 as Jean-Louis Sherman, Secretary General of EMME, so the administrator of both Eonia and Uriber. Next to him, you already met Jaap Kees from ING. Then next to him, Rick Sanderlens from ISDA, Senior Counsel, and last but not least, Neil McLeod from Erstebank, Head of Group Treasury Market. So thanks very much for joining this panel. So we have now on this panel a focus on the fallbacks for the Uriber rate. So let me state unless if it hasn't been made very clear so far. So Uriber is not LIBOR. So the public authorities in charge of LIBOR have been very clear that LIBOR has an end date at the end of 2021. Nobody has said a similar thing about Uriber. We all want Uriber to continue to existing. So there are different futures that are envisaged. But nevertheless, we need fallbacks for Uriber that was first of all a recommendation that was given by the Financial Stability Board already back in 2014. And it has also been made EU law. So we need fallbacks for all important benchmarks, and in particular Uriber. And we see maybe now with the international developments how difficult it can be to establish and to implement really fallbacks into all existing contracts, not only new contracts, but also existing contracts. But let us start. I would give the floor to Jean-Louis to give us a bit information on the state of play of Uriber, which is being reformed currently. Thank you. Yeah. I will start by something that has been stated a couple of times already this morning, but I will repeat it. It's that Uriber has already been authorized on the BMR. I think that's the tip of the iceberg. It's the most visible part of it. That's the part that every user, as I believe, was waiting for since the 2nd of July that we have received his authorization that shows that the entire methodology that we have put in place for Uriber, as much as the governance framework, is compliant with BMR. And as a result, every user can continue to use it whether in new contracts or for legacy contracts even after the end of the transitional period for the benchmark regulation. Everybody also puts a lot of attention on it, another part on the visible face, I would say, is the methodology and the input data. And here we have the hybrid methodology. It's a kind of waterfall methodology that is anchored into transaction. We have three different levels, the first one being based on real transaction, a second one also based on transaction but with some calculation and a third one that then reflects the funding cost of the panel banks. But all these different levels, I want to put emphasis on it. They are all based on transaction. It's not like the code-based methodology with more or less an observation of what is in the market. So I want to give strongly that message that Uriber, the methodology, is based really in transaction, is anchored into real transaction. We have already started the phase in, so the transition to the new methodology. We have started that in during the second quarter of this year. We are gradually, one after the other, phasing in panel banks to the new methodology. And our intention is to finalize this phasing during the last quarter of this year. Of course, when all panel banks will be phased in, we will also communicate on that. So that's the last little piece of the new methodology that will have been finalized. We spoke a lot this morning, of course, about fallback rates. And we will still speak about it more about methodology, how to define these fallback rates. But maybe I want to insist more on one cornerstone aspect of a benchmark. It's not only the methodology or the authorization that everybody sees and looks at, but there is also a huge part on governance and control framework. Maybe just to recall, we are all in this benchmark reform there because there have been some, let's say, inappropriate things happening on the benchmark. And for that, we had the whole reform that started. And the reform is not only about the methodology where everybody focuses on, but it's also, and I would say, even much more about the whole governance framework so that administrators provide to the market at large, to the users, a benchmark that is solid, that is robust, that is reliable, and that is not subject to manipulation. And for that, we have established in the European Money Market Institute a very strong governance code of conduct for the administrator, which is dealing with conflict of interest, which is dealing with an entire reporting framework with transparency and so on. So that is maybe the part where one doesn't focus so much on it, but that is very important also in the whole benchmark reform there, not only the methodology. We have also code of obligations for panel banks. It's defined exactly what are the control frameworks that are reporting also on panel banks, how submissions are made, and so on. So it's very strong requirements that are put on panel banks here. I'll take again whenever I have the possibility to speak in public to also say that we are very grateful that we have 18 panel banks that continue to stay on the panel, that take in a certain way a burden for all the other ones. So if people can continue to use the benchmark here, it's not only because we have been successful in the reform, but because we have 18 panel banks that continue for the good of everybody to provide and to contribute there. And we have also now to continue on the governance and control framework because we have externalized the calculating to a calculation agent. We have also put very strong requirement on the calculating agent. So that's a bit where we stand in the URI board. So things are going on a very good way and there is continuity of URI board. It's not, I think Mr. Mayer also mentioned that in his introduction before, there's not a new and an old URI board, the underlying interest. URI board still measures exactly the same interest than before we have changed the methodology. So we still measure the same thing. We have simply changed the instrument to measure it. Maybe two words about the future, of course. It's not because we authorize now and because we have a good methodology that we will stop there. We will continue to make regular review, improve what can be improved, but we will also soon start to look at the longer term sustainability and maybe to see how this burden that some panel banks have could be shared in a bit of a more equitable way. Don't ask me what the solution is today, I don't know, but we are going to look at that very seriously. And then of course, the last piece to make URI board much more robust is to have the fallback rates and that is, I believe, what we are going to discuss now. Thanks, Jean-Louis, for the reassuring words. So let us then come to the work of the working group on risk rates. Maybe, Yap, you could start outlining what is the mandate of the group in this respect and tell us a little bit of the different alternatives that are there to have a fallback for a term rate that's based on an overnight rate. No. Yes, so the mandate of the working group is really following the FSB guidance to really look at alternative risk-free rates that could act as a fallback rate two-year-ribe. We have, of course, backward-looking methodologies where you only at the end of the fiction period know the final rate and we have forward-looking methodologies. And we look at both, because we also got feedback from the market on our consultation paper that there is a demand for both in the market. So first approach that we looked at was the forward-looking methodologies and we already issued a consultation paper on that and we also already came with a final recommendation on that. We looked at four methodologies, the future-based methodology, we looked at the OS-transaction-based methodology, the firm-quote-based methodology and a sort of composite methodology which is actually combining both the transaction-based and the quote-based methodology. Now, as we already discussed in the other panel, future markets in Europe are not that liquid. So on the basis of liquidity in current markets, that methodology was not chosen, but we can still see markets evolve. So our recommendation now does not mean that a potential administrator should stick always to that methodology. Like Jean-Louis just explained, benchmarks are here to be reviewed, to be updated, also methodologies can be adjusted. But for now, future-based methodology in the future market does not seem liquid enough. Then we look at the OS-transaction-based methodology. Of course, we prefer a fully transaction-based approach because that is probably most robust, but the current OS markets based on Ionia are not sufficiently liquid to do it on transactions only and that could still evolve. So once the benchmark takes off, the EOSTR markets could grow, could be more deep, which could mean that we could evolve to a full transaction-based methodology. For now, not the recommendation. The recommendation was actually to do it on a firm quote-based methodology, which means that it's not just expert opinion. These are really firm quotes offered by dealers, by market makers, on levels that they are willing to transact up to a certain amount. So these are real actionable quotes. That for now is the recommendation. We also looked at the composite methodology, which actually was combining the transaction-based and the quote-based approach, but that seemed to be operationally too complex. So for now, firm quote-based, but once market markets evolve, yeah, the benchmark and also the setup could change. Then maybe go to backward-looking approach. I think Neil, you could share some light on that. Thanks very much, yeah. Before I go into the details of the backward-looking methodologies, I think it's worth highlighting, again, it's been touched upon, why we need to look at backward-looking methodologies. I mean, number one, these would be available now. So when the overnight rate is published, you could start calculating this rate. Additionally, if you look at other currency areas, we do not yet have a forward-looking rate anywhere. So people need to be aware. This is still a question mark, I think as Holger answered in the previous panel discussion, it's still a question mark as to what will be there. Additionally, this very much avoids complexity in terms of the benchmark calculation because it's actual rates that you've observed. So a lot of the complications you have with other benchmarks you avoid. And the final thing I would say is that for some currencies, they've already stated that it's very, very unlikely that there will be a forward rate, in which case you have this problem anyway. Now, why do I start with that? Because what I'm about to go through is not there is no ideal solution in terms of the backward-looking methodologies. So you need to be aware of this at the start because this does have some fundamental changes that would need to be made within the way you operate. So the other thing I would do at the start is just give a quick marketing pitch to the FSB. The paper that Stephen mentioned earlier, this is based, the work that we've done has been based on this paper, which is very good if people have further questions, it's worth looking at that to start with. What I'm gonna run through is just certain methodologies that we, as a working group, saw as the preferred methodologies, what makes most logical sense. Now, it's useful just to look above and look at the screen so you can kind of understand the logic of each backward-looking methodology. You have an observation period and you have an interest period. In an ideal world, the observation period and the interest period would match and then you have a point when the payment is known and then the point when the payment is made. Again, in an ideal world, everything would match there. The reality is that even for a derivative market, it's not quite true that it all matches. So when we run through the different methodologies, it's worth being aware of that. How did we analyze the different methodologies there? We tried to look at all of the different aspects that we thought were most relevant. And remember, we tried to consider this across various different asset classes, not focusing on one specific asset class in itself. So the first one is obviously operational ease. The question is, can you handle it if you only know what the payment will be when the payment is due to be made at the extreme the same day? How difficult is it to calculate? Remember, if you start embedding these in, let's say, retail contracts, you need to look at how you would have to adapt your core banking system. And if you have similar experience to me, that is not an insignificant task. We obviously have to consider hedging ease, how it would then relate to the derivative market, but also then client acceptance. How much would the client understand it? How much would the mechanics really work for them? And then finally, and a point that does become quite a debatable one, is period congruency. So how much do you have a matching observation period and interest period? So the three main methodology we see as viable, and the first, this payment delay, which is basically representing how the OIS derivative market works, right? So we have a matching observation and interest period, and there is a small delay of one to five days between knowing what the interest payment is gonna be and the payment actually occurring. So very, very straightforward in terms of logic, the fundamental issue there becomes you only have a very, very limited period of time between knowing what the rate is and it being paid. Now, many, many sections of the market have issues with this. The extreme would be retail, but this is why this methodology alone, we don't believe would work for every asset class. So the second methodology is the look back approach, which has actually already been mentioned today in a previous panel, because this is the method that the bond market is looking at at the moment. And what happens here is basically the interest period and the observation period or the observation period is shifted backwards. Now, the standard at the moment is around five days. So you have some level, a mismatch in the fundamental economics, but it's relatively small. The big advantage here is then, okay, you have at least a slightly longer period of knowing what the payment is gonna be and then the payment's actually occurring. The last and I think most controversial, I'll talk about it a bit more in terms of the reasons to keep it on on the next slide. Here we look at something that is fundamentally backward looking. So you take the compounded rate every day, but for the previous period. So let's take a three month rate, you're looking backwards, you're saying, okay, what happened for the last three months? And we will use that to reference the next three months. Now, in terms of even within the working group, the response we got to this range from people understanding why we need it to the other extreme people absolutely hating it. So this one has the greatest debatable points, let's say. What's interesting in terms of each of these three? So the payment delay, obviously clearly, if you can handle the complexity and the fact that you only know the rate very soon before it actually occurring is very, very attractive, right? So that would work. And if we think about the sophisticated interbank derivative market, yes, okay, this works there. Now, even with the derivative market though, you could start debating whether it would work for the less sophisticated users. And so you do have to look at other options. The look back approach, as I say, has been successfully used in the derivative, sorry, in the bond market in other jurisdictions and hopefully also in the Euro area too. The actual methodology there, there's been some inconsistency in exactly the calculation in some issues, but I think that gets ironed out over the next three to six months and we have something that's consistent there. What's additionally interesting in terms of look back in other jurisdictions, we have already seen bilateral loans based on this methodology. And as of Monday, the LMA also published documentation referencing the look back approach. So it's certainly something that needs serious consideration there. I think the debatable question there is what sort of period is necessary? And then the final methodology that gains the most debate, we haven't touched too much today on retail. That's partly because Ionia is not so relevant for retail, but obviously you're eyeball is significant. Now, of these methodologies, either the payment delay or the look back approach, there is a big question mark as to whether they really would be applicable to retail or not and why. It's a question of how much notice a retail client needs ahead of the payment. And this is where the last reset at least comes in in some fashion because at the very least, you know what the payment is going to be. However, if we take, let's say the extreme of the Spanish retail market where we're referencing 12 months you're eyeball, whether that can really work in that example is a very, very big question mark. So the debate is still out there in terms of whether we need forward-looking rates as well as backward-looking, but these are the options that we most viably see. I think maybe the final point I would make, which is quite interesting looking at these graphs compared to the other options that are in the paper, you notice that all of the observation periods are exactly the same. Now, other options that were in there would either take a one-day rate for the whole period or at least for part of that period. Now there's been quite a nice example in the US market where that causes problems. If you have a specific spike in one day and then that's then used for a much longer period, it creates a much more volatile index that if you're just using it on a daily rate. So this is the reason why we came up with these three methodologies, but I think I'll now pass over to Rick to talk about the derivative market. Okay. So thank you. You may have picked up the question of fullbacks for all products. It's extremely complicated. There are a lot of moving parts. The Euro risk-free rate working group has made great strides in that direction and you may be wondering whether ISDA is just going rogue in running its own process in relation to fullbacks for your iBOR and other iBORs. I want to reassure you that's not the case. ISDA was approached by the Financial Stability Board back in 2016 in order to create robust fullbacks for the iBORs, the global iBORs. That's not to say that fullbacks don't exist today, there are fullbacks. The ISDA definitions provide a fullback to a dealer poll. So the calculation agent goes out to dealers and asks them what rate they think the iBOR would have been had it been published. But the Financial Stability Board rightly pointed out that in the current environment, dealers were unlikely to provide a quotation. If they did provide a quotation for one transaction, would they provide the same quotation for a different type of transaction or would different banks provide different quotations? And then if you have a 30-year swap and you have to do that process every reset date for the next 30 years, it doesn't seem a very sustainable approach. And you've got to look at this, I think, in the context of the degree of exposure that exists within the derivatives market. So our first keynote speaker, Benoit Couré, in 2018 estimated that there was over 100 trillion euros by a notional amount of derivatives which reference your iBOR. So if you have doubt around what your fullback provisions will end up as, that's a potential source of systemic market disruption. So is the set about creating these robust fullbacks? We didn't sort of differentiate necessarily between the different iBORs. So we cover US dollar iBOR, sterling iBOR, your iBOR. The fullbacks are going to be triggered by announcement or information showing that a permanent cessation has or will take place. And the fullback will be to the risk-free rates identified by the various risk-free rate working groups with respect to that iBOR. So in this case of your iBOR, it would be to the euro STR rate. The risk-free rates are, as you've heard, different in nature from the iBORs. The iBORs have a term structure. So you can have one week, three months, six months, your iBOR. Whereas the risk-free rates are overnight. So we need to make an adjustment for that. And that's the kind of methodologies which we were just discussing. But also your iBOR has inherent in it a credit spread and a liquidity spread, which is absent from the aptly named risk-free rates, which would also require some adjustment in order to mitigate against the effects of valuation transfer. So in terms of those two adjustments, ISDA has run a series of consultations you may well be aware of. The first one covered Sterling iBOR, Japanese Yen iBORs, Swiss Frank iBOR, Australian BBSW. Notably absent from that consultation were the two biggest iBORs in the world. So US dollar iBOR, estimated at nearly $200 trillion worth of exposure and your iBOR, which if you include that the entire market is somewhere close to the same amount. The good news was we had great feedback. We had very wide input from around the world, not just from the buy side, but from sell side community as well. The other good news was that they coalesced around the two adjustment methodologies that ISDA will proceed with now. So for the term adjustment, we're using compounding in arrears. For the spread element, there's a historic mean or median approach which involves comparing the spread between the tenor of the iBOR versus the risk-free rate for the same period and doing that over a certain look-back period. We didn't include US dollar iBOR or your iBOR at that stage because the risk-free rates for those jurisdictions either hadn't been printed yet, so EuroSTR, 2nd of October, SOFA had only just become published and so we wanted to allow a degree of liquidity to be established. But we followed up with a consultation on US dollar iBOR, iBOR, CEDOR, and again had very good feedback. Again, people coalesced around the same methodologies, both for substantive reasons, but also because they strongly wanted consistency across currencies and jurisdictions. We will run a similar supplemental consultation on your iBOR and Euro iBOR at some stage after EuroSTR has been published. That's the timing has yet to be locked down. But we're aware that your iBOR is a very big benchmark and therefore it's not a tick-box exercise. One point I'll mention just in passing, the methodologies did not include, the methodologies that ISDA consulted on did not include a forward-looking term rate. And the reason for that was that the OSSG, which is the official sector steering group that was set up by the Financial Stability Board, gave very strong guidance that they didn't believe the derivative market should fall back to a forward-looking term rate. That's partly because a forward-looking term rate would be built on the basis of a liquid overnight derivatives market. And so there was kind of structural anomaly there. But also because the risk-free rates are so much more robust than a forward-looking term rate could ever be. So the consultations seem to be going in a good place. What we haven't done is lock down the final parameters for the historic mean and median approach. And so we recently launched a new consultation, which is due to close on the 23rd of October. I'd encourage you all to have a look at that, engage with it. The two methodologies which we've suggested are the five-year, sorry, five-year median or the 10-year trimmed mean. And there are pros and cons to each approach. I would urge you to go to the ISDA website. There's a huge amount of information there. There will be webinars made available for you to look at as well. And we're also consulting on this question of the lag or the lockdown period that should apply because, again, falling back to a risk-free rate that you only know on the date of payment I think will cause problems even in the derivatives world. Finally, we also published Bloomberg Index Services to, sorry, we also appointed Bloomberg Index Services to publish the risk-free rate plus the spread that ISDA is going to, has been consulting on. And you should expect to hear some more information on that in the near future. And finally, we also recently consulted on other types of event that might cause a fallback. So what we call precessation events. If the FCA for LIBOR, for example, were to announce that LIBOR was no longer considered to be representative of the economic reality that it's designed to measure, what should the consequences be? And we expect to publish the anonymized and consolidated feedback from that in the near future. It's fair to say that we got feedback falling into three buckets, people who wanted that trigger to be sort of hardwired into ISDA definitions, those who wanted it available but with some optionality and those who were opposed to it. So we intend to work with the OSSG on resolving that dilemma and again, keep your ear to the ground with respect to that. Thank you, Rick. So I think really the big question that comes is, what about consistency? So you have outlined that ISDA is looking at backward, looking methodologies for the derivative space, but we have heard from the previous speakers that the working group is looking both at forward and backward-looking methodologies. Does it make sense to have different ones for different product types? So a forward-looking for the cash products and the backward-looking for the derivatives or different types of these backward-looking rates that Neil has outlined, so across asset classes or is it desirable to go all in one direction, even though maybe then the chosen fallback doesn't fit the asset class so well? And then secondly, how about international consistency? How important is it to go along the same avenue that maybe other jurisdictions have chosen or is it better to have a tailor-made solution for the euro market? Anyone? Several questions again. Yeah, I think it's a very fair question. It's also a nice bridge to our next steps because this is exactly what we now need to analyze. So for some products as a fallback, what we got back from our consultation is that there is really a need for forward-looking benchmarks. But yeah, then we have derivatives potentially falling back to a backward-looking benchmark. And how can these coexist? Can we still, as bank, offer these products and then hedge the same products with a different convention? So offer forward-looking and then hedge on backward-looking. Most of it is possible and especially for large banks, but how about the markets? Can they also work with that? And or will there be certain fixing risk or basis risks? So this is all the work we now need to do. So we will look at the various cash products, we look at the very derivatives, and then we will look how can the rates, the backward-looking and forward-looking, what would be the best suited rates for these products? And how can they coexist without getting too much basis risk and fixing risk in the market? So that is a big challenge, but we have, as you've seen, a very good group on that. So there's a group five actually working on that with Christian and Anna. But yeah, big dilemma and a big challenge, I would say. Next to that, we are looking, of course, at the recommendation on the SPAT methodology. So that is also a next step of the working group. Now we have, of course, also we will closely look at what ISDA Group is doing and also try to ensure that there is some consistency and also consistency across working groups, across the globe, but we will also do our own analysis, because yeah, that is the mandate we got as a working group. And of course, we will look at the impact of the inclusion of these URIBO fallbacks, both on systems and infrastructure, but also on risk management and financial accounting. So systems and infrastructure, again, so group five, risk management and financial accounting, that's markers with group six. And finally, the legal group will still come with recommendations on the URIBO, both for new and legacy context. So we have a sort of legal URIBO action plan. So these are, I think, also a nice, so your question is actually a nice bridge to the next steps. I can't give a final answer to the question, but we will surely look at it and we will try to solve this and give more clarity to the markers. Okay, thank you very much, Jaap. Can I just add something on it? So the degree of everything that Jaap said, I mean, the basic building blocks of both the forward and backward looking are the same. One of the reasons why we chose the RS quote-based methodology is because you could have a clearer linkage between forward and backward looking methodology. So as Jaap was saying, it at least minimizes the hedging costs or the mismatches. The big question I think in the market there is, okay, who manages those mismatches that are there? Are they then embedded within banks to manage or are they transferred to the individual client? And that I think is a question mark at the moment. And I think from Mr's point of view, I mean, it's a great question and it's the question we'll be asking in our supplementary consultation. The consultations so far, as I say, have given us a strong indication that people do value consistency across the jurisdictions, but you've also got to be aware that there are two different types of user in a way in relation to each eyeball. There are global users who definitely, I would suggest, value consistency, but there are also people who only use one eyeball. They only use your eyeball or they only use US dollar liable and they may have views which diverge accordingly. One other issue that we will need to think about and again, we'll consult on is what data we can use. So if we adopt a backward, a historic mean or median approach, that requires us to have historic levels for EuroSTR. And that was the same issue that we faced in relation to SOFA. We don't have five or 10 years worth of SOFA and so the question on the US dollar liable or consultation was would market participants be comfortable with this calculation being made using proxy data or in the case of EuroSTR, could we perhaps use the Onia as an alternative source of that spread? And these are questions we don't know the answer to, but they're interesting and we'll factor into the consistency point. Thanks a lot. So a lot of open questions, but you're working hard on it and I'm sure you'll find an acceptable outcome for everyone.