My relationship with Clear has been interesting.That is how we'll put it. I don't buy that everything needsto be cleared in foreign exchange. That's why I'm quite fascinatedwith smart clearing, which you've done a recent proofof concept on with Forex Clear. -Yes.Can you talk to us about what Smart Clearing isand why you developed it? -Yes. We started about 18 months ago,and we've been developing with Forex Clear.The idea is really about innovating FX trades or riskvia new trades into the clearing house to achieve counterparty risk reductions,but certainly also resource reduction across both clearedand bilateral. Really, a nutshell is about gettingall the benefits of clearing without necessarily getting the cost,or certainly limiting the cost like margin.As you know, FX clearing has really lagged compared to interest rates.I think the main reason is that there has never beena mandate to clear, but also, in [inaudible 00:01:07-00:01:08],FX forwards have been exempt. Therefore, the cost versus benefitshas always been in a bit of a balance, and it was probably seenas too expensive in the past, but with SAC I thinkthat had really tipped it in favour of clearingas long as you can control the margin. What weare really looking to do here is, in an intelligent way,put risk into the clearing house if it makes senseto get all of those benefits, but control all the negativesof margin, etc. -How does it work in termsof how flexible it is, because some people might say, "Well,these trades can go into the clearing house,but these can't" and other people might say,"Well, all my trades should really go in there."How does it work? I think one of the problemswith clearing FX in the past has been the huge variety of tendersthat are traded pretty much every day. -Basically, that's where Quantilecomes in. Firstly, we take and look at your starting positionon the bilateral side and the cleared side.We then take all the trades that you have. Then we calculate the capitalof each trade and margin on the bilateral side and the client side.Just to take a step back, the benefits you would get from each tradeand moving them into the clearing house on the capital sidewill be from the marginal period of risk reduction, the SDM treatment,and also that the risk weight at the clearing houseis significantly lower. We see on each trade, for example,risk weight against a part of somewhere between 20 to 100.Then, when we move it to the clearing house,we drop that down to two percent. We calculate that on every single tradeand then look to see which portfolio scanned through all of those trades,all of those risks, can we move in and then get it?The ones where we deem it not favorable, we will leave it.The ones where we do, we will move it into it.Okay, so the decision has made Quantile level in termsof how this works best within the parametersgiven to you by the clients. -It's highly customizable,so each participant can give us what risk they can move maximum,what trades, what currency pairs, and what resources can we increase?How much margin can we increase in a clearing houseso that it is very flexible? It is Quantile controlling it,but is it just us setting up the problem, and the optimizerwill then find it very quickly? The thing here to bring up probablywhich I find very interesting, but optimization makes me excited.The thing where you might say SAC is a risk measure, right?As I said earlier, it tries to calculate the capitaldepending on the risk. The clearing house margin modelis a risk measure. How can we move any capitalinto the clearing house, because they're both tryingto measure risk? If I move a lot of risk into the clearing house,you could imagine the margin increasing. Now, the reason for thatis really a difference in the model. The SAC model, in particular,is very siloed, so each hedging set, each currency pair, is separate.There's no correlation. There's no offset.You can imagine your long euros and short Canadian dollars.In the capital world, they're treated completely separately,so moving those into the clearing house, you could see a reduction.However, in the clearing house, you then have to lookat the asset portfolio, historical analysis,expected shortfall, and far measures. If we weigh them correctlyor optimize the weights correctly, we can create solutionswhere they offset each other in the margin world,the CCP model, and therefore do not increase the riskor the portfolio level. We definitely see no margin increase,and we can see significant capital decreases.The other thing that the optimizer does very nicely,and definitely one of the reasons why we need an optimizer,is that if we just blindly move this whole portfolio of ethics forward,you could see massive margin increases like I justtalked about, but you could also see capital increases.For example, if you have an exotic portfoliofor your loan and you have an FX forward or any other FX tradethat you could clear, that is short, they are offsetting each otherin the SAC world. If you just blindly move that forwardinto the clearing house, you could see the capital increaseon the bilateral side and potentially also on the cleared side.It's those calculations that have to be done within secondsand find the best possible solution on what we can do.Could this have been done three years ago? Was the technology available?It seems quite a big advance in six years to get to the stage right now.Has the advance in technology actually helped drive this?Definitely. Our optimizations have definitely become faster, better,and more powerful ever since we started. We built strongerand more powerful algorithms, and that has definitely helped us.This one is probably one of the most advanced optimizationswe do because it's not just about getting the best capital solution,it's also about controlling margin, taking into account settlementrisk breaches or constraints at the clearing house.It's also about PVs. You can imagine if you have a tradethat's on a Euro CSA, you move it onto a Dollar CSAor the clearing house. There's a PV change.We need to control that. We don't want to give our participantsnegative PNL by participating in this. Then all of these very flexibleand customizable constraints that each participanthas have to be taken into account, so it becomes a massive problemthat we have to solve extremely fast in orderfor participants to be happy with it. -I mentioned earlier, fairly recently,and it's becoming further back. In the past,you've run a proof of concept. Can you give us a high-level feelfor the results of that in terms of the benefits of smart clearing?Yes. We've run two. One is probably just a year ago now.Then the second one was the bigger one, which we ran in the spring,which had 13 banks, 19 entities. The banks submitted tradesthat could be innovated for about $4 trillion worth of notional.The interesting bit here is that we know whereto get only three trillion of that, which really shows that the optimizerpicks the trades that should be moved from a resource perspective.Therefore, it left one trillion on the table,on the bilateral side, because either it hedgedexisting positions-- -It is just a trillion.It is just a trillion. It either hedged existing capital positionsor moving it into the clearing house was just too costlyfrom the margin perspective. Out of that three trillion as well,we could actually compress more than 70 percent of it.Both. Half of that was outright just long shorts on the same day,and the other was using slightly risky compression.However, the main result, of course, is capital here.We saw, on average, more than a 50 percent reductionin capital for each participant. Across the network,it was about 500 billion worth of effective notional.Really significant results that we really think are very good.On the back of that, I think they've got quite a lot of the participants excited,which means that they've invested money in the technology build.I wanted to be able to say today we've gone live,but actually it's in the next couple of weeks we're going live,so look out for that. -By the time this comes out,you may be live. -We should be live.We will keep an eye out. -We have a few banks participatingthat are going live in November. Then in 2024,we have quite a nice pipeline of banks building to it and getting ready for...Okay. It's always nice when you see these conceptual ideasactually come to fruition and actually startoperating in the market. We have a lot of good ideas in the world.They don't all necessarily get to market, which is sometimes a shame.Sometimes it's not. I guess just to close out then Esben,how far into the journey do you think we arewhen it comes to the smart clearing and our optimizationand our better use of capital? How far are we into that journey,and where do you think the innovations are going to come from next?I think we're just at the beginning. Part of the POC,we did a lot of analysis to see, "Where are we going?What is the actual path for smart clearing and also for everything?"One of the things that are very interestingis that at the moment, there is clearing happening,but it's a small fraction of the market. As we build up more and more riskin the clearing house, moving a diversified portfoliointo a bigger, diversified portfolio makes the optimization work better.This is because the margins of sets of moving tradesin are much more easily achieved, and you can actually get margin decreasesby doing this. Whereas at the moment, we are just constraining it.Therefore, doing scenarios, mimicking that,we see that the optimization resource, the capital resource,can increase by 10 to 15 percent. Looking at it in a different way,the people who participate in the POC that already have riskin the clearing house actually saw in the POC itselfabout a 60 percent reduction in capital, which shows it as well.The second thing where I think we can really see big jumps in resultsis symmetric data. The banks and participantsare generally still in a stage where the data is quite messy.We often see that they don't necessarily always even agree,whether they're long or short, whether they have positions.Doing scenarios or where we make it symmetricacross every bank, we actually see the results jumpby 20 percentage points. Those two things can really increase the resultsjust from improving everything and getting risk into the clearing house.The long-term journey is, of course, building this out,not just to the global 13 or 14 banks, but to the 20, 30,to 50 market participants that could benefit from this.On both sites and by site. -Exactly. I think there's huge potentialfor including cleaning by site as well, or regional banks,et cetera.. Maybe just to finish off, Quantile's aimwill always be to look at it holistically. That is part of what Smart Clearing is.It's putting the trades and the risk where they make most sense.Whether that is keeping them bilateral, moving them into the clearing house,moving them to a swap agent, moving them somewhere else,or putting risk overlays. There will always be Quantiles in orderto achieve the best possible results for the clients,for the participants in the network. -Well, I wish you luck in therebecause obviously the foreign exchange marketis dear to my heart, and I'd have been a little concernedabout this impact of regulation over the years as it's grown and grown.Would it dramatically impact liquidity in the market?I think what you're doing means it won't be, which is great.Esben, thanks very much for your time today.Thanks everyone for watching, and we'll be back again very soon.
