To see the digital version of this report, please click here.
To purchase printed copies or a PDF of this report, please email gloria.balbastro@thomsonreuters.com.
IFR: Perhaps the first thing to do is get an update on where we stand today in Asia, relative to the rest of the world, and where things may be heading next. I’ll ask Sean to tackle that one first.
Sean McNelis, HSBC: The Asia Pacific region is quite advanced in terms of implementation of Basel III instrument criteria. Across the region, if you look at the larger markets such as Hong Kong, Singapore, Australia, Malaysia and so on, they’ve got rules in place now that allow banks to issue both Tier 2 and Tier 1 capital in Basel III-compliant format. We’ve pretty much got certainty in terms of the instruments the banks can actually issue. Europe just got that in June with CRD IV [Capital Requirements Directive], and we’re already seeing the first T1 transactions there (BBVA, Soc Gen and Banco Populare). The banks here in Asia also have the template in place to allow them to issue T2 and T1 capital and the issuance has already begun with the UOB [United Overseas Bank] Singapore dollar Tier 1 and the ICBC (Asia) US dollar Tier 2.
There are one or two exceptions to that: the CBRC [China Banking Regulatory Commission] in China, for example, is still finalising some details of their new T1 regulations. However, overall, if we take the countries where perhaps we’d expect to see the first supply of Basel III instruments, Australia has already been busy, with the banks there issuing primarily domestic T1 transactions in Aussie dollars, where they’ve got a very deep market and, selectively, T2. With the UOB transaction we’re starting to see Additional Tier 1 instruments out of Asia – in Singapore dollars so far, but I think you can expect to see US dollar transactions in due course – and we’re starting to see banks issuing Tier 2 in their local markets. CIMB did a Basel III-compliant T2 in Malaysian ringgit at the start of September, and we’ve also seen a number of smaller transactions in quasi-private placement format, for example, from some Indian banks doing local currency 10-year non-call five.
There is still some tax uncertainty in some jurisdictions, as to whether the coupon is deductible from a corporate tax perspective. In Europe it’s been a big source of discussion and the UK has announced that both T1 and T2 are tax-deductible, whereas in Asia we haven’t had that clarity from the tax authorities except from Australia where there were changes to legislation. So, it’s going to have to be case by case. There are still some factors that banks need to take into account when they’re looking to issue but in terms of the uncertainty we’ve had in the past three years, I would say 95% of that has gone away.
IFR: How big is this market we’re talking about? You mentioned there’s been some innovative T1 deals out of Europe, but what about in this region – can anyone put a number on what we’re expecting in terms of T1 or T2?
Sean McNelis, HSBC: Maybe just some high-level numbers, and leaving aside Australia for the moment, if you look at Asia ex-Japan and take the four largest banks in each country, they have risk-weighted assets of about US$8.0trn. Under Basel III, it’s basically efficient to have approximately 1.5% of your RWAs in the form of Additional Tier 1 and approximately 2% or more in the form of T2 capital. Historically banks in Asia have been relatively frequent issuers of T2, but not T1. In Hong Kong, for example, only Bank of East Asia has done a T1 transaction. So, when we look at RWAs and legacy instruments, China is clearly the biggest potential issuer, as there were no AT1 regulations before. The banks are so large that the potential supply, over time, is very significant
Basically, taking 1.5% of RWAs, the total issuance of AT1 over time we’d expect to be about US$120bn equivalent across local currencies and G3. I stress this will be the total stock over many years and it will also reflect refinancing of existing legacy instruments. Only really the Singaporean and the Malaysian banks have been big issuers of AT1 in the past. So, for example, if the Chinese banks were all to issue AT1 tomorrow you’d be looking at around US$40bn equivalent. Much of this issuance will likely be in the domestic market but as the currency is liberalised these banks may issue in dollars or euros as well, so that’s potentially a big source of supply. In Singapore the banks have already been regular issuers of T1, so it’s more about refinancing and financing growth in RWAs, as it is in Korea and Australia too. In the rest of the region, in countries like Hong Kong, Malaysia, India and Thailand, there is the potential to see much more Additional Tier 1 issuance.
On the T2 side, Asian banks, in most cases, already have more than 2% of RWAs in T2 format – it’s more like 3% – so in the short term they don’t have a need to issue too much more unless they’re issuing for refinancing and, more importantly, to replace instruments being transitioned out. So I’d say over time, the T1 market could see in total about US$120bn equivalent and the T2 about US$160bn, over time and across both local currencies and G3. That compares to a European market that will be about US$500bn – US$600bn. So if the Chinese banks issue in size, the market could be around half the size of the European hybrid capital market.
IFR: In practical terms how realistic are those numbers?
Dominique Jooris, Goldman Sachs: Sean has done the heavy duty work in quantifying that, and the numbers are in the right ballpark. There are a number of issues that we will face and one of them is whether these instruments are tax-deductible, which will materially impact their economic efficiency and that’s still a question mark here in Asia. The second point is that AT1, in particular, is only efficient if its cost is below the return on equity. If, for some reason, you have a relatively low return on equity, that will mean a higher cost to pay on your AT1 and also a cash cost. There is still stigma associated with this concept for a number of issuers, and that will dampen the ceiling number that Sean mentioned.
Another haircut we will have to take is that a number of jurisdictions in Asia are at a senior debt rating threshold that would lead to hybrid T1 instruments being very expensive – assuming there is proper correlation between pricing and rating. One of the things we have forgotten in Europe, because of the single currency, is that there are domestic markets in Asia. Sean mentioned very appropriately the UOB hybrid T1, the first new T1 in Singapore dollars. The first Chinese Basel III-compliant T2 was issued in renminbi, not in US dollars. I think these are strong pointers to the fact that a lot of this supply will be directed to the local market. If you look at the US dollar portion, it will certainly be much smaller than the headline numbers we’ve been discussing.
IFR: On that note if we’re seeing higher interest rates and therefore higher ROEs could it actually reduce the need for AT1?
Dominique Jooris, Goldman Sachs: If banks are profitable, they may be willing to gear up their capital structure a little more, and do so safely and efficiently. But if their internal capital generation becomes excessive, they may not go through this process. It’s an argument that could go both ways.
Jonathan Cornish, Fitch Ratings: And we’re not seeing internal capital generation ramp up from current levels at this stage. If anything, there’s probably more pressure on profitability and hence we see the prospects for issuance being fairly strong. I agree totally with Sean and Dominique about both the potential for issuance and the concerns around that. We have been talking to some of the Chinese banks and clearly they will be the biggest issuers, but at this stage they’re primarily talking about T2 and not T1. To the point Dominique made, if they were to issue T1 instruments then chances are they’d be quite lowly rated. Given their nature, they are much closer to equity and with loss-absorption that kicks in much earlier than T2. Our viability ratings – that is the unsupported standalone ratings – for the Chinese banks are no higher than Double B flat, so you talk about notching from that level, theoretically. The consideration for us is whether or not we factor in state support for those ratings and that will be a big difference in terms of absolute rating and the yield.
IFR: My sense is that the buy-side here always assumes there is going to be full state support, so I’d like to ask the buy-side guys if they expect these instruments to be treated differently than they would be in Europe.
Jonathan Cornish, Fitch Ratings: Unlike Europe, in Asia we have a fairly high prevalence of state ownership in the banking sector, so when it comes to thinking whether the taxpayer or the shareholder is going to bail them out, they’re one and the same. We’re thinking in the case of China – as indeed in other jurisdictions – that support would be more pre-emptive, so while they may have defined triggers in place for T1 already, we believe support could be forthcoming in advance of a regulator wanting to call a bank non-viable. In a few jurisdictions, like China, India and Taiwan, you would probably face considerable concern across the whole industry if one of the major banks was to be deemed non-viable. That’s basically what we’re looking at.
Sumit Bhandari, BlackRock: If you take a step back, this whole conversation about regulators providing clarity on what the point of non-viability might be is really not there. For right or wrong reasons, regulators have stayed away from talking about it. Looking at other investors it definitely feels like people aren’t willing to consider that risk and price a premium for it. While I agree that there are certain jurisdictions where it will be more forthcoming than others, I think from an investing standpoint it’s very difficult to price that risk appropriately in terms of basis points or spread and decide which one should have more of a premium than others. That’s our thinking right now on these.
David Lai, Eastspring: In Asia the regulators have tended to be more investor-friendly so far. We haven’t seen many cases of the regulator asking bondholders to bear losses in previous bailouts, so we can take some comfort from that. But things can change. If the norm worldwide is changing then Asian attitudes may also change. So this ambiguity about the point of non-viability is still one of the major concerns from institutional investors’ point of view.
Ben Sy, JP Morgan Private Bank: I think retail investors in particular have a strong belief that the government will support them. So far for the T2 and T1 deals like UOB in Singapore dollars we have seen strong demand from private clients because they don’t believe the point of non-viability will be triggered. They believe, based on their past experiences that the governments and regulators in Asia will be highly supportive. If you look at the spread between the new Basel III instruments and the legacy ones it’s only 50bp difference at new issue, and some of them are trading even tighter than that in the secondary market.
I think, in general, retail investors are a bit indifferent, or they don’t know how to price the securities at this stage given the uncertainty. They don’t believe a Cyprus or Ireland event will happen here, especially for the state-owned Chinese banks in Hong Kong or the Singaporean banks. Personally, I think the pricing is probably a bit too tight for the new Basel III securities versus the old ones.
IFR: Do you see that changing?
Ben Sy, JP Morgan Private Bank: I don’t know, because there’s a lot of retail involvement and they are sometimes not that logical in pricing, they tend to be more subjective. But I think in time the legacy ones seem much cheaper, and the new ones relatively expensive given the point of non-viability risks – certainly the risk is worth more than the 30bp premium we’ve seen in some cases. Having said that, those deals went quite well.
Desmond Lee, Morgan Stanley: One important element is that there is a big home bias among investors here in Asia. Some Asian investors were previously buying a lot of European bank capital securities, so when it comes to buying bank paper from their own their country, they would feel even more comfortable. So if you look at the allocation stats for some of the recent US dollar bank capital you see 60%–70% of that is placed within Asia. That home bias helps these deals price tight, relative to what the convention otherwise might be.
Sean McNelis, HSBC:Just on that point, it really depends on country by country. Certainly the banks leading those transactions would argue that it’s a very big call for a regulator to make a NVLA [non-viability loss absorption] determination on a particular bank, as it obviously impacts the ability of other banks to fund or raise capital at that point in time. It’s a big determination to make – particularly in countries with large financial systems, such as Hong Kong, Singapore or Australia, for example. On the transactions we’ve seen to date with NVLA, out of Switzerland, Hong Kong and now Singapore and Australia, the premiums have been quite low, but it might be higher in some other cases where investors are more concerned about a regulator being ‘trigger happy’, for want of a better word.
Just one point on NVLA; what we did in the UOB transaction for example, is that in most countries – India and Malaysia are exceptions – if the AT1 is equity-accounted, you do not need a numeric trigger at 5.125% of Common Equity Tier 1 or higher, you need only to have NVLA. So in that case the risk for investors is more akin to a legacy instrument anyway – it’s simply a perp non-call five, with non-cumulative coupon deferral. There are no step-ups anymore, but it has no numeric trigger. I think that’s something you may see more of from the Hong Kong and Singapore banks, where they can do it. European banks all need to have the 5.125% trigger, whether it’s debt or equity-accounted, but in most of Asia Pacific it’s not necessary.
And one other point, issuers will tell us there shouldn’t be a premium because it’s already reflected in the price of their legacy instruments, because it’s been proven with the Irish banks, SNS Reaal and others, that governments can retrospectively introduce legislation to impose those losses. If you’ve looked at how the rating agencies have downgraded the legacy instruments, is there really that much of a risk difference between the new instruments with NVLA and those without?
Ben Sy, JP Morgan Private Bank: Sean raises a good point, because from earlier this year most of the private banks have started to raise the risk profile of the CoCos [contingent convertible] bonds or those with a contractual writedown. The Asian ones still do not have these numerical triggers to enforce the writedown, it’s just a non-viability trigger. Most private banks can still market those to their clients, and maybe that justifies a lower premium versus the European issues.
IFR: Have you seen people selling out of the old legacy assets?
Ben Sy, JP Morgan Private Bank: No. What I mean is, given the more stringent regulatory selling process on those with triggers, the 5.125% or 7% Core Equity Tier 1 triggers, those Asian T1s that come without the numerical trigger, with just the non-viability clause, are becoming more acceptable.
Bryan Collins, Fidelity: From our perspective as investors these new instruments are materially different to the old-style T1, so they should be trading at a meaningful premium to the legacy paper – regardless of whether or not there is a 5.125% trigger. Apples for apples is not what’s happening here. Exactly what that premium should be is going to be an iterative process; it’s going to develop over time as we see different structures, as we have seen in Europe. We’ve got multiple jurisdictions for euro-denominated securities, and the nature of how that security fits into the capital structure is actually quite different.
In time, investors are going to have to be more informed about that, and we’re really only just at the beginning of the beginning of this process in Asia. It’s still a very new process, so to try to use the handful of new-style securities out there as a pricing indicator is not appropriate.
On the point of government support, this is an interesting one, because the whole point of these new instruments is that governments don’t have to provide that support. Conceptually, to provide government support for the rating of an instrument that is equity-like, or even junior to equity in some cases, starts to become problematic because these are supposed to bear losses before the government gets involved. For state-owned banks, I appreciate that it’s going to be a bit tricky when it comes to this pre-emptive potential for governments and regulators to step in.
All of these things give these new instruments an extremely different set of risks – certainly very different from the corporate perpetual and very different from the old-style capital securities, and that needs to come through. Asia trades at a spread premium to developed markets, and the same should happen here. If you think about where we are in the cycle, where the US and Europe have gone through massive and painful adjustments and Asia is at the opposite end of that cycle, all of that weighs into the need for compensation. That needs to come through in price and that’s going to be an iterative process.
To the point about the cost, that’s going to be really quite important. If the costs of issuing what will be very much high-yield rated instruments – if they can be rated – are punitive then the issuance numbers we’re talking about are going to surprise very much on the downside.
Sumit Bhandari, BlackRock: From what we’ve seen from our conversations with the issuers so far, basically they’re very price-sensitive. A lot of that is because they can go to the retail market with a 5%–6% handle and that works for retail, which might be looking at yield versus institutions, which might be looking at spreads and sub/senior ratios. If you look at the redemption profiles of some of the legacy sub-debt over the next few years it’s not looking that big. So I think this price discovery process, where you need a broader investor community to be participating in these deals, may be postponed a little bit until a broader set of investors can actually get involved.
We might have to wait a little bit until we find out what exactly should be the right premium for all the uncertainties you all have mentioned, not only around how regulators will behave but also around if indeed they were to impose losses then who are those losses imposed on – particularly if retail is a big buyer of these. My sense is that there are going to be multiple things that people will need to talk about as time goes by, but, in Asia specifically, you might have a delay in price discovery until issuers can no longer rely on retail for these deals.
IFR: I’d like to pick on a point that Sean raised. You mentioned that issuers believe that their old-style notes already reflect a lot of the additional risks. How are arrangers going to fight that perception, because the institutional side clearly doesn’t agree?
Dominique Jooris, Goldman Sachs: It takes two to tango. We don’t have a statistical model that allows us to pinpoint what the risk premium should be, so it will be a price discovery process. But we tend to forget that the discussion around the point of non-viability is only a narrow subset of the discussions we would have had before Basel III anyway. Bankruptcies did exist in the past, but we might have thought that state support gave us a get-out-of-jail-free card. With the new securities, this doesn’t happen anymore.
First, take Lehman Brothers. If you were a senior unsecured bondholder, you would have gotten back, say, 25 cents on the dollar. You would have been wiped out anyway, regardless of the tier of capital you owned.
Next, no regulator but for the MAS has put out a clear definition of what the point of non-viability would be. On Lehman, the point of non-viability was, ultimately, a question of investor confidence – the last quarterly statement before it went bankrupt still showed a healthy level of capital, so 5.125% or a similar figure would not have triggered non-viability. You would not have been able to find any sign other than the fact that it was becoming illiquid and unable to fund. So the first thing is whether a regulator would enforce a point of non-viability or let nature take its course. The second thing is, assuming the regulator had said you are non-viable and you wrote down your T2, would that have saved anything? The answer is probably not – it would have set more alarm bells ringing and worsened the initial problem, which was liquidity.
The third point, and we’ve alluded to that when we talked about the investor base, is that the systemic implication of calling a point of non-viability on a particular institution may actually be significant. Regulators have done a good job of avoiding cross-holdings of sub-debt amongst banks to make sure that there are no falling dominoes within the banking system, but they will be aware of the systemic impact.
The fourth and last thing to consider from an investor standpoint is what happened in Europe at the height of the crisis, where some high-quality instruments such as hybrid T1s, perpetual non-call 10s and step-up structures from AA rated banks in Triple A rated countries were trading at 15–20 cents on the dollar. So the question then becomes how much more painful it would have been to be written down at that point than to sell in the market.
My point is the scenario where the point of non-viability causes a different outcome for an investor than the ones that they were already exposed to is actually quite narrow. This does not go as far as to say you should not ask for any premium, but we need to take a 30,000-foot view of the initial outcomes and realise that this is merely modifying one of many scenarios.
Bryan Collins, Fidelity: I think you explained those various aspects very well, but it just highlights the complexity of these instruments and the highly theoretical nature of everything you’ve just laid out. To have a situation where instruments and regulatory environments and all these various definitions depend on highly, highly hypothetical outcomes just adds to the complexity. It adds to the risk that these instruments are just so fundamentally over-engineered that they defeat their own purpose or become this minor subset of the broader market and, therefore, investors should be expecting some kind of compensation.
Your earlier point alluded to an argument that has been around for a long time and will be there in the future. In the Lehman example, whether you’re senior, sub or whatever, you got back two fifths of not a lot. So, the logical conclusion – which is a dangerous one – is that you might as well go down the capital structure because what does it matter at the end of the day. Now, that is a very dangerous argument that was used in 2006 and 2007 towards bank capital paper in the first place. It’s a little bit dangerous to hear that, because, if you’re not being compensated through 30bp, 40bp, 50bp of additional premium per year, the opportunity cost to not be anywhere near there is quite meaningful.
To continue reading this roundtable, click the relevant section. Foreword - Participants - Part 1 - Part 2 - Part 3
