillar 1 requirement (which applies to G-SIBs only). As regards subordination for the Pillar 2 requirement and guidance, subordination is discretionary and resolution authorities shall (once certain thresholds are met) assess whether subordination is necessary. Text: Council of the European Union – Banking Package – Nov 2017 Paragraph headings: ABN AMRO Group Economics

So, let’s assess the potential impact of a high MREL subordinated requirement. Firstly, if the eurozone authorities were to push for a fully subordinated requirement the regulations may not be as globally uncompetitive as first thought. The UK authorities have bitten the bullet and have set their MREL targets at 26-28% of risk-weighted assets, with a fully subordinated requirement. The Swiss authorities have likewise set a fully subordinated requirement at roughly the same level. For US banks they do not have MREL and have to conform to softer total loss-absorbing capacity (TLAC) standards. However, crucially, the US banks have been issuing their debt predominantly out of their Holding Companies (HoldCo’s) for a number of years, and therefore, they already have considerable amounts of their debt naturally subordinated.

Addressing the second issue, a fully subordinated requirement for eurozone banks would not necessarily be that costly. As the authorities demand more of the rank of debt for MREL, it actually becomes cheaper for a bank to issue it. The larger the subordination requirement, the better it is for the cost of the new debt class, and as the thickness of the layer shares the burden of recapitalisation and thus generates a positive correlation to recovery.

Our conclusion is therefore, that the eurozone authorities are likely to pursue a plan, after some political wrangling, which pushes for MREL to be a fully subordinated requirement. This view has been echoed recently in a Single Resolution Board (SRB) discussion, found here. This would induce Senior Non-Preferred issuance and be positive for the rank of debt.

Consequentially, Senior Non-Preferred recoveries would increase, while the Tier 2 recovery will not alter. It is only once/if Senior Non-Preferred performs the recapitalisation and receives a recovery of over 100% that any Tier 2 recovery can be then considered. Simply put, Tier 2 and Senior Non-Preferred debt will continue to take very different paths.

Our positioning in the European bank debt market

Aggressive: Low ranked debt  Defensive: High ranked debt, and long dated maturities

Now that we have reviewed the general themes, we look now a little deeper at the bank debt market. We take a mixed stance across the ranks of European bank debt as the monetary stimulus begins to fade, albeit very slowly, but fundamentals continue to improve. Our overall anticipation is that the gradual monetary policy reduction in stimulus does not have a sharp negative impact on spreads, a narrative we witnessed in 2017. Especially, as it will be offset by stronger fundamentals, accelerating economic growth and supportive regulation.

We therefore, underweight very expensive high quality ranks of debt and try to take selective positions in deals which offer enhanced spread. Our position is rather defensive in the long-end, and somewhat aggressive in the short-dated lower ranked debt. We underweight the traditional senior class, to finance overweight positions in Senior Non-Preferred. Then we underweight Tier 2 to take overweight positions in AT1s. As a rule of thumb, we prefer to invest in a lower capital structure instrument of a national champion rather than a higher ranking debt of a weaker name.

Breaking down the ranks of debt

  1. Additional Tier 1s:

Aggressive in the short-end, taking advantage of high reset spreads

We begin with looking at the lowest ranked bank debt first. The short-end AT1 space is our preference across all structures of bank debt. The returns available on short-dated AT1s trump practically all asset classes and sectors.  We particularly like the short-dated instruments as their low extension risk is enhanced by strong roll-down protection. This allows instruments to provide strong returns in a number of scenarios; if spreads tighten, if spreads stay stable and even if spreads widen by up to 45 basis points.

From an analyst perspective, the current low rate environment makes the outstanding short-dated AT1s which have (often very) high reset levels an easy non-extension calculation. We therefore, take advantage of these instruments which are in this short-dated ‘honeymoon’ period. However, this golden period will not last. The AT1s that are being issued now will not have this benefit in say four years. It will be fascinating to see how they trade in the future. Chiefly because the current market does not seem to be pricing extension risk on these new issues.

The very low reset spreads that banks are locking in now could mean that it is an attractive proposition to extend the maturity in the future years. This will then lead to move complex calculations for the AT1 market, meaning that the many individual constructions of each AT1 deal will begin to play  a greater role. As such, this will make the future AT1 market assimilate closer to equity than at present.

A casing point is the recent AT1 issuance from Nordea, NDASS 3.5% C2025, which was issued with a reset spread of 300.3bps above EUR 5Y swap. Under a scenario where the AT1 market is 150bps wider in seven years’ time the call of the instrument is not such an easy decision. If the curve was 150bps wider, it could mean that if Nordea did not wish to refinance the instrument that it would cost them EUR 12mn per year to refinance, roughly EUR 92mn for the length of the instrument once fees are considered. With these high costs under a benign scenario, the risk of extension in the future is increased. The decision of Standard Chartered not to call some of their preference shares, STANLN 6.409 2017, earlier this year, showed both the business case and the subdued overall market reaction to sudden extensions.

Consequentially, although we are positive on short-dated AT1s in the European bank market, the recently issued mid-to-long-dated instruments are a worry for us. Overall it appears that many of the features of AT1s do not seem to be applied to pricing. Especially the reset spreads and the consideration for subsequent call periods after extension do not seem to be playing that strong a role in market prices.

  1. Tier 2:

Attractiveness begins to fade as Senior Non-Preferred debt increases

We now transpose our earlier regulatory views onto Tier 2 pricing. At the start of the year we were positive on the Tier 2 debt class, especially from deals around the BBB area, some of which were offering over 200bps. Over this year, and even in the last few weeks, we have seen these names grind to unbelievably tight levels. Especially those at the longer end (over seven years) are most at risk and it is an area where we see hardly any compensation for the longer maturity. Therefore, we are worried that some of the trading levels could widen in the class past break-even points.

Accordingly, for positioning one can go either way. Our preferred method, based on current spreads, is that one invests in AT1s, and gains over 250bps more than Tier 2. This is a dynamic we like as we consider very similar rates of default on principal. Alternatively, one can pick strong selective Senior Non-Preferred names that should hold firm as future issuance supports their recovery versus Tier 2 levels.

It is also interesting when we consider this dynamic versus the current market prices. We can see in the market that the companies which have built up the largest buffers, often have the largest difference between their Tier 2 and Senior Non-Preferred or equivalent debt class. Below we show how the evolution of the Senior Non-Preferred class can impact the spread levels. The German banks with their already large layer of Senior Non-Preferred have the largest spread differential, while the French banks, which have only recently started issuance of the new debt class have the smallest difference. Therefore, as the issuance of Senior Non-Preferred continues the dynamics favours Senior Non-Preferred over Tier 2.

  1. Senior HoldCo and Senior Non-Preferred

Positive on the dynamics as recoveries improve under capital requirements

So now we move to the relatively new debt classes, of Senior HoldCo and Senior Non-Preferred. They too have experienced a splendid year. Banks would not have considered a few years ago that they could now issue the new debt class inside where their traditional senior class was trading. The takeaway is that although the probability of default is often similar in our view for Senior Non-Preferred and Tier 2 (we assume a one rating notch differential) the recoveries are vastly different. Crucially, the amount of CET1 and Tier 2 that is used in the loss absorption that makes us judge that the recapitalisation of a bank with the Senior Non-Preferred class can offer a strong recovery.

Especially in around four years’ time, once MREL buffers have been built up, the Senior Non-Preferred class offers a recovery which we think may sometimes underestimated versus for example Tier 2. Please refer to our publications which go into far greater detail on the relationship between Senior Non-Preferred and Tier 2 here (Part 1) and here (Part 2).

However, we must still be selective as at present the debt class is at extremely tight levels, a factor which also has been accentuated over the last two months. It is also interesting to note that largely there does not seem to be a difference between banks with larger buffers (a high issuance of Senior Non-Preferred) and those without one. The GSIBs are far more developed with their issuance of Senior Non-Preferred and naturally they should temporarily have a far stronger recovery versus non-GSIB peers. We pick short and mid duration names to pick up an enhanced yield versus the traditional senior rank.

  1. Traditional Senior:

Strong fundamentals, no benefit to our benchmark, but attractive against other classes

Now finally, we take a look at the traditional senior rank of bank debt, in reference to both the Senior Preferred and OpCo Senior debt. Once again this class has seen a great year, but to a lesser extent than the other ranks. In essence, we think that the bail-in of any traditional senior debt is very unlikely in a resolution scenario for any mid-or-large bank.

As we have discussed over the year, the large amount of resolution instruments, AT1, Tier 2 and Senior Non-Preferred, should greatly protect the traditional senior debt. The amount of resolution instruments should fulfil a key requirement under BRRD, to bail-in 8% of Total Liabilities before resolution funds up to 5% of total liabilities can be used. This results in a mammoth buffer of safety that surely could correct the troubles of any struggling bank. MREL will continue this trend and provide loss absorption and recapitalisation before traditional senior will be considered. The mammoth buffer in particular make us positive on traditional bank senior debt versus Non-Financial senior debt, as at present they often trade at similar levels.

From a supply side, traditional senior will also benefit from positive issuance dynamics, as regulators demand instead lower ranked debt instruments. Our expectations for issuance of the traditional senior debt class remains rather limited for next year, roughly EUR 21bn in the euro market. We discuss the upcoming bank issuance for 2018 in our publication here.

Finally, we also think the traditional senior class should be safe despite the softer approach by the ECB to QE in 2018. The compression dynamics mean that we now have the smallest gap between Covered Bonds and the traditional senior debt class. Firstly, we still see value in traditional senior versus Covered Bonds, especially around the 5 year part of the curve where you can still pickup 20bps. However, secondly, as we noted earlier, we expect covered bonds spreads to stay supported next year. The ECB’s slowdown of QE is to be borne mostly by the public sector program (PSPP), while the corporate bond and covered bond purchases will be maintained closer to current levels. Furthermore, Covered Bonds are reaping the reward of consistent increases in capital buffers and balance sheet improvement across the entire banking sector. The consequence is that some Covered Bonds are rated as AAA, but still have a whole six notches of unused uplift (meaning that in theory a bank could be downgraded by six notches yet the deal still maintains its AAA rating).

Despite this incredibly positive view on the rank of debt, and against other asset classes, we have to consider our benchmark. Against our benchmark, we actually go Underweight almost all the traditional senior debt. The rank of debt has simply too little carry to warrant our choice when we are largely positive on the banking sector. The underweight in this rank of debt, then finances our positions in the HoldCo Senior and Senior Non-Preferred debt class which offer greater carry.

Further reading

Please find below a number of related pieces to this publication;

21 Jul – Financials Watch – Resolutions rupture Europe, what is the impact?, click here

23 Aug – Financials Watch – Failing banks, bond pricing under the new regime, click here

3 Oct – Financials Watch – Navigating the path of future resolution, click here

6 Nov- Financials Watch – Foundations arise for a new bank funding mix, click here