Roundtable: AT1 goes global

In the space of a month Danske, Santander and UniCredit printed inaugural AT1 transactions while a non-bank UK institution, Nationwide Building Society, and a non UK financial institution Crédit Agricole opened the sterling sector. What explains the rapid and successful globalisation of AT1 after its opening in the US dollar market?

Peter Holm, Danske: I’m positively surprised at the speed the market has taken off at, particularly against the background of all the views you previously heard that there would be a limited investor clientele for these products. To start with it was in US dollars, but now you are also able to issue in euros, and we also saw Nationwide and then Crédit Agricole in sterling, so now we have basically seen three major currencies available. And there is a much broader number of investors, not only one particular group.

That it took off to the extent that it has done is very satisfactory — we had hoped that this would be the outcome. I think the yields available on these instruments in this low interest rate environment paved the way for this broadening of the investor base and has been very, very helpful in bringing about this situation.

Chris Huggins, GLG: The development doesn’t really surprise us because issuers need to build out a curve in these instruments, in both maturities and currencies. We envisage issuers funding in all the G3 currencies, and with a curve of call dates on AT1s.

While it doesn’t particularly surprise us, it’s a welcome development for investors to have a choice of currency and structures to look at. And the pipeline is clear: as the various jurisdictions in Europe get sign-off for the tax treatment for the coupons on CoCos, that is catalysing issuance in those jurisdictions. The UK and Switzerland were first to give sign-off, then France and Spain got done, and more recently we’ve seen Italy and finally Germany getting approval.

I wouldn’t say that there’s one particular jurisdiction that we are more or less excited about. It’s more at an issuer level than a country level that we look at these things.

Elisabeth Van Sante, Pioneer: The new regulatory regime is without doubt the main reason for the number of recent issues in the market; not only are the great majority of old-style subordinated instruments losing all regulatory eligibility after their first call date, banks also face higher capital buffer requirements where the equity-like and loss-absorbing CoCos have a preferred position. The Swiss issuers are even likely to issue only CoCo-style subordinated instruments going forward. The stock of old-style Tier 1s being redeemed over the next few years means we should expect close to $50bn (Eu36.2bn) of issuance per year, in a format that meets the new Basel III rules, and the fact that tax-deductibility on coupons is being agreed across jurisdictions will support issuance levels. Also, the upcoming ECB stress test is encouraging many European issuers to pre-emptively reinforce their capital buffers. For many banks, issuing AT1 is one of the most efficient ways to do so.

From an investor perspective, we believe demand is likely to remain quite solid in the short term given that the yields/spreads that these instruments offer can be interesting vis-à-vis the rest of an issuer’s capital structure.

Vincent Hoarau, Crédit Agricole CIB: The globalisation of the AT1 market is a natural step. The new regulatory regime and the requirements to expand the capital base are obvious reasons for supply to increase. Some major technical obstacles have also been overcome: most of the products are index-eligible, people are getting more and more comfortable with subordination to equity, and we also found a common basis to compare instruments and work on relative value.

But I think that the periods of frenzy we observed were mainly driven by the still-exceptional liquidity situation in the context of the low interest rate environment. As an investor, you can’t miss out on such juicy coupons and to some extent you are forced to buy. Those deals are priced a couple of hundred basis points wide of pre-crisis levels. You can hardly pass them up, even if you are not a big fan of loss-absorbing features.

Meanwhile, euro and sterling denominated bonds are offering investors very interesting opportunities for diversification while the deeply subordinated market was historically in US dollar format.

Raphael Robelin, BlueBay: I think that the evolution of the AT1 market is a consequence of the clarification of the Basel III rules and which structures will obtain regulatory capital treatment going forward. This is something that was only clarified last year and so a lot of the issuers we had met were waiting for this visibility before coming to market. On top of this there has also been affirmation from tax authorities in many jurisdictions confirming that these instruments’ coupons will be tax-deductible, which again I think is an input that potential issuers needed to be clear on before deciding on how much they wanted to issue, in which currency, as well as which structure was most appropriate.

We have known for some time now that by 1 January 2019 banks, from a capital standpoint, need to be Basel III-compliant, and so they are going to need a lot more capital. If you look at a bank’s capital stack they have the choice between having 2% of their risk weighted assets (RWA) in Tier 2 and 1.5% in Tier 1, or all equity instead. So if you think about it, for a bank with a cost of equity of 10%, and an average tax rate of 20%, it means that as long as you can issue sub debt with a coupon of less than 12.5%, it is economic for you to do so. We also know that legacy hybrid capital, in particular in the Tier 1 space, will typically cease to count as capital after the first call date if the bonds contain step-up language.

So if you put all these elements together: the need to have 1.5% of RWA for Tier 1, 2% for Tier 2, in hybrid capital, the visibility on the type of structures that will qualify as hybrid capital under Basel III and the visibility of the tax treatment, there’s going to be a large amount of supply so it comes as no surprise to us that banks are keen to be the first ones to establish a curve, and therefore we are starting to see a lot of issuance. We would expect this to be a continuing theme between now and 1 January 2019, when the Basel III capital rules will be fully implemented.

But is the opening of different investor bases surprising?

Robelin, BlueBay: It’s very symptomatic of where we are in today’s market. This is an environment where yield levels are extremely low and where attractive investments are more and more scarce. The average yield in the European high yield index is 3.75%, and the average yield in the euro corporate investment grade index is 1.75%. When the average yield on offer is so low it’s extremely difficult for investors to ignore an asset class giving coupons of between 5% and 10%.

So to me this is the one part of the equation where medium term I’m still a little bit sceptical. At the start of the year, certainly from a demand standpoint, all the stars were aligned: there was low volatility in the market, risk appetite was very high, and the low rate environment meant investors were desperate for incremental yield. I still feel, though, that there is a lack of dedicated natural demand for these new-type hybrid securities and in particular AT1. So far it has primarily been a combination of private banking investors, hedge funds, and asset managers going off piste that has led to the success of these deals. It is not obvious to me that all these investors see these instruments as core holdings, and so I suspect that we will have air pockets at some stage in this market. One of the great difficulties with this market is precisely the lack of a core stable investor base. Hedge fund managers and asset managers buying these as an off-index bet, or even private banking clients, may not have the appetite for the volatility in the asset class if bond prices go down five or 10 points, which in my experience can happen much more easily than people anticipate.

We at BlueBay tried to work with the regulators between 2008 and 2012 to make sure that the voice of fixed income investors was heard. We tried to make it clear that if regulators wanted fixed income investors to have a natural demand for new hybrid capital securities then they had to make sure that the deal features were broadly acceptable to this investor base. Unfortunately our advice was not heeded, so a number of features — such as the ability to be fully written down or converted into equity when the underlying bank is still a going concern entity, as well as the ability to pay a dividend and not pay a coupon — really mean that the capital structure waterfall is not being fully respected. As a result of this I think it is questionable whether these new type of securities have a natural home in classic fixed income portfolios.

Banco Popular opened up the euro segment in October 2013? How was the situation back then? What were the major differences compared with today’s environment?

Santiago Armada, Banco Popular Español: Yes, exactly, Banco Popular placed the first AT1 denominated in euros and it came as the third issue in Europe. At that time the market was underdeveloped, as you can clearly see from the number of investors involved and with the geographical distribution, which was concentrated in UK. The instrument was not widely understood, so the meetings needed to be much longer and deeper, and many investors still didn’t believe in the product and looked at it as a short term opportunity.

However, quite a few first tier investors saw the value proposition of the bank and went on buying bonds in the secondary. Thanks to them, Banco Popular’s AT1 has had so far an extraordinary performance, being quoted 14% higher in just five months.

What were the main obstacles that Danske had to overcome before launching the deal, which I understand you had been planning for quite some time?

Holm, Danske: We of course needed to have full clarity on the regulatory rules through consultation with the Danish supervisory authorities. But on top of that it became clear that we also had to deal with two tax issues, which held up the process a little bit further and meant that we had to leave the project on the shelf last autumn.

One of the tax issues came up last summer, if I remember correctly. The guidelines were unclear about the issue of if you should ever have a temporary or permanent write-down, and by doing this write-down create taxable income, whether you should make a deduction upfront for that potential tax liability. For some time that issue was with the European Banking Authority, and then late last year it was passed back to the domestic regulators to deal with. That eased the situation for us because we could discuss it quickly with someone who had an insight into our group, so we very quickly arrived at a solution to that — which is that there is no tax to be drawn upfront, so no haircut, so to speak.

The other issue was on the tax deductibility of coupon payments. We did the first hybrid issue in Denmark, in 2004, and at that point in time we had a change in tax legislation giving us a deduction on the coupon payments. Then you had the change in regulations requiring fully discretionary coupon payments, and that created a new question over tax deductibility, and we had to have it confirmed again. That was confirmed in legislation passed by parliament in March.

As part of our capital plan we wanted to complete our AT1 by 11 April in light of the repayment of the government hybrid Tier 1 that we took in 2009 for Dkr24bn, with that date being the first call. We had been planning for that for a long period, doing equity issues, and then adding some Tier 2 issues, which you referred to before — in euros, the Trekroner issues, and a little Swiss francs — and we wanted to complete that chain of transactions with an AT1 to be fully prepared for the repayment of the government hybrid.

Why did Crédit Agricole SA opt for sterling on top of euros for its second appearance of the year, which was not something that had been done before?

Hoarau, CACIB: A large number of UK investors expressed an interest in buying sterling exposure to CASA when we were roadshowing the inaugural US dollar AT1 in early January. The same accounts also gave unsolicited indications of interest at an early stage of the marketing process for the second issue. That interest, plus the size of the demand attracted by Nationwide for its inaugural AT1 in sterling, made the sterling tranche an easy decision.

The dual currency format was validated when we received evidence from investors that there was no risk of cannibalisation between the two offerings. The sterling market is as nascent as the euro market, while supply in absolute terms is really limited, at least for the time being. So we did not see any downside to doing a dual tranche issue while we knew from the outset that the size of both offerings would remain relatively limited.

And yes, we certainly broke new ground. It was the first multi-currency AT1 benchmark, as well as the first time a non-UK financial has sold benchmark sterling AT1. Overall, the transaction was very well received and the three points of performance in the secondary market should satisfy everyone involved.

If you hadn’t seen the euro market evolve so much so quickly, would Danske possibly have gone ahead in dollars?

Holm, Danske: For us, the fact that the euro market developed in such a way that we could take advantage of it was a very positive development. We prefer to issue these instruments in a currency that is closer to our home currency, Danish kroner, which gives us a much more stable capital element, so to speak.

You did a “Trekroner” Tier 2 issue last year in the Danish, Norwegian and Swedish markets — is Danish kroner a currency in which you could envisage doing an AT1 in, or are any of the other Nordic currencies?

Holm, Danske: I don’t think one should exclude that as a possibility one day when the need appears. Definitely we should not exclude that. And in that respect it was very, very positive for us to see that when you look at the investor clientele behind our deal 24% was allocated to Nordic investors.

We suddenly have plenty of reference points in the secondary market, in US dollars, euros and now in sterling. Are the relative value schema of these instruments fully consistent across the different currencies? What about the cross-currency element and the possibility of arbitrage across markets?

Huggins, GLG: I would hesitate to suggest that there are arbitrage opportunities, because I would imagine that bid/offer spreads, transaction costs and borrow costs probably eat up any meaningful arbitrage that exists. However, we think there are opportunities to switch between structures or currencies from the same issuer to ensure that investors have exposure to their favoured part of the curve, if you like. And different investors will have a different idea about what represents value there.

If you take the Lloyds exchanges, for example, there are now three deals in sterling with three call dates. There was quite a lot of flow in the ECNs that were going to become the AT1s, and you could see how interest shifted amongst them. While the deal was initially priced by Lloyds based on how it thought investors would see value in the curve, the market priced it slightly differently. I think more generally there’s a preference for longer dated calls.

Van Sante, Pioneer: The investible universe is certainly growing as this market expands rapidly across a number of jurisdictions. The fact that the structures differ, that spread levels vary across markets, and that the technicals are also different creates opportunities across currencies for investors. Our impression is that the market is not differentiating much between certain structural features — either within or across currencies. This also creates opportunities before the market matures and pricing efficiency increases.

Robelin, BlueBay: I think that there are great arbitrage opportunities in the market. Clearly the more issuance we see, the more diversity we will have — be it the different currencies in which we have instruments outstanding, the different structures in terms of high and low triggers, conversion into equity and write-down, the different levels of coupon resets and so on. There are many inputs and in a young asset class the market will tend to be quite inefficient in pricing them. I also think that there is still a bit of a question mark for investors as to whether they should value the instrument on spread or in yield.

Filippo Alloatti, Hermes: Yes, between currencies there are from time to time some opportunities. But the way we would approach the question of valuations is that there are many different angles. We look for example at all old Tier 1s, we look at the bank’s cost of equity, we look at trades banks may do in the reg cap space, and then we take, if you like, a holistic view on valuations. But the fact that we can invest in the three main currencies, and eventually could consider other currencies, also allows us to express some views in different currencies.

Hoarau, CACIB: The number of reference points in the secondary market increased significantly at the end of the quarter, with Danske, Santander, KBC, Société Générale, CASA and UniCredit surfacing within 15 days in euro, dollars or sterling. But there is still a lot of discrepancy in the valuation schemas. Structures differ across jurisdictions and issuers, while the handful of instruments available is quite inhomogeneous.

I also think that there is not enough credit differentiation in the asset class. We meanwhile have to bear in mind that the quality of the placement in the primary market and the sizing of the tranches also play an instrumental role in the level of performance in the secondary market as well as volatility. So, it’s not an easy call when discussing consistency of valuations. There are always different ways to look at a trade.

When UniCredit’s inaugural US dollar AT1 surfaced at 8%, you had investors valuing the trade three-quarters of a full point higher! The 8% yield was certainly eye catching and high enough to ensure a good trade, but if you looked at it in spread it was quite tight, in particular versus BBVA AT1 perpetual non-call five, to which the UniCredit perpetual non-call 10 came flat. Nevertheless, versus a UCGIM 9⅜ 07/29/49 “soft CoCo” trading at 6% back then, the new deal looked cheap.

There was also a lot of price discovery around the recently launched CASA perpetual non-call 12 AT1 tranche in sterling. We received indications ranging between 7.25% and 8%! Some investors worked on relative value terms versus Lloyds. Others simply reflected the euro-sterling yield curve differential to come up with a valuation in sterling. In addition to that, whether or not you crystalise the inversion of most subordinated sterling curves was debated. And ultimately I have to admit that there is also a gut feeling element that interferes with this objective valuation process, plus the envisaged deal size and market tone consideration at the time of the launch.

When you see a new issue where there is more than one currency, for example CASA in euros and sterling, how do you position yourself?

Robelin, BlueBay: There are many different inputs involved in our investment process. Usually for a multi-tranche deal you would expect the structure to be identical but the initial spread will be quite different. Clearly that means that there is more risk of one bond not being called than the other, so that obviously is one differentiating factor between the different tranches: you’ve potentially got far more extension risk for one deal than for another.

Then there is the specific analysis of supply and demand for the different currencies based on the issuer and how much outstanding debt the issuer has in the respective currencies. Where is the domestic market? How much supply have we had over the last few months in the respective market, and how much supply do we expect in these different markets going forward?

All else being equal, we think that there is a natural bias for European accounts to buy the debt of European banks, so we are not surprised to see a strong demand for euro-denominated deals, for example. And so it really is an analysis of all these factors. Clearly for the dollar bonds you have to assess whether there is demand out of the US but also what Asian demand there is. Asia tends to buy far more dollar-denominated debt than euro-denominated debt. These are examples of the types of inputs we take into account before deciding which tranche to invest in.

What can we say about the spread differential between Tier 2-hosted CoCos and AT1? What about the value of the deferral element?

Van Sante, Pioneer: At the moment, the key differentiating feature between T2 and AT1 CoCos is centred upon the risk of coupon deferral, which therefore commands the bulk of the discount over Tier 2 structures (currently in the range of 130bp-150bp). This is valid in our view when considering that coupons on AT1 instruments may be suspended well before the capital ratios fall to the official trigger point. In theory, coupons may be suspended from the moment capital ratios fall into the capital conservation buffer; although when precisely this happens may differ between national regulators. For example, the UK PRA and Swiss regulator appear to have a more rigid view on the point of non-viability. As national regulators’ views on PONV become clearer, this is something which may require a more nuanced pricing approach. Going forward, the market may focus more on other features such as extension risk, which may mean that AT1 trade at a greater discount.

Hoarau, CACIB: I fully agree with the risk differential implied by the two instruments, but I think that in terms of valuation, the differential should be a bit higher. And this is what the market is telling us: there are roughly 200bp of spread differential between Barclays US dollar 8¼ 12/29/49 AT1 (PERPNC5) and Barclays US dollar 7¾ 04/10/23 (T2 10NC5). In the euro market, CASA 8⅛ 09/19/33 T2 (20NC5) is trading in the context of 4.35% or around 275bp over five year swap rates. This implies 205bp of differential between Tier 2 hosted CoCo and AT1.

I expect the differential to widen further, because AT1s are much riskier instruments with regards to the risk of coupon deferral, and I think that this risk is undervalued and will be reassessed.

Elsewhere, the scarcity element regarding Tier 2-hosted CoCos will likely push spreads tighter and lead to some squeeze situations, while — in the meantime — AT1 supply might weigh on valuations.

Does issue size play as important a role as the credit itself or the distance to trigger? What are your priorities when evaluating these instruments?

Van Sante, Pioneer: We would usually require a minimum size of 500m to invest in these instruments, and above that figure we would not really differentiate between 750m and 1bn. But once the deal size is above our minimum threshold, the size would not play a huge part in our decision-making process.

Our priority when assessing those instruments is rather the quality of the issuer and the structure of the instrument, including the trigger point, the capital cushion above the buffer, the write-down mechanism, duration and structure (AT1 or T2 host). Our preference remains for structures with a recovery mechanism (be that a temporary write-down mechanism or equity conversion versus permanent write-down).

Hoarau, CACIB: Triggers as a percentage and distance to trigger are key parameters. Even more important is to look at them in absolute terms with the balance sheet in front of you. You can better assess the risk of burning through the buffer and triggering a write-down. But yes, the issue size is also critical to the liquidity of the bond, and more importantly the size of the deal has an impact on the performance of the instrument in the secondary market. So it should also be taken into account when modelling the valuation of the instrument. The market recovered from the wobble experienced mid-March after some hybrid capital new issues performed poorly after pricing. And that was the result of the size of the print versus the coupon offered. Too big and too tight, perhaps!

When we brought the CASA euro/sterling AT1 transaction to the market, one of the objectives imposed by the funding management team was to protect the performance of the transaction in the secondary market. The sizing strategy was therefore designed around that and I think sticking to £500m and Eu1bn was key for the overall performance of the trade even if we could have easily printed a much bigger trade.

What structural features do you prefer or consider key? And are there any you think are underappreciated?

Alloatti, Hermes: The so-called Maximum Distributable Amount (MDA) is very important, and if you look at some transactions then you may get the feeling that this aspect is underappreciated by the market.

And I would add in respect of that two aspects that are very important and should also be borne in mind: first, the attitude of management vis-à-vis the bondholder and vis-à-vis the equity holder; and second, the attitude and the track record of the regulator, because in some tough situations the regulator could actually swing the pendulum towards one side quite easily.

You really have to look at the management’s track record in terms of putting the shareholder or the bondholder first, or the other way around, in recent decisions. And also look at the track record of the regulator because in the end these people can have a big impact. I’m thinking of the FSA in Denmark increasing the RWA requirement for Danske Bank to the tune of 100bp, or the UK PRA taking a selective approach to the implementation of the Leverage Ratio, or the Swiss imposing additional RWAs for operational risks at UBS. And I don’t expect those to be the last examples of this type of regulatory intervention.

Also — to complement the answer — the distance to trigger is important. Some people are suggesting the distance to trigger is so far away that these instruments basically are non-CoCo, i.e. straightforward bonds with a tail risk attached in the form of massive systemic loss and a triggering of the bonds. But we don’t feel this is an academic debate.

Huggins, GLG: We greatly prefer equity conversion triggers to write-down structures because we think it aligns the interests of issuers and shareholders with those of bondholders. There are a number of scenarios one can construct where a write-down structure perverts the normal hierarchy of capital, i.e. if a bank is close to a trigger on a write-down structure, then management might want the situation to deteriorate in order to bail in junior bondholders, or CoCo holders, without equity shareholders being diluted.

Alloatti, Hermes: I think it is a question of taking them on a case by case basis, because, for example, the Société Générale write-down, write-up is to some extent different from the Crédit Agricole write-down, write-up. Philosophically and speaking generally, perhaps equity conversion is better because you have a little bit of upside.

But then also it depends on what the floor on the share price is, if there is some type of pre-emption right for the current shareholders, and if it is converted 100% equity or if there is a combination of cash plus equity. There are a lot of elements that need to be considered.

Were both equity conversion and write-down open to you? And if so, why did you choose what you did?

Holm, Danske: I would say that, in respect of regulation, all the options were available to us. But the bank has been very clear in this respect, that we would like to deal with this as a debt-like instrument and to separate the elements of this from the shares of the bank.

Regarding bookbuilding and order inflation, is a $25bn order book and 1,000 investors involved a healthy development?

Huggins, GLG: You can’t read a huge amount into it. There will always be an inflated order book for a bond where there is anticipated to be new issue premium — there’s never a shortage of investors who want free money!

Our concern on inflated order books is that obviously it persuades issuers to push the terms to a point where there is no longer value, so investors just have to be disciplined about not following the crowd too much and ensuring that they know when to stay in and when to come out of an inflated order book. In some of the more recent deals that have subsequently struggled in secondary, perhaps the size of the deal was influenced by the size of the order book, and that weighed on its secondary performance.

Van Sante, Pioneer: While overhang was not a major issue for us in the first deals that were issued in the market, we are obviously becoming more concerned at the size of the books and inflated orders. Part of it is a welcome increase of real money investors in the investor base, but the somewhat disappointing performance of the most recent deals — which were many times oversubscribed — shows there is still a very high proportion of inflation and some faster money in the books.

Hoarau, CACIB: If you are referring to Crédit Agricole’s inaugural AT1 and its record order-book of $24.5bn, the answer is yes, it is manageable, but it is not healthy at all. The total book size is completely misleading and it is a real challenge to appreciate the level of inflation and the level of real demand for the security. As Elisabeth said, there are lot of fast money investors on board. On the one hand you want to zero them all to best protect the secondary market performance, but you can’t discriminate too much. So in the end it is just a question of how you make everyone equally unhappy while favouring smart and real money accounts. But more annoying is when an issuer decides to upsize a deal because the book looks many times covered. And in that respect I like very much Chris’s comments on investor discipline!

Alloatti, Hermes: I can understand why some books get oversubscribed so fast, given that there is in general in the market a hunt for yield and we have basically no issuance out of the old-style securities, the old Tier 1s, for obvious reasons. But at the same time when you see the books on some recent deals being more than 10 times covered then of course you start asking yourself some questions. There is a little bit of froth in the market, to put it politely.

Holm, Danske: I don’t think we have ever had a roadshow where we have seen interest to the extent that we saw for our AT1 roadshow. And when we opened the book we knew that even if we were out there simultaneously with Santander, there was great interest in this. I think we stood at Eu17bn at one point in our bookbuilding before we reduced the coupon to 5.75%, and then some investors left the book, and we ended up at some Eu13bn with some 700 accounts.

Was this a problem? I don’t think it is a problem that you see investors showing such an interest in the credit of your bank, in the instruments you are offering. But of course it leaves the syndicate and at the end of the day also the issuer with an allocation issue to deal with.

Is it hard to judge the right size in light of the demand? And what guidance did you give on size, if any?

Holm, Danske: We will normally guide investors as to what we intend to do, and in this case we indicated that we would do a benchmark issue, which in our terminology is at least Eu500m. But we also gave guidance that investors should not expect us to do a jumbo. So I think investors who saw us on the roadshow had a fairly good idea that this was going to be an issue in the amount of Eu750m. There might have been some investors believing that we would drive that to Eu1bn, but Eu750m was what we had internally been aiming at, and that was what we stood by when we did the deal — even if it would obviously have eased the allocation process a little if we had done a bigger issue. So we stood by the Eu750m and we took the allocation problems. And I think the fact that we stood by Eu750m, and we had a quality book, can be seen in the aftermarket of the issue, which was relatively good compared with some of the competing supply.

Spanish national champions now trade very close to UK names in AT1 — are you surprised at this?

Armada, Banco Popular: It is no surprise. It is only the speed of the economic upturn that is surprising. We’re at the point when recovery is translating into figures at a pace much higher than expected. As an example, revised Bank of Spain projections for 2014 are 1.2% for GDP and 1.1% for private consumption, while just a few months ago we were expecting a quite weak recovery. This will also bring closer the end of the Spanish banking restructuring, from which we will come out strong and maintaining our recurrent profitability.

In this context, the appetite for Spanish assets that started to pick up around November 2013 is also accelerating. We are seeing inflows basically in all asset classes, and I think the AT1 category is a very attractive recovery play.

Do you feel investors are being too indiscriminate resulting in too flat pricing across credits?

Huggins, GLG: Our concern on inflated order books is that obviously it persuades issuers to push the terms to a point where there is no longer value, so investors just have to be disciplined about not following the crowd too much and ensuring that they know when to stay in and when to come out of an inflated order book. In some of the more recent deals that have subsequently struggled in secondary, perhaps the size of the deal was influenced by the size of the order book, and that weighed on its secondary performance.

Some exceptions apart, most of the trades so far have shown a tremendous performance in the secondary market. Is the current liquidity situation the main theme in the spread tightening across asset classes?

Van Sante, Pioneer: Without doubt, high cash balances and a search for yield resulted in significant spread tightening for most of the AT1/CoCo deals launched last year. However, the performance of recent deals has been quite mixed given tighter valuations and greater supply. A heavy supply pipeline is likely to lead to a far more selective approach amongst investors over the coming months, notably as MDA considerations become mainstream. The difference between banks that have front-loaded that risk and others is not yet reflected properly in valuations so we would expect more differentiation going forward.

Hoarau, CACIB: Credit investors are sitting on mountains of cash and are eager to increase credit, spread and duration risk because they are all convinced that the normalisation of credit markets and the convergence across assets classes will continue. So the expectation of further spread compression is pushing everyone to get involved in the AT1 market. As said before, in a very low interest rate environment when you are a portfolio manager you can hardly afford to miss out on the juicy coupons being offered by top tier European financial institutions. And when you get disappointed in allocation in primary you complete your investment in the secondary market. So, it is not a surprise to see most AT1 deals performing well off the break.

What explains the poor performance of some of the deals?

Robelin, BlueBay: As always, when a market has just rallied 10 points, you end up with a dynamic of investors chasing the market, and perhaps those who did not participate at the start of the rally suddenly feel that they need to start getting involved. This usually means they end up buying deal structures that are ever more aggressive and less favourable for investors, buying at valuations that are ever more stretched.

And remember initial valuations are important for two reasons in AT1: they are important because this is the headline coupon you are being paid; it’s also important because the reset of the coupon is based on the initial spread, and so if a bond was allowed to come at the top end of the range in price, low end of the range in coupon, and then you move back to the middle of the range, it makes it far more likely that this bond will not be called at the first call date. In other words, the call option that you have sold to the issuer as an investor — i.e. the option to call the bond typically every five years after the first call date — is far more valuable because the coupon reset is that much tighter.

So from that standpoint you have to be particularly careful to not buy new issues at the top of the market, and maybe the deals you are referring to performed poorly specifically because of that dynamic. On the back of a 10 point straight line rally in this space, more and more investors who are not natural investors were getting involved. That is typically a sign that the market is due for a little bit of a correction and when there is a correction it’s always the most recent deals that came at the most challenging valuations that will tend to perform the worst, having been allocated to new entrants who are less naturally biased to buy the asset class, and being the deals with the least favourable structures.

Do fundamentals justify the performance of AT1? Should we expect a correction at some point, such as AQR?

Van Sante, Pioneer: The European banking sector is undoubtedly on an improving path and the pressure to de-risk and rebuild capital will persist until the results of the AQR/BSA (Balance Sheet Assessment) are released. In this context, some of the spread tightening we have witnessed has been justified by improving fundamentals.

However, we believe that some AT1/CoCos are now looking a little rich, and given the compressed nature of the market we are moving up in quality. We try to consider valuations in the context of the broader IG/HY market as well as an issuer’s full capital structure.

Hoarau, CACIB: We already observed a correction move in mid-March after some issuers pushed it too far in primary and deals underperformed off the break. But markets recovered quickly. Nevertheless, since then investors are soberer, more selective and price sensitive. Inflated orders in order books have not disappeared, but things are getting more reasonable. Valuations in the AT1 segment will remain very volatile, anyway, because at current levels you see more and more buy-side accounts that dislike AT1. But I have to admit that most of the time those same investors end up buying because they are forced to do so.

In the end, fundamentals do not justify the level of performance. Current valuations reflect the fact that market participants are trading the levels of liquidity in the market and the negative net supply across asset classes. And this is one of the main drivers and supportive factors out there in my opinion — together with the yield offered by the asset class, relative to others, of course.

The main question is where do you put the floor in AT1? A way to approach the question could be to look at the high yield index, because in the AT1 segment we have investment grade issuers issuing non-investment grade instruments. And, as Raphael suggested before, you will see that there is still much room for performance. But we will continue to have some periods of indigestion and congestion on the back of oversupply.

In the end, I think the real correction might happen in the second half of the year when people will start focussing on the AQR and stress tests. A failure could theoretically lead to the write-down of an AT1 instrument and if investors were to anticipate that, you could see the appetite for the product becoming much more volatile and investors getting very selective.

Huggins, GLG: Probably the more macro volatility there is, then the more investors will congregate around the stronger structures and the better credits, and that will probably drive dispersion.

In terms of events on the horizon to worry about, yes, I think the AQR is definitely one, but more generally Europe is very vulnerable to a deflationary shock right now, and some of those economies are more vulnerable than others. We think that’s something that all investors should bear in mind when they are underwriting a deeply subordinated, high trust instrument from a levered issuer with a mid-single digit yield.

Robelin, BlueBay: AQR will definitely have some impact because clearly for European banks it’s one of the key game-changers for this year. I suspect that there is as much possibility that it will have a positive impact as a negative impact, because I don’t see many banks who are particularly at risk of failing the stress tests having AT1 bonds outstanding. The single biggest risk to AT1 instruments in our opinion is the non-payment of a coupon and the rules around stopping distributions by banks will only go live on 1 January 2016.

Alloatti, Hermes: The AQR does play into our strategy in the sense that we may expect some issuance of AT1 hosted CoCos because some banks may have to resort to the public markets in order to increase their capital, even before the results of the ECB Balance Sheet Assessment are published — some banks said to us that they have been told to raise additional capital by their own regulators. And as long as raising capital through CoCos is tax efficient in terms of coupon deductibility, and non-EPS penalising, it is quite attractive from a corporate finance point of view to have 1.5% of RWAs in AT1 and the finance directors of these banks may see an opportunity to tap the bond market as opposed to asking shareholders for more money. So that’s how it could impact the supply.

For the investor, or course, the question we ask ourselves — and which is very difficult to answer — is how much supply there will be and whether it is digestible. Recently, for example, we had the Lloyds exchange, which resulted in an additional £5bn of AT1s on the sterling market. So we are constantly monitoring the situation to see if there is too much coming into the market too quickly.

But more generally I think there is an important point to be made, which is that in the context of similar securities, be they financial or corporate, with a similar volatility, it is not obvious to us that these CoCos have outperformed. If you compare the performance of CoCos against the share prices of their issuers, then you can find some unexpected answers. So for us the question, do fundamentals justify the performance, actually we think this should be the other way around.

And regarding whether we expect a correction at some point in time, for us the relevant question is how the space will perform when we eventually have the first case of a security having the coupon switched off.

These securities have been spread so broadly across the market — you have hedge funds involved, real money investors, private bank investors, and you even have individual retail invested — and an important point is that the buy-side’s response in a coupon switch-off event will be very difficult to predict.

We can take some examples from the past: we had some German issuers in the so-called old-style securities that were Basel II-compliant — with the dividend stoppers and pushers that are prohibited under CRD IV — and French insurance company Groupama switched one coupon off on its Tier 1. The market reaction was quite severe and the bonds went down by some 10 points in a trading session — that was because the market was discounting a situation where coupons were switched off for many years, which turned out not to be the case.

To what extent are rating constraints a factor in AT1?

Alloatti, Hermes: Ratings are market relevant, and we take them into consideration, but they are not the principal driver that governs our investment decision.

Van Sante, Pioneer: Ratings — whether the bond is non-rated or high yield — can be an issue for some specific funds, but we have quite a lot of flexibility across most of our funds to take off-index positions. Also, we rely more on internal ratings rather than those of the agencies. We put a lot more emphasis on the investment case, the structure of the CoCo or the distance to trigger and relative value. Linked to this, the inclusion of such instruments in broad credit indices could help those instruments become mainstream.

Robelin, BlueBay: The one investor base where the rating constraint is real is the investment grade fixed income space. To me, it is questionable whether this is a natural asset class for an investment grade bond fund, because again, as I mentioned earlier, banks have the option to turn off the coupon at any moment. The regulator also has the option to turn off the coupon at any moment. The coupon is non-cumulative so this income is lost forever. Banks also have the option to make these instruments perpetual, and while some transactions have been designed more favourably — i.e. if they are not called after the first call date then the coupon is reset for five years, and only callable again five years later, other transactions have been issued to the market where after the first call date there is a mismatch between a coupon that is reset for five years but then from the first call date the instrument is callable every three months, which I think will be a nightmare for bond funds because then you don’t know the interest rate duration of the instruments. That will make it quite difficult for institutions to own this type of paper.

The vast majority of the universe is rated high yield so if you are ratings sensitive it doesn’t really make sense to get involved. But also, to me one of the single biggest risks is that there will be from time to time one or two bonds where there is a suspension of coupon payments because the bank needs to raise capital and then the bank does raise capital and then coupon payments are resumed in due course. The problem is, with the methodology that ratings agencies employ, whether you just miss a coupon payment, even though it is in the docs that you have the right to skip coupons, or you default, it’s the same outcome. The issuer would be downgraded to single-D. So if you are ratings sensitive, between the risk of changes of methodology that we currently see with S&P and the risk of a straight downgrade to single-D as soon as one coupon is missed, to me you would have to be extremely brave to be very active in this space. I think it only makes sense to be active in this space if you are not ratings sensitive and if you have full flexibility to invest across the ratings spectrum.

The other aspect is to what extent these bonds will eventually make it into the fixed income indices. Right now most of these are high yield rated, so I think that therefore there is pressure from high yield investors to buy them, less so for investment grade investors. In due course, as banks improve their profile and as the asset class becomes mature, will there be pressure for rating agencies to move a lot of these bonds to an investment grade rating, and will they be eligible for index inclusion? This is very much what we saw with the old-style Tier 1 bonds that started to come to the market in the late 1990s, entered the indices in the early 2000s, and became very much a mainstream asset class for investment grade investors over the following few years after index inclusion.

German investors have proven reluctant to buy any type of loss-absorbing fixed income instrument, while German issuers have been absent from the AT1 market. Is this frustrating?

Van Sante, Pioneer: Clearly it would be better to have German investors participating in this market, but liquidity seems to be improving given that the investor base is growing significantly in many other jurisdictions. However, now that tax-deductibility on coupons has been agreed for German AT1s, it is only a matter of time before Deutsche Bank hits the AT1 market and likely brings with it German investors to this market.

Hoarau, CACIB: It is a bit frustrating. But Germany lagged behind in putting in place the tax framework and deductibility for this new generation of loss absorbing contingent capital trades while, so far, the bulk of the supply has come in US dollar-denominated format. And as euros becomes established and the German authorities have sorted out the legal and tax issues we will see the first German AT1 transaction. This will force most of the reluctant German investors to get involved. It is mainly a matter of time. But I also believe it is a matter of education.

Robelin, BlueBay: I do think that a Deutsche Bank AT1, for example, could be potentially interesting to German retail. But again, it will depend a lot on how the deal is structured and marketed including details like the minimum lot size for example. Having access to Deutsche Bank at a high single-digit coupon could be appealing to German investors.

On the institutional side, I still think that the level of demand is questionable. First of all, as you know, the capital treatment for that type of investment is becoming quite penal, so for other banks or for insurance companies AT1s are very expensive to hold in terms of capital usage. So this is why again I think so far it has been primarily hedge funds, private banks, and to some degree asset managers who have been buying them. So it is not necessarily obvious to me that the level of institutional demand for a Deutsche deal would be a game-changer. I think it really depends whether they want to tap the German retail market or not.

Do you already have further plans in AT1 for 2014, or any intention to diversify in currency and build up your curve?

Armada, Banco Popular: As one of the first players in this market, we probably maintain a high profile for most of the investor base, so a follow-up transaction would benefit from Banco Popular’s inaugural AT1 success. We would like to build our T1 buffer well in advance, but we cannot anticipate if it would take place in 2014.

We could go for longer tenors and other currencies, but we are currently keener on sticking to euros. The final decision should be taken in line with Banco Popular’s strategy and targeted capital structure.

Holm, Danske: I would say that if we’d had any actual plans for doing AT1 in the immediate future, we should probably have availed ourselves of not only Eu750m, but perhaps done Eu1bn, which investors probably would have liked to have seen. So we presently have nothing on the agenda in AT1. But as for what the future will bring? Well, we will probably be back at some point in time.

How do you anticipate your activity for the rest of 2014 in the CoCo space?

Van Sante, Pioneer: We will be active but highly selective in this market — we will maintain a quality bias — focusing on high quality investment cases and good quality structures.

Robelin, BlueBay: We at BlueBay are in the process of launching a dedicated CoCo fund to take advantage of the opportunities in this space, because we think that banks are one of the improving sectors. They had a terrible time during the 2008 crisis, but it is our experience that the sector that has the worst crisis the previous time around very often has a much better crisis the next time around — telecoms, being the sector under pressure in 2002 and then subsequently faring far better in 2008, is a typical example of that dynamic. With banks, I suspect this will be amplified by the fact that not only did management teams have a shocking crisis and would typically learn from that, or get replaced, but so did the regulator, so the regulatory environment is changing dramatically. The pendulum is going from one extreme to the other in terms of oversight. So banks are going to become far more utility-like in our opinion, the volatility in quarterly earnings is going to be a fraction of what it used to be. For a creditor, this is good news.

And so I think that this is clearly one of the key selling points for this new family of hybrid instruments, which are giving you attractive coupons when the underlying credit risk will improve over time. So the extent to which we see a growing investor base, as asset managers come to this same conclusion, and further dedicated funds are launched, will be key to the success of the asset class. This is the only dedicated money that I can see right now that when the market goes down will have a bias to buy more instead of thinking: “Ouch! This is not my core universe, my benchmark doesn’t have these bonds, I don’t have a bias to own them. I’m a hedge fund, or I’m managing against the investment grade index. I want out.”