CASA hits target with dual-trigger

Crédit Agricole’s AT1 transaction carried new structural features and targeted multiple investor bases. Neil Day spoke to Vincent Hoarau, managing director, head of FIG syndicate, Crédit Agricole CIB, about the challenges of pricing and executing the innovative structure and handling a record US$24.5bn order book.

Vincent Hoarau, CACIB

How challenging was it to value this inaugural temporary write-down loss-absorbing instrument?

The market has evolved constructively since BBVA printed the first CRD IV-compliant multi-trigger AT1 trade in April 2013, but there are still a lot of controversies in the price discovery exercise. It was therefore a fairly iterative process.

Whether the second trigger requires a premium of 15bp, 25bp, 50bp… or nothing… was one of the key elements we needed to find out. We obviously looked at the distance to triggers, the coupon payment and cashflow schedule, as well as the nature of the callability structure to take a view on pricing.

But there was no scientific recipe for coming up with the right number for initial price thoughts (IPTs). It was a slow process, while the parameters can also change dramatically in dynamic markets.

Above all, people had to understand the structure of the Crédit Agricole Group and how the issuer, Crédit Agricole SA (CASA) is positioned. And the logic says that if, as an investor, you think that the bank has a more than significant probability of hitting the trigger you don’t buy the AT1 instrument whatever the parameters and coupon level look like. So it is all about the relative value scheme and analysis.

In terms of pricing references, the direct comparables in the secondary market were the Société Générale single-trigger (5.125%) US dollar perpetual non-call 10 issued at 7.825% in December 2013 and the perpetual non-call five issued at 8.250% in August 2013.

In terms of pricing rationale, apart from Société Générale, did you look at anything else?

Some investors, mainly some UK hedge funds, worked on a pricing rationale from outstanding old-style Tier 2-hosted subordinated issues. Looking at various reference points in the market, they saw plain vanilla Tier 2 with five year call protection trading at around 350bp over US Treasuries. Using old-style Tier 2 and AT1 relationships, they came to the conclusion that the add-on in yield was around 200bp. Going from perpetual non-call five to perpetual non-call 10 means another 100bp extra. Then, while they think coupon deferability is priced in the 200s, they valued the dual premium at 20bp or so. This implied a yield in the low 8% area or thereabouts for the inaugural Crédit Agricole US dollar AT1 in perpetual non-call 10 format.

Elsewhere, it was impossible to rely on the relationship between US dollar Tier 2 CoCos and AT1s in the market to construct a useful pricing rationale. The spread differential between Barclays US dollar Tier 2 hosted CoCos (BACR 7 ¾ 04/10/23) and Barclays US dollar AT1s (BACR 8 ¼ 12/29/49) was around 100bp including some curve and loss-absorbing feature adjustment. Crédit Agricole Tier 2 hosted CoCos (ACAFP 8 ⅛ 09/19/33) were trading at around 5.30% ahead of bookbuilding. Using the above relationship, Tier 2 to AT1, we would have come up with something completely inconsistent with Société Générale’s trading levels.

How does the dual-trigger structure play into valuations?

Crédit Agricole Group is a co-operative banking group. Its 39 regional banks and central body, Crédit Agricole SA, benefit from a cross-support mechanism. This was the central element for the valuation of the dual-trigger. The low trigger at issuer level cannot be breached before the high 7% trigger at Group level gets hit. These mechanisms were emphasised in the investor presentation. This was also a focal point during the 52 face-to-face meetings and various group meetings the funding teams ran across Asia, Europe and the US between 8 and 14 January.

The buffer of equity that CASA would have to consume before its 5.125% trigger is hit is Eu10bn, or 3.2%. At the Group level the buffer is Eu21bn, or 4%, before the 7% trigger is breached. That compares with a 5.7% buffer for Société Générale’s two AT1s in dollars with 5.125% triggers. The equity buffer of Société Générale is Eu18bn. So we looked at the instrument at the Group trigger level, and outlined systematically the greater equity buffer compared with that of Société Générale.

What other relevant features are there in the structure?

The structure offered many other investor-friendly features that you might not notice at first glance. The CASA AT1 cannot be called before it has been fully written up in the unlikely event of a writedown.

In addition to that, the coupon structure is fixed rate per annum, but in each case payable quarterly, so the cashflow schedule looks better for investors and offers an add-on in yield compared with semi-annual payments. Furthermore, the instrument offered a one notch better rating than Société Générale from Fitch.

Elsewhere, the dual-trigger adds a lot of consistency in the hierarchy of the capital structure. It was important to be in line with the CASA Tier 2-hosted 20 non-call five high 7% trigger CoCo launched in September.

Finally, many investors also considered the retail footprint and the limited level of exposure to the CIB business.

How else did investors’ views play into the transaction?

During the roadshow we purposely didn’t say whether we were looking at a perpetual non-call five or a perpetual non-call 10 until the day before pricing, because we wanted that to be determined by investors. And we told them that.

We are seeing a general hunt for yield and this logically crystallises throughout investor meetings. The preference was for the perp non-call 10.

With regards to the main feature of the trade, many buy-side accounts said that we introduced complexity with the dual-trigger and that we had to pay a premium for that. This was not a surprise, so the challenge was to compress that premium through convincing meetings and calls, emphasising all the above-mentioned selling points.

In meeting after meeting investors gave feedback via quantified indications of interest on perpetual non-call five and perpetual non-call 10. Numbers surfaced in the context of high 6% and low 8%, respectively. The main two secondary reference points, Société Générale 8.250% 11/49 and Société Générale 7.875% 12/49, were trading at around 6.45% and at around 7.75%, respectively. This implied a new issue premium of 15bp-20bp and a dual trigger premium of 15bp-20bp, including some minor credit adjustments.

How did pricing evolve during bookbuilding?

The roadshow ended on Tuesday 14 January and we formally announced the call format that day with a view to starting bookbuilding on Wednesday morning in Asia first. We avoided announcing guidance before Asia opened because we wanted to communicate the biggest possible book when marketing into the US started. We already had a shadow order book of US$4.5bn from about 150 accounts out of the UK and US on Tuesday.

The IPT level was set at 8.125%-8.375% ahead of Asia opening. We decided to give a hard number although we could have started at “low 8%”. It’s all semantic, but that way we could respond to some criticisms from investors over the lack of guidance in pricing definition during some other bookbuilding. With IPTs of 8.125%-8.375% we demonstrated that we would consider all pricing indications reflected by institutional accounts during the roadshow. At this guidance the door was also technically open to pricing inside the 8% mark.

When we opened books in Europe, we were in a position to communicate a US$7bn book very quickly and we fed demand further in Europe. When the US session opened, US$20bn was already in the book. So few US investors were willing to miss out on the extraordinary performance in primary and the book enjoyed another key round of momentum.

Guidance was refined smoothly to the 8% area and we announced that the deal would price within a plus or minus 0.125% range. We closed with an order volume of US$24.5bn and nearly 900 different participating accounts on board. Half of the orders were made with tickets of US$10m or lower, showing an exceptional level of granularity.

The pricing level was set at 7.875%, suggesting that nobody demanded a premium for the dual-trigger — not because there is no differentiation between structures, but because market participants ignored the CASA AT1 low trigger. In the meantime they valued the fact that the buffer to trigger in absolute terms was higher for the Crédit Agricole 7% Group trigger than for Société Générale’s 5.125% trigger.

The margin for the reset after five years and every five years thereafter if not called was fixed at 489.8bp during the pricing call. The final size was discussed at length as the Group cares about secondary market performance. But the book supported a good size, so we decided to go for US$1.75bn.

How did you go about allocations given the size of the order book?

Allocation was challenging given the number of individual orders in the book and the level of oversubscription: Crédit Agricole enjoyed one of the biggest order books for a contingent capital trade, if not the biggest. But if everyone was made equally unhappy, people recognised the strong performance of the deal and enjoyed two full points of profit straight off the break.

The performance of the trade in the secondary market was another key objective achieved by the group — it was Crédit Agricole Group’s inaugural CRD IV-compliant AT1 and it had to be a success in primary as well as in the secondary market.

US and Canada-based accounts took half of the total, with the UK taking 22%, and Asia 8%. Asset managers bought 61%, banks and private banks 15%, hedge funds 14%, and insurance companies and pension funds 10%. The deal enjoyed a phenomenal number of new investors for the asset class. More traditional asset managers got engaged because their mandates now enable them to buy any type of deeply subordinated transaction. They are all betting on further strong compression across the capital structure in a context of a normalisation of the markets post-crisis while interest rates remain low.

How do you expect the investor base for AT1 transactions to evolve?

Looking ahead, as we go from one deal to another, the level of oversubscription is becoming misleading. Investors now expect AT1 transactions to be heavily oversubscribed and to perform in the secondary market. There is a greater number of investors who are inclined to inflate orders. The contribution of Asian investors pre-allocation was 15% (US$3.6bn). But you need to take into account the level of inflation coming from Asian private banks, although there was no rebate offered on that deal. Staying with the distribution into Asia, it represented 135 different participating accounts and 80 tickets of US$10m or less.

So this was an excellent outcome given the fact that the participation of Asian investors in deeply subordinated instruments was fairly volatile in 2013. The Asian bid tends to be determined more by absolute yields, which dropped significantly in 2013, while the European and US bids are driven by swap spread and relative value analysis.

Looking at the demand more globally, the UK investor base remains instrumental for the growth of this asset class, because of their capacity to buy dollars as well as euros in size. The euro market investor base is getting more mature every day, but in terms of investor liquidity and depth of the market for this instrument, the US dollar market is the biggest. It will remain that way for a while even if issuers can get success and print in size in euros. And Crédit Agricole seriously considered the euro road.

One of the reasons why the dollar market is bigger is also because you cannot rely on German demand being there in size and granular for a deeply subordinated transaction. This will change and it’s just a question of time and education. Only 30 German/Austrian accounts participated in the Crédit Agricole AT1 trade for a total allocated amount of US$37.8m. In contrast, US demand is highly developed — roughly 50 high profile real money accounts with pretty chunky orders in the CASA AT1.

For more on Crédit Agricole’s AT1 transaction, read:

Crédit Agricole points to AT1 future

CoCos: View from the buy-side