 Felly, roeddwn i'n ddweud y FFT yma yw'r wychau, ychydig wedi'i gweld y cwrdd yn ysgrifennu Prof. Ricardo Rebonato o'r edrych ar y Cymru, yw'r cyfnodau clymysgau, yw'r cyfnodau clymysgau sy'n ddweud yw'r cyfnodau'n mynd i'n dweud yw'r cyfnodau cyfnodau. I Speaking of which, that has been one of the many topics that we are going to be delving into over the next hour in this panel. So the panel is essentially on the climate biennial exploratory scenario learning exercise, commonly known as SEABES. That exercise was designed to stress test the UK's largest banks and insure as portfolios against climate risks. Dw i'n gwybod yw'n gweithio am y cyfrifio i ymddangos ac yn ddiddordebeth ac yw'n gweithio'r lles. Rydyn ni'n amser i'r panel sy'n rhoi am brosesau eich hwn, mewn gweithio'r hyn, a'r ddweud i'r ymddangos a'r dyfodol ymddangos ymddangos. Mae'r ymddangos, rydych chi'n gweithio a'r anod o'r ymddangos, yn Gweithfyrdd Mike Wilkins. Ryd yn gyfnodol ddechrau sy'n cyfnod gyda'r Llywodraeth Cymru, ond yw'r coleid yw'r CGFI, a'r cyfnodd yn y Llywodraeth Llywodraeth Llywodraeth i'w ddweud i'w Nicola Rangyr a'r Profesor Ian Clacker. Nicola Rangyr wedi cael ei wneud hynny, ond i'n ddweud i'w ddweud i'w ddweud i'w ddweud, wedi ei wneud i wych i'r newydd a yn ystod i ddweud i'w ddweud i'w ddweud i'w ddweud. Fighting the sense, dywed ybau yn yma. Mae gen i ddweud i'n gweithio yn arweithio, seef nhw wedi ddweud i'r ddweud i'r ddweud. So, ddim ddweud i ddweud i'r ddweud i fynd, ond wedi gynchno'r ddweud, a wedi ddweud i'r ddweud i'r ddweud, a nid i'n ddweud i fynd i'w ddweud i'u ddweud. mor hefyd o'r cyfle brydaeth ar y bryd. Gweithiwch ar gyflwyno ac mae oedd doctors ar y bethau cyfrnach ac mae'r gwrando gyda'r cyhoeddau. Mae'n ddiwedd ymlaen i'r panel. Rydyn ni'n chyfwng i ddigon doedd, ac mae'n ddigon iddyn ni'n ddigon. Calon i'n ddiddorol i'r panel. Mae'n ddweud a gwysig wrth hyn ymlaewch, oeddwn i'w Nicola ac mae hyn yn maes o gyflwyno gafiaethau I found out I was doing this at midnight last night or this morning, depending on how you do time. I'm actually not just protein international, I'm also a professor of pensions and finance, so I'm not some random bit of the academic machinery coming to do this, and I was also involved in the CBes exercise and doing the research. I think what we want to get to is a reasonably quick overview of what we did because I think the more important bits, the more interesting bits are the what next that come from this. So when I got the slides it was also quite a surprise to see myself. That is myself, Nicola, Mike and people at Imperial doing a round table with some of the participants through the Climate Financial Risk Forum. One of the things that's important about that is, and I was just discussing this with Chris, what I found very interesting in this piece of research is working with the Bank of England because the Bank of England is, as a central bank, very cautious, and that because it has such a fundamental role. So the CBes exercise and being able to get this level of access and this freedom to actually investigate what worked, what didn't is actually, I don't think, very common in working with central banks globally. So I think it speaks to something about the approach taken by the Bank and more broadly in the UK because we see other examples of this within the FCA and so on. So what we're looking at is this is like a proper partnership and this is, to some extent, an unfiltered exercise. So we sense check things with people in the Climate Financial Risk Forum. We sense check things with the Bank but actually this is a wholly independent piece of work and so the views expressed here are our own. That's further caveated with the fact that anything I say between now and the end of this session are just my views, they're not Nicholas, they're not CGFIs just in case I get it wrong. Right, so very quickly the objectives of the CBes exercise help strengthen knowledge and capability across the UK and globally because actually the CBes exercise was a paradigm shift in terms of how central banks do things. So any learning that comes from the CBes exercise can be shared globally. We wanted to understand what the impact CBes had on the capability of those that participated and we also wanted to come up with some recommendations which I think will be the general subject for the next panel discussion. So very quickly, started in March, April 2022, made some informal interviews. We then built and developed a survey and that was where it's interesting from a research sense because you're bringing together climate scientists, you're bringing together social scientists and various other groups to try and make this work because you're dealing with something which cannot be analysed through one lens or by one discipline. We then ran the survey and actually despite those small numbers that's actually a really high number given how few institutions participated. So actually this is really quite strong evidence and also it's what we call in social sciences, hard to reach populations, the people that we got access to are the people who are doing this. That's very, very rare to be able to do something like that. We then also had some formal interviews with banks, general insurers and life insurers. There was a two workshops in January and the report was then released later, earlier this year. So in terms of highlights, I think the strong evidence for capacity building was there and so we feel that the CBes exercise was hugely positive in this way and particularly because what happened was a much greater awareness at bold and sea level. So the top end of these organisations started to become much more cognisant and aware of these types of risks and that's the only place you're going to get traction. If we want to see systemic change in any of these institutions it has to get that high and the CBes exercise for me delivered it. Now one of the things we also found was the lower impacts at the client facing level and that's really about the fact that the process was fairly short and a lot of those direct engagement with clients didn't take place. That wasn't something which was an objective of CBes but it was one of the findings that we had. Broadly speaking, the design we've seen is adequate for the purpose of it. Now there were challenges and we're going back on to the static balance sheet because it's in the recommendations. But what we're looking at is, as an exercise it worked but things like the static balance sheet, which are controversial and I know Mike has discussed this in many forums as well it had a use because we could do something with it. It was not perfect and I think the bank acknowledges that. What the CBes exercise also did and it was one of its objectives is it allowed us to learn lessons for future work. So actually what worked in this exercise and what was good and then what needs to change going forward. One of the things that again comes from the exercise is there was this capacity building within the sector more generally because nobody had ever done anything like this before. It was done in a fairly constrained period when other things were going on but what we did see is there was this upswell in organisational capacity and personally I think we have to maintain that in some way. How we do that I think is an open question but it would be a tragedy to see the capacity that has been built up lost. And as I said the findings can be shared so the lessons from the CBes exercise do not just apply to the Bank of England in the UK they can be shared globally and help other central banks and supervisors around the world to gain their exercises more carefully to build off of this. So for our recommendations I think the one that I'd like to focus on because it's the one that will always come up is addressing data gaps. There are always going to be data gaps and one of the big challenges we have is with data gaps what do we do? Do we do nothing just now and hang about and wait until we fill those gaps or do we proceed with an imperfect analysis to make progress? Many a think the approach of an imperfect analysis to make progress which is where we arrived at allows us to identify those data gaps and I think some of those gaps speaks to the previous panel about transition plans and so actually some of those data gaps are already getting filled and so as we bring this together what we're going to be able to really do is accelerate these types of exercises going forward. The other thing that is going to be really important is the collaboration in terms of methodologies, guidance, knowledge models, scenarios, the whole thing because what is a real risk and we've seen it in other places around other things is that everybody starts to veer off in different directions and well-intentioned as they are that non-standardisation, that a lack of commonality means that we're going to potentially miss things and so I think very, very importantly if we look at things like dynamic balance sheets that's an extremely hard thing to do it's something that emerges as something that has to be done in response to CBes but it's extremely challenging and so I think as we forward look that we have to be honest about what the direction of travel is and how we get there. We also can look to things like the Climate Financial Risk Forum as a way of convening relevant actors in this space to try and get those standards, that commonality and learn from each other because it's the way in which we'll accelerate forward. Last thing sorry, I just looked down and it said 33 of 76 which panicked me, I've only got seven slides. Very, very quickly just to finish before we get into the discussion I've got to say thank yous, okay? So I've got to thank the whole team of researchers because while myself, Nicola and Hannah who's loitering somewhere in the audience led on a lot of this that just gives you an idea of the amount of people that are involved from the CGFI side we've also got to thank the Climate Financial Risk Forum for all the support and access they gave us and also GALP and also to Chris at the bank because as I said right at the start I am not aware of the bank previously being so open in the way in which it engaged with a group of academics because as you can see we are quite random. Now as I said this is my views this is not Nicola's or CGFI's. Anyway if you want more information there's Nicola's details. Thanks Ian. So we'll get into our discussion now and to avoid the death by panable syndrome I'm going to try and keep this fairly light and breezy in terms of a conversational style but quickly to introduce the panel in front of you from your left to right we have Professor Ramakand from the Mathematical Institute of Finance at University of Oxford Joe Paisley who is President of the Global Risk Institute at GALP Jacob Tomai who is from Two Degrees Investing Institute and we have Chris Faint who is a leader of the Climate Hub at the Bank of England and Ian you already know. So that's the panel we have this afternoon. So let me go straight for the jugular so to speak and I'll address this to Joe to get things going. Was or is Sea Bears fit for purpose? Good question and I would say yes it was fit for purpose because the purpose was really about building capability and let's not forget it's kind of world leading so there wasn't a blueprint out there to just follow. Now one thing I've noticed already is people were slipping into talking about it being a stress test it wasn't a stress test it wasn't designed to be a stress test so when we're talking about scenario analysis and stress testing we need to think about what is the purpose that we're doing this exercise for and I see the Sea Bears very much as a kind of world first building capability in the firms which wasn't there. So for those of you who don't know GALP with the Global Association of Risk Professionals and we are passionate about risk education I run the thought leadership arm which was set up five years ago and was given a kind of blank canvas look at any areas of risk that you don't feel that risk management is paying enough attention to and practically the whole five years I've been focused on climate risk, sustainability, nature increasingly nature which gives you an indication of how important it is how systemic it is but also the gaps we see in terms of knowledge, capability, understanding which need to be urgently addressed and I think that's what the Bank of England was doing but I can't really put words into Chris's mouth but in terms of being fit for that purpose absolutely and we can argue about other purposes like was it really stressful was the design of a scenario appropriate how do we move on and I'm sure we'll come to that. Jacob what's your view was it for purpose? So I have to agree on the first bit but then the question is is that an appropriate enough purpose so it is worth interrogating whether simply saying capacity building and education and the comments that Ian made as well is that the right level of ambition for an exercise like that and I think we've talked about the conservative nature of central banks already and some of the constraints or handcuffs that are perceived or otherwise are put on these exercises as a result of that but there is just I think a lack of courage in actually moving away in making sure that the kind of economic risks are properly translated into financial risks IPCC is 3,000 pages long plus so nobody reads this I guess but the chapter on the economic risks in the IPCC are about 6 to 10 pages out of 3,000 pages and they tell you that quite a large number of scenarios suggest the UK will economically benefit from climate change under most temperature outcomes this is the IPCC so if the central bank wants to exercise wants to implement a risk exercise whether you want to call it stress test or not we can debate the semantics they have to have the courage to challenge the true risk that may materialize and that requires looking at the mortality effect of climate change the potential conflict and civil and international conflict effects of climate change that requires looking at massive supply chain disruption from climate change and if you are unwilling to do that then at least on the physical risk side of the exercise you may be wasting your time Chris I don't expect you to be able to say anything in an unbiased way but if you can give us your view on whether you felt CBEZ fit for purpose bearing in mind its limitations I'd be happy to I'd probably am a bit biased since I designed it and implemented it so I want to pick up this point on ambition because I think with hindsight it might be easier to frame ambition compared to where we were so in 2019 when we announced that we were going to do the CBEZ no framework existed in the world for doing this so we started with a blank piece of paper we locked ourselves in a room and over a couple of months we actually looked at how to do these exercises and that was a huge exercise in itself and we didn't only think about how we can stress test the banks which is what we've done in the past but for the first time ever we looked at doing an exercise of a scenario exercise not a stress test that included both banks and insurers so that's huge, that is the first time all of those institutions have been put in an exercise together because we didn't want to be unambitious if you just looked at banks you would only get part of the picture we also looked at further ambition we could have looked at including asset managers pension funds but there's capacity in what you can do and this is the story of the CBEZ in every aspect of the CBEZ design we could have been more ambitious we could have included the trading book we could have looked at more granular data from institutions but if you go too ambitious then the issue is that I think the quality of what you do get goes down there's only so much capacity within institutions to do new stuff so what we sought to do was balance ambition for something that's never been done before with reality of what we could actually get back and I think what we sought to do was get a sweet spot now in some ways I think we got that right I think it was absolutely right to include banks and insurers but maybe not include asset managers and pension funds at that time so that's been too big similarly I think it was right to not include the trading book I think that would be a totally different framework for doing an exercise we can get on to frameworks if that's helpful so I think that was also the right decision now could we have gone more stressful potentially could we have included some different kind of narratives around the CBEZ absolutely which goes to the value of what you've found from your exercise so I would argue we were ambitious I would definitely state that we would do the exercise differently if we were to do it again today but that was the point of doing the exercise in the first place doing a world first exercise so we could understand what works and what doesn't and then just be open about the fact that if we do it again we're going to do it differently sure and that's one of the learnings from the exercises what could be done differently maybe if I can come to you Rama and you know I know you've looked at this from a mathematical modelling perspective what do you think could be done differently and what could be improved on for future scenario exercises like CBEZ thank you so first let me start by commanding Chris and his team at the Bank of India for actually doing this because I've been involved in regulatory exercises elsewhere and actually carrying out such an ambitious project is very difficult so that's already commendable that they have done something now from the point of view of the methodology there are many things that can be improved so one of the things that Joe said initially is that I don't think this CBEZ should be seen as a stress test even though some people may have intended it as such because the way well if you look at how stress testing is done in the banking sector for example there are actually two different way stress testing is done one is supervisory stress testing which is probably the one that Bank of England built upon to design this exercise there the regulator picks one or two scenarios an adverse scenario and a baseline scenario and then they compare the banks or the insurance companies losses in these scenarios and that's the here the choice of the scenario is crucial because it should somehow embody the risks that the institutions face in an extreme but plausible situation so this choice of the scenario is crucial and in a banking or in a bank stress test setting often there is a dominant market factor or the index or interest rates so the scenarios correspond to large adverse moves of this dominant risk factor so in but then there is the internal risk management of the banks and financial institutions where stress testing and risk management is done in a much more complex way institutions typically look at a range of scenarios and the reason is that given the complexity of the portfolios it is not obvious at all at the initial where the risk lies so you look at actually a very large range of scenarios tens of thousands and you try to assess the exposure of the financial institution to these scenarios and then identify as an input the scenarios to which the financial institution has the largest exposures and then you may want to hedge that risk or take that risk that's another decision so we have a range of scenarios very large number versus the choice of a single scenario in the supervisory stress test obviously banks do not rely on supervisory stress test for their risk management policies that would be highly risky to base your risk management on exposure to a single scenario no nobody does that so this is a reporting thing really and financial institutions really have an internal risk management based on a very wide range of scenarios now coming to this exercise we're looking at climate change very very complex phenomena in which a multitude of factors intervene so having a single scenario means you're picking one single pathway for all the complexity of factors which enter this problem including demographic, physical and also financial economic that's certainly if you know in advance where your risk lies perfect but that's highly unlikely it's even much more complex than a purely financial portfolio so here I think it's even more necessary to look at a range of scenarios probably wider than what we do when we look say at the fixed income portfolio so and now if we go to the details the CVS scenarios I believe if I'm not mistaken we're based on one one of the so-called shared socioeconomic pathway scenarios that were an output of the climate science and IPCC work this is the one so they produce five scenarios shared the so-called shared socioeconomic pathways one to five depending on whether you think there will be active engagement in favour of transition or not and this my understanding is that CVS was based on SSP2 the second scenario which is a middle of the road scenario there is some engagement for transition but not full so it's a if you want it's a middle of the road scenario as they call it it's certainly not an extreme scenario so even in the framework of the five scenarios the SSP scenarios we're not looking at to stress this in the sense that we're excluding the extreme scenarios and even the SSP scenarios whether the extreme ones or one in five or the middle ones you should think of them if you're a financial institution you should really think of them as a way to centre the scenario universe you're thinking of it defines the expected evolution of a certain number of macro variables and then around those variables you'll have volatility somebody pointed out this in the previous panel and the volatility would build around these scenarios to generate actual market scenarios which would be then an input to your stress test and I think that's for example how if we look at the bank stress test scenarios that's really how it's done also you have a macro scenario for defining terms of inflation and GDP and so on and then the financial institutions base themselves around that to generate scenarios for market variables so there's volatility incorporated at a second stage and here I think we should really think of the CBIS scenarios as a baseline around which there will be other things coming on the specific factors related to market volatility so this is about the scenarios my second remark is something that came up in the GARP survey it's about what's called static balance sheets versus dynamic balance sheets so these scenarios in CBIS are long term scenarios I'm not talking about three months but later decades so over decades it is unrealistic of course to assume that any financial institutions balance sheet will be static there will be rebalancing at least at the annual level and the balance sheet 10 years from now may look completely different so the input to these exercises unlike bank stress test which are short term there you can the inputs are typically the static balance sheets as observed today as a certain date observation date here it cannot be balance sheets because balance sheets change over time and 30 years even 10 years is a very long time scale for balance sheets so as was indicated in the previous panel the input should be strategies describe your business plan your strategy to face your transition plan your strategy for climate change and I will then tell you whether your transition plan will survive such and such scenario so a transition plan is dynamic is a commitment to do certain actions which will then affect the portfolio in a certain way if I can then implement that in the in the scenario analysis that's going to give me a more reliable answer than how does your current static balance sheet grow through the scenario because that may not correspond to any actual profit and loss of the financial institution unless they're sleeping at the wheel hopefully not so those are my two remarks I think the static balance sheet is not a technical issue it's somewhere in the documents but it's really we need to think about strategies scenario analysis of strategies just as financial institutions today do backtesting of their investment strategies they don't backtest static portfolios they say ok this is my investment strategy I'm going to do a bad backtest of this or through counterfactual scenarios they know how to do that but you have to you have to incorporate the strategy as deployed over time especially if we're looking at such a long horizon maybe just to pick up on the usefulness of this from a risk management perspective I mean despite the limitations that have been highlighted in this exercise there are still certain aspects of it which can be useful but need to be adapted going forward maybe Joe could you just give us your view on that what do you think going forward could change to make seabares that much more useful for risk management professionals seabares is kind of gone unless you're going to do another seabares but the question is how can banks and other financial institutions model the risks that they are exposed to understand them mitigate them and help the transition so one thing that might be quite interesting is that before the seabares actually started we set up a little working group with many of the banks that were involved and it's interesting that we're critiquing the seabares now because I know that that group before they even knew what the templates were going to be on the instructions or anything literally did not have a clue what they were going to do so it's a kind of mark of success isn't it that we're here critiquing the fact that we shouldn't be using necessarily SSP2 as underpinning you know to come on enormous enormous way but one of the exercises we did with the banks was just a very simple exercise to benchmark their probabilities of default because one of the problems with the climate modelling is you haven't necessarily got the historical back data to kind of look well you know with a macroeconomic scenario you look at what happened in the financial crash that happened in the 2007 or whatever with climate it's not the same you don't necessarily have that back data so what do you do when you model something that's quite new you present it to your senior management and you go ok do you think this is reasonable I have no idea if this is reasonable literally no idea so one way you can overcome the lack of time series is you can kind of go for a cross section so what we did we did some benchmarking we got the firms to submit to us how their PD's evolved in the disorderly scenario that the Bank of England published and my goodness that was illuminating because there was such a range now what we did we split it down by credit ratings for different sectors I mean it's a pretty broad brush thing it's not an exact science but the learnings they got it was just to see sort of you know were they a million miles from the centre of the distribution if they were why was it because they were just ridiculously prudent or the others were not prudent enough and so that kind of learning exercise was huge now what comes after see bears well to be honest it's kind of more exercises like that but critically more short term scenarios as well one of the things that the banks really struggle with is kind of trying to think 30, 50 years ahead you know if you want them to take action now risk management action then it sort of has to be within the next five years really and that is the huge challenge isn't it because we don't expect many of the physical risk to become kind of extreme until beyond that which of course depends on the action they take today so and it's also contingent on policy so I think we're going to have to have more short term scenarios we're probably going to have to have more regulatory scenarios on a range of different SSPs but we're absolutely going to have to see not just better scenarios but also better modelling within the banks as well better more expertise, more benchmarks that they can use because this is not just about the biggest and most sophisticated banks and insurance companies it's also about all of the other all of the other firms as well that are smaller and arguably probably more concentrated more exposed because the bigger firms can probably be more geographically diversified so yeah short term scenarios more integration within their risk management better modelling better exercises like the benchmarking one I was talking about so they can learn from one another as well and no doubt CPEZ 2.0 which will be even better and incorporate some of the risks that Jacob is talking about that's quite a shopping list of things that need to be done and it's urgent absolutely and Jacob maybe I could bring you in here because do you think these challenges that currently face financial institutions, financial sector in general can actually be overcome with the provision of new tools new data, new approaches what do you think could actually lead to these fundamental changes required so I mean the one thing is obvious to make sure that any of the evolutions is centered on the actionability right and that's where you have, it's great if you have a number but if you can't action the number and Ian you highlighted data so one of the things that we're doing with the Bundesbank for example is developing an SME data set for the German SME sector over 350,000 SMEs in the data set now so that's an exercise where then if you run an exercise we can debate PDs and model scenarios but you know okay here's SMEs both business fine but you can think about how to develop your strategy around SMEs not just based well I'm a local bank in Germany, I'm not going to stop blending to my SMEs locally that's not a viable, actionable strategy whatever a stress assess for me right especially in a more diversified banking sector I think that's the one thing, I think the other thing is what was already said about connecting this to the multiple scenarios right I mean it does seem that we need to think about under how many futures am I surviving rather am I fine under this future and I think that in that point to Chris point about the level of ambition Monday morning quarter backing which is obviously my favorite thing to do is to sit this American phrase sit there and go yeah I would have done all these things differently probably would have done it more or less exactly like you probably not as good right so easy for me to say that now but I think you know I used to work in France saying there you used to say oh this is a lovely prosecco but a horrible champagne right this is one way to think about this right this is a great prosecco it's a poor champagne and if you want to get to this overcoming this challenges you need quality champagne so you need SME data bulk I think the German Bundesbank is really pioneering this field you need an integrated nature social climate risk which is really a stress test and to call it that and have the courage to call it that and if you're thinking about five year scenarios you're not going to be in the spirit of the op ed from financial times you quoted you're going to be in a situation where you say you have an El Nino then coming up now obviously so the next couple years going to be pretty hot so what does happen if we're looking at maybe uninhabitability in parts of the trophic trophics affecting maybe upwards of 50 to 100 million people like that's what we should be asking ourselves if we want to run stress test because that is what the frontier of climate science tells us is the frontier of our climate risk and that feels uncomfortable because first of all the IPCC feels very uncomfortable talking about this they don't do it properly you're basically asking people to believe what most people think is science fiction and so this is not a comfortable place for the Bank of England to find themselves in but if you are a bank internally and you are serious about thinking about these risks thinking about integrated nature and social risk I mean the iris troubles here in the UK they're a 10% shock of Northern Ireland GDP the iris troubles have a bigger GDP impact the historical empirical iris troubles impact from a range of studies then what most of the IPCC studies say about climate change so why are we not looking at these kind of impacts as part of the climate change not instead of but as part of the climate change change discourse and I think that's when we get to the point where we can talk about I mean basically start drinking champagne right which is why we're here I understand Yeah I'm sure we are and we'd all love to be drinking champagne right now unfortunately we're talking about stress testing the scenarios so I'm just wondering I mean we can slip into quite a technical discussion here about stress testing, modelling, probabilities you know whatever from a layman's perspective from a person on the street trying to understand all this and realise are they at risk are their savings at risk in the future I mean Chris looking at it from that perspective do you think there's more that can be done in the future moving on from where we are now to try and get these messages across in terms of how exposed we are to risk from a financial perspective so this goes to the multiple roles of the Bank of England so we've got many different remits but to summarise them we've got the micro-represential remit looking at the health of the banks and insurers which speaks to the security of savings and banks we've got a financial stability remit which talks about the financial disability of the system and we've got a monetary policy remit which people will be familiar with I think the point that Jane made is a really really good one that we talk about scenario analysis as if it's one size fits all one exercise that can speak to each of those different remits and a number of different objectives but unfortunately that's just not the case so the exercise that we did that we did as part of the seabares it was a 30 year exercise to look at the vulnerabilities over time so we kind of spoke to the question of how secure your savings would be but actually if you wanted to do a micro-represential test focusing on that specific question for banks and insurers what you might want to do is a shorter term much more stressed exercise that's more akin to the solvency exercises that we do if you want to look at financial stability as the UK travel to note zero and how that might evolve then maybe you'd want to do another 30 year exercise so what this points to is the need for a number of different scenario exercises in an ideal world you'd have a number of different scenarios of different levels of stress that would allow that to be undertaken and a number of different frameworks the reality is that we cannot do all of those exercises there's not enough bandwidth to do it so what we need to do is we need to prioritize and understand where do we want to put our focus what is the most important thing for us to be looking at at the moment what is the proximity of the risks to where we are today to your point on where are our savings or where are the financial stability risks most acute and that is a discussion that we are having within the Bank of England it's a discussion that international central banks are having as we think about how do we build up the toolkits how do we prioritize the work that we're doing but at the same time it's not all about the scenario exercises that we're doing of course it's not it's also about the work that we're asking firms to do so we're asking banks and insurers to understand internally the risks that they're held to and I think a really important point to understand if you look at banks is that on their own they can't understand how their risks are going to arise over the next 5, 10, 15, 20 years they need to speak to their counterparties usually in the real economy to understand how they think their risks are going to arise and this comes to your point Robert actually a bank cannot, it can form a judgement over how one firm might seek to transition over the next 10, 20 years but they need to speak to that counterparty they need to understand how they're thinking about it and actually that was to my mind one of the greatest successes of the SEBAs we asked institutions to speak to that 100 largest counterparties 100 counterparties in the vast majority of those cases have never been had before and the counterparty were like we don't know, we've never thought about this so it triggered those discussions it triggered a process where they would have to think about those risks and I think that has to be magnified a number of different times over the next years for us to truly have an understanding of how risks not only happen in the real economy how they get magnified within banks and insurers and what that means for us so it's very complex I've been up to questions in a minute did you want to come in with a point Jake? I just wanted to add because I was struck by the use of the word savings here I think it's quite an interesting discussion I always think about the risk to savings but not the risk to levels of savings aka wealth and it is just a missing piece of the climate puzzle there are no models that I'm aware of and I've talked to dozens of modelers on this topic that measure the impact of climate change on wealth and we're not just talking capital stockings and put into the production process we're talking a country's wealth Germany recovered its GDP level in the late 1950s after World War II still hasn't recovered its wealth levels from World War II 70 years later and that's I think a really interesting part of thinking about savings and centering that as part of the different use cases you described Chris I'd just like to make a comment about what you said so I think it's very important to note that this exercise was conducted by the Bank of England because it's the only institution in the UK which could have done this because unlike stress testing done by financial institutions for their short term exposures to market risk and credit risk any given financial institution does not have the visibility on the elements that are needed the inputs that are needed to conduct such an exercise it's only you need a broad cross section of view on many things that are not available to the financial institution itself and I would argue that actually for a realistic assessment of one of the points that was brought up the short term impact of a long term risk of climate risk you also need to broaden beyond the banking sector because most of the assets in the economy are held by asset managers that are not banks insurance companies pension funds, other asset managers and we speak about transition risk as some kind of abstract mechanism whereby climate risk will affect financial markets but in fact it's very concrete it's the rebalancing policy of these large asset managers which will determine capital flows from one sector to another divestment or investment in new sectors and so on this will determine how markets will react in the short term or the medium term to the long term climate risk so I think including the behaviour of these large asset managers in the model in a model which can only be a macro model run by the bank that can give a concrete way to model what transition risk is concretely and how it will impact financial institutions it was two small things one of the findings we had in the report which was really interesting and it wasn't actually one of those objectives of the exercise where people were saying they didn't have enough time to talk to their counterparties so actually it was a really positive thing because going talk to your biggest 100 we've never done that before and it was that we now actually would have wished they would have had more time so we could have engaged in that more fully so I think that was really important in terms of something that came out of it I think the other thing was something Joe said is just large banks and insurers these are the banks that have the most resources and the most capacity to do this and they're constrained so once we go outside of that pool and we then start to go into smaller banks or smaller insurers that constraint is going to become much much more binding and so a lot of the heavy lifting being done by large banks is going to hopefully allow us to develop capacity further down the market because the other the resource constraint is human capital and there's not enough of it for the global banks, large global insurers it's definitely not going to be there further down the chain so that gap has to be filled and it has to be filled quickly Chris, I'm going to ask questions after Chris's comment here I'll just make one quick comment and I'm paid to be a bit cynical and focus on risk but when I hear that banks and insurers said they didn't have enough time to speak to their counterparties I'm dubious about that they had two years, potentially two and a half years from the point they were told they'd have to do it it's the point they had to submit now I don't know how long it takes to pick up a phone to a counterparty and have a meaningful discussion with them but I suspect it's somewhat less than that so I think what this tells us is that in a lot of cases it's really important that there is a catalyst for that discussion to have to happen so you could take this and that discussion might not even have happened now had we not run the exercise so it would be interesting to get underneath that one Question over here, thanks Can you say who you are as well please? Scott Agay, I guess it is an economist I can say I don't think William Nordhouse won a Nobel Prize for climate stress testing for sure, yet there's a methodology that's been carried throughout from the top down my AM side and I want to get into those details just a simple question for the whole panel if the current scenarios don't really reflect risk because they are trend changes in macro factors over 30 years and risk credit specifically is easily observed over the last 30 years with three systematic recessions that are driven by unexpected shocks that's risk so is it not true that the next step for climate stress testing and scenario generation is to add risk because there is no risk now that's the problem Yeah, no I completely agree the interface scenarios are sorry They're smooth they don't have the volatility that's required absolutely and in fact I would say we've concentrated so much on 1.5 degree pathways 2 degree pathways we should be looking at 3 4 degree pathways properly not incorporating the full range of what we can expect which will be unfortunately migration conflict the whole thing it's not going to be 4% drop in GDP by the end whatever William Nordhouse or it's going to be much much worse than that and just one sentence and if we don't do that I think we should stop if we're not ready to really do the stress test exercise and let's stay with the capacity building and the education and the looking at things like this but unless we start doing that I don't think there's a point to continue Chris and you could turn the question around and say okay well if there isn't any risk included in the scenarios and still look at the level of impairments that come through for a scenario that doesn't include risk there's a lot going on in the economy there's hundreds of billions worth of write downs that's before you include trading with scenarios and that's if you assume they've spread over 30 years but they'll be concentrated in certain sectors, in certain regions so I don't disagree that the scenarios could be more risky of course they could but for a non-risky scenario if that's the case they're coming out in the back a bit quite notable Now in the ECB How is the question over there in the middle, thank you Hi, thank you I have a question relating to the previous one which is how do we include this risk like Rama was talking about the fact that we are currently just centering the scenarios how do we get the volatility to have the whole ensemble of scenarios so that's one question and the second how do we move from reporting which is currently done to enforcing actual action at the part of the financial institutions Thank you So how do we include the risk and how do we actually lead to action from reporting Maybe I think building on what I said previously I think that there are two two really sources of risk in such a type of model one is exogenous risk for example physical climate risk but I think the driver of the main driver of risk in the short term which is what really focuses the attention of financial institutions is here the reaction of the financial institutions themselves to the announcement of this climate risk at some long horizons so the rebalancing of portfolios which will cause impact on the financial markets in the short and medium term and that needs to be modeled in some way in order to make these models realistic otherwise so there is this external volatility of course you can always input market volatility into such a model but I think even greater impact will come from the rebalancing of portfolios and the shift of investment strategies in the short and medium term that's one of the things that we should try to understand and that's exactly what transition risk is about What about the leading to action Anybody want to take that part of the question I could take that so we are taking action already we've got supervisory regulatory expectations on the banks and insurers that this nest into so we said those in 2019 we said that all of the banks and insurers across the UK have to be doing scenario analysis to inform their risk processes so we are rolling those out at the moment all firms had to have that in place to some degree with appropriate ambition at the beginning of last year so we see these steps as being absolutely complimentary I can speak about that for a long time and get very boring about it so I won't but we see that as the absolute linchpin of how we actually enforce an understanding of risk and the embedding of scenario analysis There's a question at the back Thank you Thanks, I'm Tom Fult from Maximum Information This has been a really great panel actually and I've really appreciated the diversity of the opinions and voices on it I don't think I've heard an answer though to the question of what's next after Sea Bears I think I'd pose that directly to you Chris and Nicola Ryfdy was here but maybe to represent that there's a lot of us in the market looking at what's next, what comes how to respond and how to plan a free plan for what's going to happen Are there any plans in place? Chris what's coming next? Thanks to the question So we've said publicly that what we want to do is to ensure that there is a period where firms can embed the significant learnings that they need to undertake if we run the exercise every single year we just see very similar results every year So what we need to think about is what is the appropriate frequency for a climate exercise 2, what does the climate exercise look like? Does it look exactly like the Sea Bears with the same objectives? Do you look to focus in on some of the other aspects that we've talked about shorter term trading books all of these questions need to be considered I think the third point is at the point that we do it how does it nest into all the other aspects that firms are being expected to do bit transition plans, bit regulatory principles those kind of things So I guess my overall answer to that is watch the space There's a question from this side Hi Albertine from Can I go through a little asset management? My question really goes to kind of circularity of risk here so one of the things that was mentioned by drama is getting asset managers allocation into the models for example but also at the same time that our current scenario analysis aren't really accounting for risk so what is the risk of putting scenario analysis results from insurance and banks into the market which aren't risky enough in terms of asset managers being able to understand the risk because you look at the risk they don't look bad but then they're not actually considering the things Jack is talking about which is contributing to drought etc That's the third point, full solution Anybody want to address that? I think it's a great point actually and there was a danger from the regulatory stress test that they haven't been stressful particularly when people are going on what's the problem and that kind of plays to your point Jacob that don't bother you're going to be like that we need to actually ramp up the stress so I think I completely agree with you and the other thing there is a danger that the financial system now is kind of really moving towards just using the NGFS scenarios so we've seen it in our own surveys and the problem is if those scenarios aren't risky enough they're not extreme enough then we're going to be complacent so we absolutely need to introduce riskier more volatile scenarios so completely agree with that the other thing which is a slightly different point but we talk a lot about reporting as if it's going to save the world and it's not it won't sorry to tell you that now it makes things sound like it isn't real the thing is reporting is great and there's a process of putting together a TCFD report, a transition plan which is incredibly beneficial but let's not pretend the reporting is going to save the world or save humanity it's absolutely not what we need is we need robust policy and we need to get on with this as soon as possible because the more we delay the more abrupt the transition will be and the more disorderly and the more risky and the more costly to people who are putting together TCFD plans they get their lawyers to go through them they look at anything that's remotely risky or market moving they just strip it out, the lawyers just get there so my question is how much value is there in the scenario analysis that we're seeing published I don't know but I don't think that these reporting frameworks are necessarily they're part of it but they're absolutely not going to solve the whole problem briefly and I think just to be clear because of the language we use the point isn't necessary to make the more risky per se this isn't oh we don't see enough risk so let's add in some risk until we get that this isn't sort of a data mining exercise it's to reflect the risks better I think this is the important message I had a 1.2 billion potential climate refugees by 2050 there's not a single economic model on the IPCC that tells us what might happen economically in this scenario which exists in some academic research studies this is the key it's reflecting the risk properly not mining the numbers in order to give it the risk that we'd like to see because to be honest it's great if we don't have a lot of risk at the end of the day isn't it it's just not great if it's because we do an incomplete exercise this gentleman here is dying to ask a question so you've got my last word thank you Mike that is what's death after sea bass I'm going to encourage Jacob in the scenario work we are doing El Nino figures it is quite like the 1998-1991 will be exceeded in the scenario work we are doing La Nina follows it the pension fund is looking at the real world sorry Mike Tark we know who you are they are taking the real world there are no numbers let me say that again there are no numbers they are looking at the real world I disagree with the quotation of the talent economist that you presented if you were a board before climate scenarios came along did you get modelling to make strategic decisions around the board table? No you thought about the future to Joe's point you thought about the future numerically deal with risk you might have considered it but you thought about the world you were heading to the financial sector has sadly entered the world of financial economists who are seeking to model something that is probably not modelable often trope reelected mass migration the question is do the panel believe that a large greater proportion of narrative scenarios are what comes after C-Best was good C-Best scenarios are very poor for decision making so the question is do we need more narratives? Do we need more narratives? Yes So first let me say I agree with your disagreement with the Italian economist quoted in the FT we absolutely do not need assigning probabilities to scenarios to do risk analysis in fact the whole methodology of stress testing is you have some stress scenarios you look at your exposure to the stress scenario you don't need to have a probability and often it is impossible to come up with plausible or realistic probabilities associated with any of the SSP scenarios and even less if you start adding market variables so we don't need probabilities so do we need more realistic stress scenarios what was said in the panel we do need I think you can see C-Best as a useful first step for capacity building basically the idea is if financial institutions have the technology and the channels to assess their exposure to the C-Best scenarios they can also use the same technology and channels to assess their exposure to other hypothetical scenarios which we may come up with later for more imaginative and in that sense it's useful is it a stress testing exercise as we said as Joe and I said in the beginning we think no it's not a stress testing exercise it's just a scenario analysis exercise which is useful to have and let me repeat what I said in the beginning unlike current stress test conducted in some financial institutions where you have some idea where the risk lies because you're long equity so the index falls is a good stress scenario here we don't really know exactly where the risk lies there are many many factors influencing each other in this in this universe so we need to consider a very broad range of scenarios and then expose the output of the exercise will tell us which scenarios are risky provided we have modeled them realistically and all the things inside so it's very important to consider a broad range of scenarios otherwise we cannot have final words so I think it's I wouldn't call it either or but we definitely need to do better at narrative just one example I think is inevitable policy response scenarios which now have dedicated narratives that don't try and put a number on everything but tell the storyline of the transition together with the scenarios but I wouldn't give up on the financial side just to suggest because we haven't really tried to integrate social nature risk into these stories yet so we might as well give it a go or narrative conversations for sure and IPR I think is a good example to point to here Chris? I agree we need more narratives we need more choices of scenarios I guess the one thing we need to balance though is we need to have some form of comparing scenarios across institutions as we know at some point firms can make anything happen with this scenario because often it's what's under the top level narrative where key modelling decisions happen so we need to make sure that people understand what they see is being modelled when they see the results coming out but absolutely agree we need a bigger suite of narratives and scenarios Thanks Chris I think we're going to stop there we could carry on forever but we'd rather not I think one day I've learnt that we need champagne as well as prosecco and that Santa Claus really doesn't exist sorry to tell you that so thank you very much for your questions I've been an excellent panel thank you everybody for an excellent presentation it's now lunchtime everybody so just to let you know that the keynote at 2.15 will also be streamed live in the Geoffrey Mitchell Theatre which is down two floors so thanks again to panel and enjoy your lunch