Thread by @jeuasommenulle

Who am I to say no to @fwred? So here it is!

https://twitter.com/fwred/status/1222064474775080960

@fwred The ECB (SSM) just released its annual SREP results. As a reminder, it’s the annual assessment of risks large banks are facing with the corresponding capital requirement add-on (called the “P2R”). Lots of very useful info this year.

@fwred Imho, the big picture is this: NPLs are the worry of the past. Now the SSM will focus on governance, OpRisk and profitability. The profitability issue will obviously feed the thinking on the ECB’s strategic review as <0 rates are a culprit.

@fwred OpRisk is a also a big issue for two reasons: managing OpRisk is very expensive (compliance costs etc.) and the industry hopes that Basel 4 implementation in the EU will water down some expensive Basel provisions on OpRisk.

@fwred This includes some important choices to make on the OpRisk RWA formula parameters. Not too sure about the EC’s leniency on that, now. (Again: this is a BIG issue, number-wise.)

@fwred Looking at the report now, step by step. First, capital.

@fwred As usual, you guys want to know who the bad boys are. Well, we’ve got 7 of them this year! One bank refused to disclose its P2R (and this can’t be good) and six had capital below their P2G.

@fwred What’s P2G? Very good question. It’s the Pillar 2 guidance, which is an expectation from the SSM that the bank will keep capital above that level, but not a legal requirement.

@fwred The difference is important, because if you are below the P2R you face distributions constraints (e.g. no bonus or dividend) whereas you don’t face any if you are below the P2G – apart from feeling the breath of the supervisor on your neck.

@fwred Last year, only 1 bank was below the P2R – and we had no data on P2G. This year here’s the P2G chart with the 6 weakest links, but it’s impossible to know who they are because P2G are not public.

@fwred What else do we learn from the capital bit of the process? Not much. Capital requirements have barely changed, global P2G at 1.5% of RWA and global P2R at 2.1% of RWA.

@fwred The lowest is 0.75% & the highest is 3.5% and this year we have a nice tool to get individual data easily. THANKS @ecb

@fwred @ecb The report also shows a split of P2R/P2G by business model and I believe this shows how the governance of the buffers is totally broken.

@fwred @ecb What do I mean by that? The SSM sets the P2R/P2G but after that, other (domestic) authorities set systemic buffers or countercyclical buffers. And the results don’t really make sense. Here’s why.

@fwred @ecb That’s the requirements before other authorities set other buffers, depending on the business model. Hum sorry, what? Asset managers or custodians need more capital than G-SIB? Or small retail lenders? Seriously?

@fwred @ecb Well, maybe that’s because the SSM reverse engineers what other authorities will do, and then decides to get everyone at the same (ballpark) level? (see chart below.) And honestly you can’t blame them when you realize that…

@fwred @ecb …Small domestic lenders have, on aggregate, a SYSTEMIC BUFFER, TWICE as large as large corporate/wholesale lenders! This is absurd. And it’s only because it’s different authorities with different approaches.

@fwred @ecb So here's one big topic for future work on the banking union…

@fwred @ecb So if capital is ok, what is the SSM unhappy about? Because they are unhappy. See this (SREP score = 1 is the best, 4 is the worst). This is RWA weighed and the numbers are going down quickly: from 71% of SREP 2 to only 49% now!

@fwred @ecb And there are 3 reasons A) GOVERNANCE B) OpRisk and C) Profitability.

@fwred @ecb On governance, it’s always hard to know the specifics, but the numbers are a bit scary. 30% of specific measures requested by the SSM (i.e. improvements required) were about governance. And the problems are broad based, as you can see.

@fwred @ecb Seriously, this governance problem is getting out of hands and I predict some serious soul searching in many bank boards. Only 18% (!!!) have a governance SREP score of 1 or 2! From 58% in 2017 there are now 76% with a score of 3 !

@fwred @ecb Switching to profitability, the SREP scores are stable, but the business model split is fascinating. I’m sure you all know which bank is a GSIB with a profitability SREP score of 4 😊 (maybe starts with a D)(and they must feel lonely in that sea of yellow)

@fwred @ecb …But the SSM is also very worried at the profitability of… custodians! And that’s a big issue, because it’s a high cost, low fee, all-about-scale business. If you work in M&A you know where to look for a deal.

@fwred @ecb But the one chart I love the most, is this one: return on equity, vs. cost of equity as reported by the banks themselves.

@fwred @ecb This is fabulous. Not only are most bank below or WAY below their cost of equity, but some banks are delusional enough to believe their cost of equity is ….1% 😊 Honestly, ROTFL.

@fwred @ecb Hey, you, at the very left of the chart, I can’t wait to see you trying to raise equity. (Or maybe you’re a German bank and then I can’t wait for the EC to finally investigate that market…)

@fwred @ecb Final problem: Operational Risk. As I said, this assessment is important because it could feed into the choices the EC makes to implement Basel 4. Again, the scores are deteriorating rapidly and a key reason seems to be conduct & AML.

@fwred @ecb Full quote: "1. “Some significant institutions underestimated their 2018 losses in the stress test even in the adverse scenario.”
OUCH 😊. Their stress test modeling was so bad that they took a bigger hit in 2018 then in the adverse scenario!!!

@fwred @ecb Then this nice chart, which confirms that conduct is 50%ish of Op Risk and gives a nice sector estimate (for SSM banks) close to 18bn€/year. Not small change.

@fwred @ecb The two areas of the report that you should not really care about are 1) liquidity (because the conclusions were basically in the liquidity stress test report – on which I had a thread.) and 2) NPL.

@fwred @ecb Why not NPL? Because clearly everything is on track and the SSM is very happy about it. Global NPL have gone down from 1000bn€ in 12/2014 to 543m€ on 9/2019 and the SSM expects 143M€ by FY2021.

@fwred @ecb So all good on this front? Clearly, it’s not a worry as it used to be, but I’d still have a detailed look at this very interesting flowchart of NPLs in 2019.

@fwred @ecb Expected flow for 2019 is -57M€, so all good? Actually the quality is a bit worrying: sales are -37m€, repossession of collateral is -6M€ and cures -20M€, so -63M€ in total, more than the actual outflow.

@fwred @ecb And those are not “good” ways to bring the numbers down, because in many cases, the NPLs are still there! It’s just someone else holding them or a different accounting category.

@fwred @ecb This is obvious for sales, repossession is just the same risk booked differently… and cures can be anything from a real cure to a maturity being extended, with a grace period (and being careful to meet EBA requirements.) So please do keep an eye on this!

@fwred @ecb That’s all, folks! Now you can relax and re-watch that Federer game.

@nicolas_veron