Ahead of the curve podcast

Counterparty Credit Risk: How regulators are raising the bar in the wake of the Archegos default

Overview
More episodes

In this episode of Ahead of the Curve, John Pucciarelli – Head of Industry Engagement and Stuart Smith, Co-Head of Business Development discuss the paper issued by the Basel Committee on Banking Supervision (BCBS) on the Proposed Guidelines for Counterparty Credit Risk Management. They provide an insight into the guidelines and share their perspectives on the why the regulators felt more stringent controls are required in the wake of the Archegos default.

The paper referred to can be found here:

Guidelines for Counterparty Credit Risk Management (opens in new tab)

Hello, dear listeners, and welcome to another edition of Aheadof the Curve Acadia's podcast. With me today, I have Stuart Smith.He is Co-Head of Business Development here at Acadia.My colleague, welcome. Stuart, this is our third podcasttogether, and it's the third different studiowe've been in. -I like this one, though.I love this; is my favorite of all three. I'm very happy about that.We're going to be talking about counterparty credit risk today.I know it's near and dear to your heart. It's near and dear to Acadia's heart.Just to set the scene, there was a consultation paper from BCBSthat came out in April. It is just a set of guidelinesfor how to manage counterparty credit risk,and it has quite a few different parts to it.I think that's what we're going to focus on today.We're not going to pull the entire paper apart,but we're going to talk about the things that I think are important to usas a business, important to our clients as well.It's a very interesting paper, and I think it's one of these spaceswhere it's a wait-and-see where it goes. I think in the interim,what we can do is digest it and ensure that we understandfrom a business perspective what it's going to mean for our clients,potentially, and how we can help. Why don't we just dive in?I'll ask you the question of why this paper?What does it mean? What's its purpose, and then we can take it from there.What is it? What are we talking about? -I think risks have exploded across banks,and if you go to risk conferences, there used to be a credit stream,a market risk stream, a liquidity stream. Now there are dozens of streams.Cyber risk is seen as the biggest risk among other things.In some ways, counterparty credit risk has been somewhat lost in the background,but in the last few years we've seen many events,traditional counterparty credit risk defaults and like Archegos,which then triggered the default of a household nameand Credit Suisse and other defaults as well, which have broughtthis back into the firm view of firms and regulators.Realistically, after 2008 and Archegos default.It should be able to happen. It's always going to be defaults,but it shouldn't have had the consequences and knock-ons that it had.We thought we'd done enough post-2008 that you couldn't have eventslike a bank, Credit Suisse, being taken down by that event.This, as you can imagine, caused regulators to go away and rethinkand relook at what they're doing and say, okay, have we got everything right?Are the capital rules right? Is the guidance right?This is not about capital. We talked about that a while agoaround Basel III, and the Endgame that's going on in the US.This is on the governance side. This is more of a set of directivesto the bank. They need to interpret, understand,and then implement within their structures applicable to them,and the regulator can check that they're doing it.This is still a consultation. Nothing's final yet,and there's a period for comment open till the end of the summer.This is really about setting up your bank and understanding your responsibilitiesto act in a way that the regulator sees as fit and proper,and hopefully, this time, means we can't have another Archegosthat brings down another big bank. -It makes sense.Anytime an event like Archegos happens, it's always going to triggerthe regulators to rethink, like you said, and I thinkthat's what this paper is intended to do. We already have establishedprocedures and risk counterparty credit risk mitigantsthat already exist in the market, like initial margin.Let's dive into that, and what the potential changes are,or what the recommendations might be. I'll say enhance, not change.Because we do have a foundation. What we have to make surewhen we're talking to people and listening to this podcastis that the consultation paper is clearly not lookingto re-engineer what already exists. To me, it's pointing out thingsthat potentially have weaknesses and to enhance them,maybe potentially make them better. Let's talk a little bitabout that in terms of paper recognizing initial margin,and then the difference between static IM versus a risk-based approach.Worth noting that Archegos wasn't part of the reg IM.They weren't large enough to be posting to SIMM,and therefore, they were probably on a house IA background,and posting probably a static amount against the trading that they were doing.That's called out as being maybe something that's not fit for purpose anymore,and something that the banks should look at in that,either if they're going to use a static measure.That measure may well need to be quite a lot higher,or they should be moving to a risk-based measure,which can better capture some of the risks around complex transactions.They also highlight the case, the case is whereyou're going to see these spectacular defaults,is probably going to be quite linked to posting margin in the future.It's a class of default that maybe we didn't think aboutin the past, where the trigger for defaultingare very large margin calls, and those margin callshave been a self-fulfilling loop. In that case,the margin becomes very important. Having a static measure, I think,really called out as being something that's not the way to go anymore.If you think about the traditional pushbackagainst that, it was always, oh, we can't do that.It's too complicated. People couldn't understand it.People couldn't handle it. They couldn't replicate it.All of those questions have been-- -There's a standard risk-based modelthat's out there, has been used for quite a few years.It's called SIMM. -I think SIMM has said a lotto the industry about what's possible and what biocides are capable of doing,and a lot of those excuses, if you want to call them that,that were used in the past no longer seem relevantand no longer credible to say, oh, I could never cope with somethinglike this. It turns out they can, and they can do it quite effectively.From an Acadia perspective, it's something that we look at in termsof using the SIMM for what we call house IAor independent amount from a non-reg perspective.Can we talk a little bit about that in terms of the appetite for that,and maybe this paper will help firms who are eithersceptical or, like you said, don't think they're able to do something like that.Look at that as an opportunity or a possibilityto do something different, and potentially it could satisfywhat the guidelines are being proposed now from this Basel paper.Definitely. I think firms want to take on a risk-based IA.They've all got a similar challenge, which is, can I get my counterpartyto agree to this? Can my counterparty still challenge it potentially?Can they understand an incremental cost against it of doing new trading?I don't want to put them off to a new trading with mebecause it's too complicated. A lot of firms have stepped in.Acadia definitely one of them has shown that.Yes, absolutely. We can provide a bridge for youto help those firms understand the calculations,whether they are the same or whether they will be somethingelse as well. There's variance.There's also just different metrics and measures,help people understand them, help people understandpotentially incremental costs and optimization opportunities,and then facilitate them going live with thatand using your operationally as well. I don't think we know right now.I think there's a period of time that's going to goby when people decide how they're going to go with this.Maybe they say it's easier just to increase my IAand being more control of it. If they do go down the risk-based, I thinkit'd be an interesting space for vendors again,to get involved and support the buy side in understanding thisand using it day to day. Exciting times.It's interesting to see how it plays out over the summer.See what some of the commentary is back from the organizations.I think the comment period runs through August 28th.At the time of this recording, I think we're at the end of June here.It will be interesting. We'll touch a little bit lateron what we think we'll expect out of this. Because we areconnected to the regulatory community in a nice way.It'll be interesting to see. Shifting gears a little bitto another topic within the paper. Again, we're focusing on the paper today,is scrutiny on wrong-way risk. Now, wrong-way riskcould mean a few different things. I know from my other days,it was specifically spelled out from a collateral posting perspective,meaning. I'll ask you in a second what it means in this context,but from a collateral perspective, specific wrong-way risk is if I'm bank Aand I'm posting bank A shares to my counterparty,that's specific wrong-way risk for obvious reasons. If I go down,so does my collateral that I post to you. It was always an obvious thing,and it makes complete sense, but it was called outin the unclear margin rules globally and was implementedin this guidance paper. They spend a lot of timeon wrong-way risk. It's different from what I just explained.Can you just explain what it means in this context?You've given a couple of examples of specific wrong-way riskwhere there's an intrinsic link. Posting your own shares in equityas collateral against a margin requirement.As soon as you start thinking through the consequences of that.In the case of default, actually, those shares are going to be worthless.This is a very specific case of something that doesn't work.You can come out with lots of different examplesof specific wrong-way risk, not normally quite rule-driven,and they're just things you shouldn't do. It sounds easy.Legal structures are complex organizations,and hierarchies are complex. It can happen.This is something banks are quite good at in applying those rulesand detecting them. What's much harder is general wrong-way risk,where there isn't a direct link, but there's an inferred linkbetween the two. The normal examples, if you go to a textbookand read about this, they'll talk about airlines.If you look at an airline, its chance of survivingand still being in business is intrinsically linkedto the price of jet fuel. As jet fuel price goes up,the chance of your airline making it into next year goes down.You're selling fixed-price tickets against a variable cost.A sensible airline might say, okay, there's too much risk there for me.I'm going to enter into a transaction with a bankto hedge out some of my risk on jet fuel. I might sign a total return swap, say,with a bank where I'm going to pay a fixed amount monthly,and I'm going to receive an amount linked to the price of jet fuel.Then I can sit on a fix, and the bank is taking the riskon the price of the jet fuel. -We're doing the ratesin the rate space all the time. -Exactly. In the bank's case,you're going to get paid out when the price of fuel goes low.That's the case where the airline is definitely going to survive.That's a fantastic case for the airline. All the airlines are going to do great,and this is fine. In the case that the price of jet fuelgoes very high. It's been bad for your bank.You've signed a bad transaction. That has to happen.It goes both ways, but the airline is going to survivebecause you've hedged them out. This seems easy.We'll just align all firms, understand what's goingto cause them to default, and then we'll make surewe're not aligned with that in risk. Of course, when you getinto the reality of it, it's quite a lot harder than that.Maybe an airline does sign one of these towerswith one bank on the street, and then they realize,we've done too much. We haven't sold as many ticketsas we thought we did. I need to unwind a bit of that hedge.They place it in the opposite direction, hedge with a different bank.Of course, the other bank. It looks like they're speculatingon the price of fuel, and they're wrong-way on that towers.It's very hard to understand. Are you right way or wrong wayagainst an organization, especially when you don't know everythingthat's in their book. That's the simple example.Then you go into the really murky world. Murky is the wrong word.Firms like Archegos are holding huge amounts of equityin the same things they're taking punts on.This created a huge wrong-way risk for them. Which,from the bank's perspective, is very hard to detectunless they went and looked at their balance sheetregularly, understood exactly what was there,and their other transactions. It's very hard to seethat that was the risk that you were taking,and that it was the wrong way. -Why is it hard?Is it just because of conflict of interest or privacy,or there isn't a mechanism? Because to me, this is all about--it's transparency. Everybody wants transparency,but when you're talking about placing bets and investing,you don't want to show everything, from what we're talking about here.There's this balance. It's an interesting balance.It's a really interesting space. The more I unpack it in my mind.Two things that make it hard. The first thing that makes it hardis the thing we were just talking about, just getting that informationfrom your counterparties regularly. This is really what your credit officeis there for. You should be talking with that counterparty,understanding their business, understanding their funding mechanisms,and continually making those judgments on where they areand then also informing the quantitative teamabout the types of risks we're worried about.Which is done now, this is normal practice.It's done, but that's a difficult job, especially when firmsmaybe don't want to release some of that data,and you're being a bit more forensic about what you're doingto try and understand it. Then the second half of itis what you do with that information. We talk through a very simple,straightforward example with a corporate. We can normally get those.It's the ones where you get out of that spaceinto the more complex space, then you're reliant on modelsand PFE in general. Potential future exposure,thousands of scenarios. Then I have to somehow turn ona defaulting mechanism within that model. What we've always seen is when you tryand turn on that defaulting mechanism, forcing a counterparty into defaultin the scenarios you're looking at, the computation effortgoes through the roof. What we've seen certainlythrough the 2000, 2010s, a lot of firms implementing PFE,but very rarely implementing anything to dowith default conditioning in there. That means they're not measuringwhat we're supposed to do, which was the exposure at default,which was always what the regulation said. You should measurethe exposure at default. What they were actually doingwas just measuring the exposure in a number of scenariosand saying if it defaulted, it's subtly different,but it's definitely different. The reason they didn't do itis because they just didn't know how. It was such a difficult problemthat I think everyone accepted it couldn't be done computationally.The data wasn't that good at the first place.We may as well do the thing we can do. I think this baby was saying,okay, call time on that. This area of wrong-way riskis so important now, racking up with margin coststhat made it even more important. We need to nail this area.Otherwise, we will see more spectacular defaultswith really colossal losses, and we need to fix thatbefore we go forward. There's going to be a lot of workdone on PFE engines. There's going to be a lot of enhancements made.There's probably going to be new system implementations.If ultimately the scrutiny on that wrong-way side does increase.A lot more requirement of data as well from firmsto understand when it occurs. -I guess it depends on the sizeof the firm and how much tech they've investedand what they're doing now. Like you said, there is a structure,but it's hard and it's costly and it takes a lot of time.I would assume, like from an onboarding perspectiveand an onboarding governance perspective, this is an extra layer that will add on.I would assume that maybe the tier ones might be able to do ita little bit more efficiently. Maybe not.I'm just speculating. What would be needed?If this consultation paper and what they're laying outin terms of wrong-way risk and looking at itthe way we're explaining it comes to pass, where it's either a ruleor it's something that everyone needs to do.Are we looking at new systems to be built, new solutions, or new software?Where does Acadia fit into something like this?Can you speculate on some of that? -Yes. I think a common placepeople have looked at, and I think a lot of bankshave already done this, maybe particularly in the US,they go to a dual limit framework. In the past,they used to have a counterparty credit risk limitthat would say maybe with this most exposure over 10 years,this much over 20 years, and maybe nothing beyond 30 years.Now they're likely to have that same limit plus another limitwith a wrong-way risk type calculation. That's likely to be a larger number.They have what they call a dual limit framework,where the exposure in the free case has to work,and in what we call the default condition, or the wrong way test.I want to stop you there. Is this on the onsetof taking on the book, or is this, I took on the book of business,and going forward, are we talking about more of an onboarding exercise?Yes. Let's take a step back. When your firm comes on boardand says, I want to start trading with you.See, the first thing the bank's going to dois okay, well, who are you? Show me you have some assets behind youto do this. There's a process of onboarding.There's KYC, there's AML, there are all these thingsthat already exist and are rules for that. -As part of that,you're going to agree on some margining approach.Maybe it's going to be a very high house IA.Maybe you're going to do a reg IM because you're already very large.Then, from that, the credit department will work outwhat credit limit you should have, not dissimilarto your personal credit card. This is the amountwe're willing to take with you. The more collateral rules,the better collateral rules that you've got in place.Probably the higher you let that go, the more secure the balance sheet.Obviously, the more you're going to let that go up.Smaller firms, poorer collateral terms. You're going to seethat limit squeeze down. Like we said, now it looks a lotlike it will be two limits. One, just the day-to-day limit,maybe that you can do, and two, maybe periodicallyrun a weekly run of this stressed but default conditioning stressto say okay, right when this counterparty is going todefault, am I still under control of what's there?If you've done the same thing without Archegos,you would have found that wrong-way risk number.That limit would have been orders of magnitude higherthan their credit exposure. They're hoping that this mechanismwill force banks to find that and to fix it,or that they'll just weed out firms that aren't in a good positionto trade at that onboarding step. -There are a coupleof different angles to this. It's not just that I'm going to dothis wrong-way risk exercise, and determinethe appropriate levels of margin. Maybe I just don't do businesswith this type of entity counterparty. Maybe that will forcedifferent behaviours too, that we won't speculate into this,but maybe that's another element as well. Sorry, I interrupted youfor a second. -You also asked about what that meansfor system change. There are plenty of PFE systems out therethat don't have a default conditioning mode.Most systems are going to have to implement something new here.There's a lot of new thinking coming out from regulatorsand other people around the right way of doing this.We saw it at Quant Summit, an interesting presentationaround some mechanisms. You might choose to bringthat mechanism in. There's still a lot of thinkingto do here. Then there's going to be a lot of coding,and then there's going to be a lot of system upgrades.It's going to be an interesting time. Always handling this areawas right for the industry. I think it's going to be greatto see it done, but I think we can expect to seequite a lot more computation load off the back of it as well.There's going to be a lot of work to do. -From your position here at Acadia,is this one of your main priorities in focus right now?Yes. I think we're focused on-- -I know you're focused on a lot of things.Understanding the regulation and seeing how the consultation goes,and seeing what sticks ultimately next year.This is definitely an interesting area, and we're definitely engagedin the quantitative community around the research and understandingwhat's possible and what the new standards might becomeoff the back of this. I think we're still very muchin a thinking phase, but it's definitely an interesting areathat we want to be at the forefront of. -We know what we do best,but I do think that you're right. We'll see how this shakes out.Then, based off of that, I'm sure Acadia will be well-positionedto help in a lot of these areas for sure. Before we end,what do you think the timeline is? We are looking at next yearto see the final paper. I would assume so if the comment periodruns into the end of August in 2024. Usually, they have to digest all of thatdepending on how many responses there are, and I know there are a lotof industry efforts underway to do that. Acadia is always,when is appropriate for us. Sometimes we'll join in on a response,or we do our own response, which we've done before.I think largely this seems positive. -Generally speaking.From our perspective. -You read the documentand maybe say that 90 per cent of this is common sense.This is surely what all banks should have been doing.It's good to write it down, though, 10 per cent of it.It's definitely pushing beyond maybe what standard is today.Definitely in particular geographies. I think it's a positive step forwardfor the industry. More focus on not allowingvery large losses to happen like that. It makes the financial industrymore stable. -It's all logic.I think there's a lot of thought put into it as well, and lessons learned.It's going to depend on what the feedback is.Of course. That's the point of having a consultation.Thankfully, it's not like passing a capital rulewhere things have to pass into law and ultimately affect directly.This will happen much quicker than that if it does go forward.Individual regulators can choose to adopt this governance,and then they can apply it to their local regulationsand local requirements of banks quite quickly.I don't think we'll have a 10-year wait, as we have for some of the other metricsto go through for this one, and the next year,maybe, is a realistic time. -We can't compare Basel III to this.Basel III is huge, and it touches every part of the economy.This is specific to our world in OTC derivatives.I guess we'll wait and see. I'm going to assumethat we'll probably do another podcast to talk aboutwhen a final proposal comes out and what it's going to mean for Acadia,what's going to mean for our clients, and, of course, always, how we can help.Stuart, anything else before we sign off? Anything else you want to say?I guess it's just reflecting on what Acadia is here for.You can get lost in the details of what you do sometimes.Acadia was set up off the back of the financial crisisto reduce counterparty credit risk. In some ways,this is just a great space for us to be in,and we get to continue on that journey of saying,okay, we've dealt with this area, we've dealt with this area,this feels good. This area turns out we haven't dealt withas well as we thought. How can we help solve this next area?I think it's just a journey. It's a journey that Acadia'sbeen extremely successful at. Excited to be part of it again,for the next round. -This is our podcast,so we'll do our own horns. That's what we do.We solve problems. Thank you so much.Pleasure as always. I'm sure we'll be back in this wonderful studio again.Thank you very much. -Thank you all for joiningin this episode of Ahead of the Curve. I hope you found it interesting.I know I did. If you want to find out more about this podcast and other podcasts,you can find us on acadia.inc or any of your other streaming services.Thank you so much for joining us, and we'll see you again soon. Thank you.