 and ask the colleagues from the Izabee who have the microphones to pass them on. I would say Richard has his hands up, I see Adrian, and afterwards, Francesco, please. Thank you, Stefan. Well, first of all, the link between banks and shadow banks, they're not linked anymore, you know, shadow banking has been transformed into resilient market-based finance, didn't we all know that now? Sorry, I'm being a little bit difficult. But what I want to be positive is on Stein's smell test. I like the smell test idea, and it's not just looking for implicit backstops, I give you another clue, if you like, and that is the relationship between risk and return. We regard that as something fundamental in finance. And regulation tends to reduce risk, that's part of the point, and of course, reduces return. One of my favorite examples of that relationship, and where the smell test should have operated and did not, was AIG, AIG's CDS market activities in the period leading up to the crisis. There's this small London office, 20 people or so, producing a very large revenue stream. Nobody in New York on the board or in the management seemed to have noticed that this was a very profitable activity. What was going on that was so profitable, the answer is that they were taking huge risks. That was a shadow banking activity, par excellence, and it seems to me that we might add that criterion, as it were, to your smell test toolkit. Do you want to respond directly? No, no, I agree completely. The market discipline doesn't always work, and there are some obvious signs for sure exposed, but even ex-ante. I would even take a broad, I mean, we don't necessarily always try to make the market discipline the system, and we know there is a volatility paradox, et cetera, but we could do some data provision for sure help the market better discipline, I would hope. We could also think of some structures, mutual insurance, I mean, if you go back centuries ago, there was a lot more mutual operations in terms of clearing houses and what have you, that exercised quite a bit of discipline in the system. We've done away with that, and we've replaced it largely by the states, back to the backstop, but we have to introduce that to some degree, because we're never going to be fighting this war and winning it if we keep relying on the regulators to do this. We have to allow the market to do a little bit better, and that goes back to the data. We don't necessarily provide analytically useful data for the market to identify these risks, and we're always hiding behind, oh, it's confidentiality and what have you, but at some point we have to stop saying that argument, because we're not going to win the war this way. Adjo. Yeah, I have a comment on the modeling, so I think in the general equilibrium exercise, you want to think about additional frictions as well, like behavioral biases, like neglected risk, or contractual incompleteness. I think that shadow banking per se might not be dangerous, but if it's combined with things like biases and ratings, or counting ways of hiding risks, or behavioral biases that drive up asset valuations above fundamentals, it's that combination that makes it dangerous in the aggregate, so it might be useful to study that. Yeah, I completely agree, I mean, definitely, yeah. Now, next one is Francesco. Thank you. I remember the days when we started this institution, at that time we were all reading the papers of Gorton saying that basically the crisis has been a run on shadow banks, and we were mostly worried about money market funds as being one of the first recommendations we have been issued. Then at a certain time, a new phase came in which people were saying, well, you know, it is very unpolite to call these people shadow banks because it's almost a terminology which is negative, we should call market finance. And now if we will understand basically the thrust of the discussion is that you can continue to call them shadow banks because they are not dangerous. Now, what I'm asking myself, and if any, the things which you have to do is an empirical work to look at the numbers, but at least in line of principle, more shadow banking is better because it is less dangerous than banks. So I would like to ask them the question, which is to a certain extent a question I've asked myself also in terms of policy work. Does it mean that systemic risk is originated only from banks? I guess, Leon, as this goes right, I'd like to do you. Well, I mean, it depends. I mean, of course, not always, right? In the US, certainly not. But it just goes at the heart, and then it goes also to what Tobias said. I mean, it's not so how current modeling frameworks haven't agreed upon yet a consistent behavioral framework that would encompass all those issues that Tobias was talking about. So this is hard to model. And therefore, if you do just the regular frictions that can cause excessive risk taking, you will conclude that shadow banks may not an increase in those shadow bank activity may not be as harmful. But I'm not saying at all. I mean, that's my personal view because I'm very aware of those other features. And I'm thinking it's very important. We need to study and we need to understand this much better. And as a consequence, as we know, that the risk may emit from many other institutions as well, not just traditional banks. That clarification was a great relief, I think, to many people here in this room who dedicate their lives to understanding the system. There is a risk. So there are risks. There may be risks, right? Fantastic. I think David was also interested in asking a question. If I take the BIS data, so global GDP, 85 trillion. Banking assets, 160 trillion. So two times global GDP. Equity, it's 5 trillion. So basically we have a system that is extremely highly geared. The losses incurred in the systemic risk is only a function of one thing, leverage and gearing. Because what you thought was safe, Greek bonds, Argentinean bonds, triple-A mortgages, you have it in massive amount. And that's what kills you. Anything that you deem risky, equity, shadow banking, because you can have so little, it never kills you. So the question I ask the panel, when you say risk today, is the shadow banking system leveraged or not? Because I live in it. And I have 400 of equity, 85 of risk. OK? And when I look at any banking or any insurance, they run between 15 and 20 times leverage. So I ask myself, is the leverage in the shadow banking such a risk? Because operationally, we don't set. And for the simple reason, because we are forced to invest in anything that is risky. Because if it's not risky, someone comes with zero financing from VCB or the Fed. And then, of course, they can win, just maths. And so I'm asking your question, if you're so worried about the shadow and not actually what is triple-A, which is what is me the most. Because then someone wakes up, Korea, Lehman, Greece. Stan, you want to respond to that? You may well be in a segment of the industry that I wouldn't call shadow banking for that reason, because your low leverage, you're doing intermediation that doesn't pass my smell test because you're well covered. And in that sense, the regulation should be proportionate. It's largely disclosure. And the rest is up to you and your shareholders and investors to worry in the governance sense. So that goes back to definition. Yes, we have to be careful about what we define shadow banking and how we apply the regulation accordingly. So there are some commands that shouldn't be there. Having said that, your interactions with other parts of the financial system, not you personally, but that part of the industry could create prosyclicality that should be a concern to us, because, again, we can get an overall bad outcome. There, the toolkit is a little harder to apply. So I'm not yet ready to say, OK, we're going to now regulate that part in the following way. But nevertheless, we should be wary about it. And certainly, I can say a word about it. Leverage, in any of those entities, is of central interest to us when coordinating and defining our analytical work. And I'm speaking here for the ISABEE, for ESMA, for other institutions involved. It is the core identifying element for our risk analytical work, both at a monitoring level, but also for individual studies. And to give you just an example, we are concretely looking at leverage in hedge funds, for example. And that brings us just to say that at a number of very interesting problems. I mean, first of all, we need to differentiate between conventional financial leverage on the one hand and then synthetic leverage through the use of derivatives. For example, we have a European regulatory system where leverage is not in a unified way defined across entities. So we're dealing with different technical definitions of leverage, even if we as analysts have a joint view of what it is. Technically speaking, it is different for different entities and so on. And then on top of that comes the data question, I think, that has been highlighted by all three of our panelists, Stephen Juliana and Stain, that yes, the data are becoming available, but it just will take a lot of time to become as routine and as perfectionist as our banking supervisory colleagues and the central bankers are in monitoring the banking system. So it is for us a real central concern. I see no more direct hands up in the audience, but let me therefore ask a closing question to the panel. And this is more like a more popular sort of question. Maybe each individually, what do you think is the most pressing risk at this stage in the shadow banking system? And I will start maybe in reverse order with Stain and then Stephen and then Juliana. This is too low of the question. That's not, well, I'm going to cop out on this one. So I think we have to keep monitoring things on a regular basis. It's very difficult to tell. I think we have a situation in which we have elevated asset prices around the world. There's a lot of disconnect between what we see in terms of the financial markets of volatility and what we see in terms of policy uncertainty. So as a general observation, I would say we should be wary. And to be as it's always been wary. So it's sometimes the crying wolf, but I think this is a particular time when that may be a good thing to do. Stephen. Yeah, let me pass this question because I'm waiting for the data. Juliana, waiting for the data too? No, you've got some. Well, I can reiterate what I would say before. That I mean I also would wait for the data where actually it shows itself. But I think conceptually it's anywhere where those behavioral aspects can come into play. So what are the assets or constellation where we think we are safe? So the wrong belief of being safe is typically what triggers the problem. So availability of data is a determining factor. In this game, I think that is one major takeaway from this discussion. We know that more will be coming. So maybe it is both things, having more data and also having the capacities and the resources to actually tackling them that will be important going forward. Also take it from your findings and what you said that the definitional question remains acute just because it is a moving target. It remains a mutating concept, not least in light of financial innovation, correct? And that will need to remain active and agile about exactly understanding what shadow banking is and how we define it. And that in turn is important for us to identify the main risks that we actually want to get our heads around. And the items that we have tackled today was in particular the risk of arbitrage from within the banking sector, outside the banking sector. And we've seen some very interesting insights both from Juliana and from Stephen. And in addition also questions of leverage and liquidity in the system and the interconnectedness of those entities, both in our own jurisdictions but also then in international connection. With that, let me thank Juliana and Stephen and Sting for the great presentations, for the great discussion. Thank you very much. And also to the audience for the active involvement. Thank you.