Suisse finished — how did a G-SIFI collapse?
- Owen Sanderson
In 9fin's podcast, recorded on Thursday, I described Credit Suisse going down as "a huge deal and a massive global disaster…we’d be in crisis territory”, and questioned how the AT1 bonds had been left untouched in a bank that required CHF50bn of extraordinary liquidity support.
The nature of reporting on a banking crisis is the tendency to be overtaken by events; as of Monday morning, Credit Suisse's entire stock of AT1 is nothing more than a charred heap of ash and an epic source of lawsuits. Credit Suisse itself has been taken over by arch rival UBS, not quite for a song, but on generous terms.
How did it come to this?
That's a much harder question to answer for Credit Suisse than it is for the ill-starred Silicon Valley Bank. There were no losses hiding in plain sight. Credit Suisse had enormous reserves of liquidity (144% liquidity coverage ratio) and sky-high capital ratios (14% CET1). It raised CHF4bn in capital in December! It hedged! By any reasonable metrics this was a strong and stable bank.
But thanks to a series of scandals, stupidities and what the kids call "self-owns" it had become almost a reverse meme stock, a bank which the wits of Twitter dubbed "Debit Suisse" for the succession of fines, losses and screw-ups. The annual reports of most investment banks have a darkly amusing section on active litigation, and Credit Suisse’s is a chunky 12 pages.
The post-GFC shakeout meant billions in penalties relating to RMBS, mis-selling, reps and warranties; the period shortly afterwards saw the Libor scandal and a host of other ugly cartel-like problems wheeled out into public view, costing the overall banking sector billions in regulatory fines. FX, SSA trading, every species of Libor fix, odd-lot corporate bonds, Treasuries, CDS manipulations. As well as the fines and investigations, most of these also give rise to civil lawsuits or class actions, pushing the final bill up further.
Credit Suisse stacked up some idiosyncratic failures as well. The "Tuna bond" debacle in Mozambique, the inverse volatility ETF XIV, the deep involvement of asset management unit CSAM in Greensill Capital, and the failure of Archegos Capital Management, which hosed CS harder than any of its competitors, caused a string of huge losses. The bank also spied on its own head of wealth management Iqbal Khan. Whether this is bad luck, bad judgement or bad risk control is moot; perhaps each individual issue was explainable away, but the cumulative impression was a badly managed bank that was out of control.
It wasn’t just the bad impression these incidents left — the bills for fines and failures left a real financial mark. Archegos cost CHF4.3bn. Credit Suisse took a further CHF1.5bn of litigation provisions in 2022, though it also settled several cases. This Twitter thread is an excellent place to get up to speed.
In truth, post-crisis Credit Suisse never really found a niche. It was really at least five institutions — First Boston (CSFB), the investment bank which dominated the Eurobond business in the 1980s, Donaldson Lufkin and Jenrette (DLJ), a hard-charging leveraged finance and latterly securitisation investment bank, plus a global private bank/wealth manager, a giant Swiss domestic bank, like a Lloyds by Lake Zurich, and a credit-oriented asset manager, CSAM.
There was really no reason for these institutions to be under one roof, but successive generations of CS management tried to invent one, while simultaneously trying to unlock the jewels within.
So there was a scheme to float the Swiss retail bank separately, initiatives to stuff wealth management with the abstract creations cooked up by the credit structurers in markets (synergies!), initiatives to lean on wealth management to get their entrepreneur clients to give CS some IPO mandates and so forth.
The logic of breakup has taken hold more completely of late.
Credit Suisse First Boston is to be spun out completely as a boutique investment bank run by former CS board member Michael Klein. Securitized products has been sold to Apollo and rebranded Atlas SP Partners. The distressed debt desk was selling its portfolio; European leveraged finance was heavily pruned back. Govvies were already long-gone.
The bank’s “reverse meme stock” role reached a crescendo late last year, with a long stock slide and waves of deposit outflows in October in particular. According to the bank’s annual report, it saw customer deposits down by CHF138bn in the fourth quarter, and, as it noted, “a failure to reverse the outflows and to restore our assets under management and deposits following the developments in the fourth quarter of 2022 also could negatively affect our ability to achieve our strategic objectives, including as to our capital position”.
The bleeding was arrested, for a time, by a fat capital increase — another CHF4bn — announced at the end of October and executed in early December. This, in contrast to SVB’s rescue capital, was done the right way, with a big anchor investor, Saudi National Bank, pre-commitments for CHF1.85bn from others, including the already-invested Qatar Investment Authority, and a hard underwriting from the banks.
This wasn’t great for the stock — Credit Suisse ended the year at a price to book of around 0.25 — but it did get a lot of cash in the door, and the newsflow dampened a little as the new year dawned and the market backdrop became stronger.
So what actually blew up the bank?
As with SVB, a simple loss of confidence was the issue.
Credit Suisse released what will go down as the worst-timed accounting issue in the history of banking, saying on March 9, as SVB stock was already plunging, that it would “delay the publication of its 2022 Annual Report and related Annual Report on Form 20-F following a late call on the evening of March 8, 2023, from the U.S. Securities and Exchange Commission (SEC) in relation to certain open SEC comments about the technical assessment of previously disclosed revisions to the consolidated cash flow statements in the years ended December 31, 2020, and 2019, as well as related controls.”
This release finished with the announcement that “the 2022 financial results as previously released on February 9, 2023, are not impacted by the above”.
But impressions matter when there’s a bank run taking hold. The headline numbers — CHF14.9bn in revenue for 2022, with a CHF7.29bn loss — were indeed identical, though, y’know, did include a CHF7.29bn loss for the year.
On the day that SVB was taken over by US regulators, Credit Suisse stock was down 14% on the year. As nervousness ripped through markets, Credit Suisse took a further beating, to the point that by the close of last Wednesday, we were 35% down on the year. Some CHF45bn of tangible book value could be had for under CHF8bn.
The real pain could be seen in the credit default swap market, where the market reached 1400/1600 bps for front end protection during Wednesday. Here’s CDS don Boaz Weinstein (a former DLJ trader!).
Around this point came a second spectacular unforced error, one of the most value-destroying interviews in history. Bloomberg TV interviewed the chairman of Saudi National Bank, which anchored the recent capital raise and held 9.8% of Credit Suisse, about possible support.
He said there would be no further support, pointing out that becoming a shareholder with above 10% of a bank triggered a different regulatory regime of which Saudi National Bank wanted no part.
But in the crisis environment of Wednesday, Credit Suisse needed its largest shareholder saying “absolutely not” to a question over support like a hole in the head. Some strategic ambiguity would have been the smart move; “absolutely not” was like chucking a bucket of gasoline onto the briskly crackling fire of bank panic.
Through Wednesday, chatter also emerged that some trading counterparties were cutting their lines with Credit Suisse, and at that point it looked kind of over.
The issue is not so much that Credit Suisse needed these counterparties for funding, but that the bank appeared to have little purpose left. What’s the point of an investment bank that nobody will trade with, paired with a wealth manager which nobody trusts? Even if Credit Suisse’s liquidity buffers could withstand the chaos, it was clearly headed for trouble.
Facing the end
After market close on Wednesday, the Swiss National Bank stepped in, asserting that “the problems of certain banks in the USA do not pose a direct risk of contagion for the Swiss financial markets”.
This was wishful thinking of the most egregious kind, but it also came alongside a massive CHF50bn liquidity facility, which allowed Credit Suisse to launch a $3bn-equivalent bond tender. Later reporting from the Financial Times pegged deposit outflows at around $10bn per day in the back end of last week, which is definitely crisis territory, but could have been covered by the backstop, and by the huge liquidity buffers the bank already held.
Thursday was therefore a better day, but the crisis was back on by Friday — and over the weekend the talks with UBS kicked into high gear. The leaks came thick and fast, presumably as negotiating leverage; at various points UBS was said to be demanding a backstop related to CDS spreads, allowing it to back out using a “material adverse change” clause if its spreads widened more than 100 bps. It also wanted a regulatory backstop for future Credit Suisse litigation losses, under the “never only one cockroach” theory.
Neither clause made it to the merger press releases, but the paltry CHF50bn backstop announced on Wednesday was juiced up to CHF200bn in the form of a “liquidity assistance loan with privileged creditor status” for Credit Suisse and UBS, and the option for the Swiss National Bank to grant Credit Suisse a “liquidity assistance loan of up to CHF100bn backed by a federal default guarantee”.
UBS also said it has CHF25bn of “downside protection” to “support marks, purchase price adjustments and restructuring costs, and additional 50% downside protection on core assets”.
So it has bought CHF45bn of theoretical book value, along with CHF25bn of downside protection, and a CHF16bn recap (more on the AT1 below)…for just CHF3bn in shares. One can quibble about exactly where these marks should be, but it’s hard to argue with UBS chairman Colm Kelleher’s assertion that “This acquisition is attractive for UBS shareholders but, let us be clear, as far as Credit Suisse is concerned, this is an emergency rescue. We have structured a transaction which will preserve the value left in the business while limiting our downside exposure.”
Is there just a touch of triumphalism in the tone?
The most darkly comical element was the fact that Credit Suisse’s release said that the two banks were to merge, while UBS phrased it as “UBS to acquire Credit Suisse”. Credit Suisse corporate communications head Dominik Von Arx only joined CS once the trouble had started in 2021, after 26 years with UBS….now he’s heading back.
UBS offered a few details on its plans for the group — it’s most interested in “acquiring Credit Suisse’s capabilities in wealth, asset management and Swiss universal banking”. As far as the investment bank goes, the news is not good for the bankers remaining at Credit Suisse.
“UBS Investment Bank will reinforce its global competitive position with institutional, corporate and wealth management clients through the acceleration of strategic goals in Global Banking while managing down the rest of Credit Suisse’s Investment Bank.”
The most controversial part of the takeover concerns the Additional Tier 1 bonds issued by Credit Suisse. Additional Tier 1 (AT1) was cooked up in a regulatory lab in the aftermath of the financial crisis, and really ramped up from 2014 onwards, as successive European countries sorted out the tax issues of the asset class.
The idea was to fix the failures of pre-crisis bank hybrid capital, and build a form of loss-absorbing security that could actually absorb losses. Crucially, AT1 was supposed to absorb losses on a “going concern” basis — that is, while a bank was still functioning. So these were “fixed income” securities with a coupon, but a coupon that could be skipped without default, were perpetual, with no call incentive built in, and in which skipped coupons would not roll up. Once bank capital ratios hit a certain pre-defined level, the AT1s would “trigger”, either converting into equity or simply being written off.
As well as the predefined capital-based triggers (which were either high or low, meaning 7% or 5.125% CET1; these numbers were furiously fought over in the years 2010-13), there was a further trigger for a writedown/conversion — a regulatory judgement that a bank had hit the “Point of Non-Viability” or PONV (pronounced Pon-V).
AT1 effectively gave banks a prebaked capital raise — but in practice all banks cranked the CET1 ratios well in excess of the capital-based triggers, so “PONV” triggers were all that really mattered. There were other wrinkles over the coupon; an unprofitable bank might struggle to service AT1 coupons, which are treated similarly to dividend payments and limited by a regulatory concept called the “Maximum Distributable Amount”. It was this that led to the panic over Deutsche Bank’s AT1s in 2016.
The particularly painful part of the UBS-Credit Suisse takeover is that the entire Credit Suisse AT1 stash has been written down, at the instruction of Swiss regulator FINMA — but shareholders have not been zeroed (they’re getting UBS stock worth CHF3bn). This is not a great price for a bank with a tangible book value of CHF45bn, and a market cap of CHF11bn at the beginning of the year, but nor is it nothing, and the CHF16bn of AT1 write-down is essentially a straight subsidy to UBS for buying the bank.
AT1 is supposed to sit above common equity in a bank’s capital structure; investors are outraged. Here’s TwentyFour Asset Management:
“We believe that the Swiss regulators have ripped up the rules for investing in a company. Equity is obviously the most subordinated security in a capital structure, nevertheless, Credit Suisse equity holders got bailed out by UBS (along with government guarantees) while AT1 debtholders are written down to zero.”
That’s a little punchy; lots of corporate workout situations end up with equity keeping some value and mezz/junior debt getting toasted, one way or another. In the only other AT1 writedown, Banco Popular Espanol in 2017, equity got €1 as part of the Santander takeover, so the hierarchy was respected there.
The counter-argument is essentially “look at the bond docs”, which do indeed give regulators a very wide discretion to do what they please with AT1. But for sure there will be lawsuits over this decision — Goldman’s traders already started making markets in AT1 claims.
Regulators elsewhere are also questioning the Swiss decision.
EU banking authorities put out a joint statement on Monday stating “the reforms recommended by the Financial Stability Board after the Great Financial Crisis has established, among others, the order according to which shareholders and creditors of a troubled bank should bear losses.
In particular, common equity instruments are the first ones to absorb losses, and only after their full use would Additional Tier 1 be required to be written down. This approach has been consistently applied in past cases and will continue to guide the actions of the SRB and ECB banking supervision in crisis interventions.
Additional Tier 1 is and will remain an important component of the capital structure of European banks.”
Through Monday, AT1 instruments, which are present in the capital structure of all large EU and UK banks, many large Asian banks, and, uh, the one remaining large Swiss bank, have been selling off sharply, and there’s been lots of panicked commentary declaring the end of the asset class.
As perpetual instruments, there’s no real refinancing risk for the banks that have already issued AT1 — but this is going to hurt a lot of bondholders.