 eloquently describe two views of the world, the Schelling view, to characterize the first view which is that potentially you should grow your way out of poverty and that should also address the issue of climate change. But I gather from your model that the key assumption there is that as you grow you build that adaptive capital. So of course if you assume that there's you know by growing you're not you know there's no assumption here that you know growing you know in a way leads to a build up of adaptive capital you get the conclusion that of course you need to do something about adaptation and build that stock separately from the way you grow. Because I believe a Schelling view is that there are elements in development that induce the build up of an adaptive capital in I guess that's probably the view. So I guess it's you know if you make that assumption the model is very different and then you... Well I think the issue here is to be very clear about what we mean by adaptive capital right. So in the model it's a very very stark comparison and I sort of said it during the presentation is that adaptive capital is something that gives you no development benefit whatsoever okay. It's only something that's beneficial if the climate changes. So yes it's certainly true that if you build a dam it has a developmental benefit right. But it's only say the extra foot of the dam that protects against sort of water availability variability that you build specifically to counter the change in climate in the future that we can't as adaptive capital right. But the other type of capital doesn't have any adaptive... No it's a very strong assumption it's not it's not a strong assumption it's a necessary assumption in order to talk sensibly about what you mean right. Because if you're talking about if you're saying the vulnerable capital has adaptive capacity then what you're saying is that really it's adaptation right. Then you're saying that the that foot of the dam is in fact adaptive capital. So it's a sort of difficult thing to get your head around because the capital is not actually different pieces of capital. Each piece of capital has both... Potentially both adaptive and vulnerable components. And we're talking about you know how much of each of those things do we need. So it's not... I don't think it's correct to say that the vulnerable capital sort of has no adaptive... I mean the key point is that when you make that distinction cleanly and you allow the model to choose those variables itself optimally then the results are make a lot of sense right. Then it's really you're sort of getting at exactly the kind of trade-off that you want to put in the model yeah. I turn to Finn. Thank you very much. I get your point. I mean did you think just for example thinking about roads right. I mean you can climate-proof roads that that would then as I understand it that would then be the adaptive capital that you're adding right to the vulnerable capital and then you're kind of trying but I mean to which extent are you sort of including this thing that okay as you move forward then you might adapt by not putting the road the same place. Right. How does that sort of come into your thinking about it. So that is it's a sort of level of detail that the model doesn't resolve right. It's a very high level macroeconomic model which really can't speak to sort of locational issues at all. Nevertheless I would say that you know that if you are trading off the position of a road so imagine that in today's climate the road would be optimally built in position X and when you consider that the climate in the future may change you build the road in position Y and that incurs some additional cost right. Well that's a cost that means must be sustained through adaptive capital and in fact the position of the road is in some sense an adaptive capital right. So that's really the level of at which the model can answer that question it's perhaps you know very unsatisfying from the perspective of that particular example but at the macro level you need to sort of make this very clean distinction between the two types of capital. Following up on this question of adaptive capital I understood your definition to be the kind of investment which ascertains a protection in the future should the climate change. Right. Now applying that idea to the question of afforestation we have vast areas of land which are degraded or need afforestation because they're totally destroyed. Now the question that arises here is whether we shall have forest with monocultures for certain benefits like fuel wood and so on or should we have forest and try and get back part of the original natural forests. Now when you come to climate change the danger of this kind of adaptive investment is that there is a fear among many that when climate changes the natural forest might be partly destroyed. The composition of the plants would change usually degraded in terms of biodiversity. If you put a monoculture the trees used for monoculture afforestation have got a good past track record and probably we can predict their track record in a future climatic change situation. But it's rather expensive in a way to invest this so-called long-term adaptive capital when you don't really know what will be the result and your whole investment might be totally destroyed because those particular plants trees also cannot stand up to the change because they haven't had time to evolve a change in their capability to sustain in the new climatic conditions. So I just want your comments on that. Well I mean I would say that it seems that the sort of core issues that you're raising there are issues to do with uncertainty. So really about how do you and I would say that really this model is just not appropriate for thinking about very sort of small scale impact specific issues like that. But there is of course a large literature that looks at sort of optimal investment when you really don't know what's going to happen and moreover you're going to learn in the future what's going to happen. So there are sort of real options techniques. There are sort of quasi-option value techniques that allow you to sort of build into your investment strategy that the value of the new information that you're going to get in the future as the climate changes and it's revealed to you how bad it's going to be. And I would say that for the kind of problem you're describing those methods are probably much more appropriate. Now it's true that as I said at the end building uncertainty and irreversibilities and things like that into this kind of model would be desirable and would likely change some of the conclusions. Which way that would go I really don't know. I mean that's a sort of frontier of current research. Yeah but I mean I guess the short answer is that you know it's just it's a mismatch of tools in this case. Thank you very much. So we have time for one more question. Yongfeng. I'm Yongfeng from Yuanyu wider. Thank you very much Tony. And I think your work is of great quality. I have two questions. I can see you basically modified the Ramsey model by introducing the adaptive capital in the temperature to the model. So my first question is could you roughly describe the balance growth path of your model as well as the equilibrium in adaptive investment, adaptive capital. It is steady state. Now second question. I can see the concept adaptive investment is very interesting. Which could be regarded as the climate finance for adaptation. But in your paper you basically say the change in adaptive capital is the difference between the adaptive investment and depreciation. And the change in the normal capital should be the difference between the normal investment and the depreciation as well. However the normal investment has disappeared. Could you describe the difference between these two? Okay thank you. Okay let me start with the last question. So the normal investment hasn't disappeared we've just rewritten it using an accounting identity. So there's an identity between investment, consumption and output and that's an identity that holds throughout time. And I've just written the equation in terms of the consumption control variable instead of the investment control variable. So you could write exactly the same model using an investment variable and you get two equations that look identical. But then you need to enforce the accounting identity at every point in time. And it's just computationally and actually analytically easier to use the consumption control variable instead of the investment control variable. It's identical. So let me show you just quickly. Yeah we have to wrap up. Okay so on all right so all right the point is that basically you get two equations that look exactly like that. One investment in a vulnerable one investment in adaptive. But there's an identity that contains all the other terms that and that's an exactly equivalent formulation. The issue of steady states you know this is something the macro economists do a lot of. They sort of write down a model like this and they forecast right into the infinite future and look at what happens in the steady state of the economy. The issue with the climate problem is that it's completely irrelevant what happens in the infinite future. What you care about is the transition between today and the future. All right so the infinite future is heavily discounted in the model you really don't care about what happens. And moreover it's the dynamic aspects of the problem. It's the fact that the economy is adjusting dynamically to the change in temperature that's really interesting in this problem. So yes you could linearize around the steady states and calculate it and look at the stability. But it's really immaterial to the actual policy question that you want to ask which is what is the optimal dynamic trajectory of investment today and for the next 100 years. That's not a steady state question. All right. Thank you very much. I'm afraid this is all the time we have for question. So I want to use that opportunity to thank the presenter and to thank the audience for a very rich discussion. Thank you very much.