Harrison Hong, Professor at Columbia University and who won Fisher Black metal awarded to the most outstanding financial economist under the age of 40, presented at Luohan Academy's Frontier Dialogue. Lars Peter Hansen, a distinguished David Rockefeller and a professor of economic statistics at Chicago, discussed Hong's presentation. The following text features transcripted excerpts. It has been lightly edited for clarity and length.
Speaker presentation by Harrison Hong
The next topic is on "Sustainable Finance and the Transition to Net-Zero". Professor Harrison Hong is going to be the presenter. Harrison is my colleague at Columbia and my co-author. He won Fisher Black metal awarded to the most outstanding financial economist under the age of 40 by the American Finance Association. Harrison, please take it away.
Let me start with the observation that the world has been trying to pass some form of carbon tax for close to 50 years. Obviously, it's not been very successful. So now sort of in the last 5 years, you're seeing this immense pressure on the financial sector and regulators to do something to basically keep global warming as to meet kind of pointed out somewhere to be between 1.5 to 2 degree Celsius. There's not a lot of room left. And sort of one of the widely discussed solutions in the financial industry are these sustainable investment mandates where the idea is that wealth of households, wealth of institutions, or institutional asset managers can only be invested in companies that meet net-zero emissions targets by some day, 2030, 2050. The amounts of money are quite sizable. I think that the pledges are quite sincere.
So for instance, one such initiative, is the Net Zero Asset Managers Initiative, exactly outlines these mandates. It's got about 87 signatories, accounting for $37 trillion in assets under management (AUM). And that's just kind of name one initiative.
And there's also now basically something called "the Network for the Greening of the Financial System" that basically encompasses banks, central banks, where the idea is exactly the same that, rather than going to equity markets, you might go to bond markets and impose exactly the same mandates.
The way that this works in practice is that should your firm to be qualified to be held by this pool of money. You need to contribute to sort of what we'll call the decarbonization capital stock N, as opposed to sort of typically making investments in productive capital stock K. In practice, what this means is some flow mitigation spending. The firms have to engage in a portfolio of decarbonization measures, like whether there's like planting trees, regenerating forest. It could be some carbon sequestration. Give you a boundary, carbon capture, it could be air capture.
I think the important thing to note is that so IPCC in their last release said that almost all the pathways now that one can envision in terms of meeting these targets involved, basically accumulating, a fairly significant amount of the decarbonization capital between now and in the next 10 to 20 years.
So the idea of these mandates is that it's going to send price signals for firms to decarbonize. The problem is that there's not a carbon tax that's sending the signal. You need some other signals in the system so that firms can do something, can have something instead of it. And we have to take the view that firms are profit-maximizing. So they have to be sort of indifferent between being sustainable or not. Right? It's got to be sort of this voluntary action that these mandates are presumably going to send these price signals, and then there's going to be some equilibrium tradeoffs that the firms can engage in what they're going to be indifferent. And the way it's going to work is these portfolio restrictions of investors, which also implies some shorting restrictions. In equilibrium means that the sustainable firms are going to get a lower cost of capital, or either going to get a higher price. Right? And that's going to basically offset and pay for their contributions to decarbonization. In other words, sustainable finance is a form of a capital tax. That is kind of premise on being sustainable or not.
I did not coordinate with the meet on this, but I like this slide because Microsoft is one of the firms that are very enthusiastic in supporting such a solution, essentially. Right? What I like about it, it's fairly specific. Right? There are numbers. There are some at least sort of projections and some technology or some accumulation of capital to indicate how this is going to get achieved.
Since we talked about this, I'm not going to have to spend much time except that the green is exactly what we're talking about.
To meet earlier thing about why Microsoft is doing this, one of the reasons I think is exactly the sustainable finance mandate. There's a huge incentive for firms exactly, because in fact, these mandates of trillions of dollars are going to be enforced. That's going to move the needle. But it does raise a lot of difficult questions.
For how effective will these mandates be compared to sort of what we might think is the planners for his first bet? How much sustainable finance do you need? How many years do you need for this sort of transition in this accumulation of decarbonization capital stock? What happens in the interim, the economic growth when you're building up all of these forests? What are the consequences for firm cost of capital? Who ultimately is paying for all of the stuff, basically? Okay, and these are the difficult questions.
What I'll talk about is to give some preliminary answers using kind of a growth model. That's going to feature what we'll call kind of decarbonization, which is non-productive capital stock, versus like a productive capital stock, what we typically work with. This is based on a paper, examining the welfare consequences of sustainable finance with Neng Wang, our moderator and Jinqiang Yang.
The basic idea is the following. You start with a growth model. Okay, that sort of often used in in terms of climate economics. And you're going to have sort of these extreme temperature events. Let's call it 2 ℃ above the country's historical normal, right? In years, when you have lots of these extreme temperatures, that's going to have some destruction. You can kind of pick, you can be like destroyed capital. We could destroy kind of some productivity. But regardless, it's going to increase a lot of aggregate risk in the world. So think about firm emissions is roughly proportional to this, to their productive capital stock.
There's going be from a planet perspective, there's a big willingness to pay for some decarbonization. But there's going to be externalities, the commons problem, as Steve pointed out. There's going to be some costly and gradual adjustment of both capital stocks, right? So it's not going to be, we can't build all these forest overnight. Can't plan on this forest overnight. It's going to be some adjustment cost. And sort of. Ultimately, if you kind of believe in the science, the ratio of decarbonization to productive capital stock, which we'll call little n, is going to be the main state variable and helps mitigate the damages associated with the emissions.
And then firms are gonna get qualified on these mandates based on their flow contributions to this stock. The model is not perfect, and it's gonna miss some important elements that are kind of climate economics. But I think what it can do though, is to answer more of these practical questions associated with sustainable finance. The way we'll do it is we'll kind of calibrate the model, using historical data on how excessive annual temperatures at the country level reduces GDP growth.
Now, obviously, these 1.5 ℃ events, 2 ℃ events are relatively rare in historical sample. But when they do happen, there's damage. And we're going to use data as kind of an input into calculating the cost and benefits of this decarbonization capital. We'll combine this with some parameters from kind of a long run risk literature, from asset pricing essentially, that basically kind of gives a fairly high willingness to pay on the part of agents for reducing risk. Then we'll get some adjustment cost parameters from the literature.
Obviously, we have much better estimates for physical capital K than we do for N, since we haven't built up a whole lot of decarbonization of capital in our history. So there's going to be some uncertainty there. Then we'll pin down the costliness of accumulating this decarbonization by assuming that in the past, optimal mitigation was kind of negligible, because things were not that bad. Basically. And they will ask, well, now, as the world gets kind of worse and worse, right? I think this is the moment that we're in. There's obviously increasing incentives to begin this accumulation. And so we can kind of then ask a bunch of standard, try to answer some of the difficult questions. I believe it, too.
So I think that's the set up. There's kind of a mixed bag. I think there's good news for mandates in the sense that in our calculations, in the steady state, you need like 6.25% of this n*.
Think about $600 trillion of physical capital. That means you need about $34.5 trillion of this decarbonization capital stock. That's quite a lot of money, because each year that means the world has to spend something like $3.86 trillion annually on this. That's a big price tag, to kind of Steve's point. Those mixed news for growth, I mean the European Commission, I think has this projection saying all the transition to the net zero is going to be super rosy because I think there's an assumption that you can do all of this with just renewables. Right? But I think kind of you take the decarbonization capital stock view that the IPCC has. This is not very productive capital. So there's going to be basically some economic trade-offs, although the benefits in the longer run could be quite substantial.
And then for institutional investors, it's pretty costly, too. There's a lot of rosy projections from Blackrock and others on all these benefits without articulating what the numbers are. Our model gives you some numbers. We generally think we can get 20% of global wealth to be indexed to sustainable firms, to send these market signals. The representative investor has to give up basically about 3.25% of expected returns on this 20% of wealth. Right? And these qualifying standards, are going just to rise over time, because basically right now nobody's doing anything, so you don't have to compensate it, but so to be calling yourself sustainable is pretty cheap because nobody's doing anything.
If we're serious to get to like $34 trillion where the capital stock, people have to start spending some money. Right? That has to get paid somewhere in the system. The investor, the representative investor has got to basically give up 3.25 percent of yield to find this. There's a lot of caveats which I think Lars is going to talk about. There's a lot of parameters that we have to kind of deal with, et cetera. But these are very preliminary and give you the gist at least some preliminary answers to these difficult questions. Thanks.
Discussant presentation by Lars Peter Hansen
Thank you, Harrison. Discussant is Professor Lars Peter Hansen from University of Chicago. Lars is a distinguished David Rockefeller professor of economic statistics at Chicago. He's a Nobel laureate, and also a member of the academic committee of Luohan Academy. Lars, please take it away. Thank you.
Lars Peter Hansen:
Sure, thank you very much for this opportunity to talk about this very important subject. Reading Harrison's slide. Let me go and read some papers written recently with various collaborators, I found very interesting and fascinating. It's obvious a very important topic. To address climate change, we think of this as an externality. Therefore, we have to think about for a commons' perspective about how we made restrictions. As Harrison talked about, things like carbon taxation has been somewhat frustrating in terms of its ability to be implemented, cap and trade policies and then certainly slow down, sort of materialize. My own work has been connected to things like uncertainty contributions to hypothetical social cost of carbon, and which can be substantial. But I agree that's going to policies to send. It's been very slow to be adopted. Some much recent discussion about central bank policies about what central banks can do.
We are at the times so we might be asking central banks to be doing, hoping to them to be doing too much because there's somewhat distant from the political arena. We could have a very interesting conversations about that, certainly play walls and financial stability. Let me go to the very interesting. A policy approach that is discussed by on in this paper, and in this talk, as well as by the two coauthors, and the quantitative work that follow.
Let me say at the outset that I really appreciate the fact that the authors are writing down this kind of fully specified dynamic stochastic model. It's got some nice, innovative.
It builds extensively on some earlier work of on the walls with multiple capital stocks. And that what was done just several days, she said it's built on some stuff that means done with them. I can't say certainly, but this paper who pushes it in a very important and positive way. I appreciate that aspect of it. The intertemporal technology here that underlies their calculations. It's a productive. That's all statistically. There's an external decarbonized capital that also evolves sarcastically, and that would drive down to admissions. We take output and split it between the consumption and investment, productive capital, and investment and decarbonize capital. There is a jump process that then governs the stochastic degradation of the capital stock. The basic research allocation problem is what fraction of output should be allocated the mitigation investment? And how should this be achieved? So this is not about what firms do to look at their own long term interest. This is for the person. You're discussing that in the context of Microsoft. This is about the section amania.
The part that firms do not enter the price, and then, how do we get resources into that? How should we proceed? And so that there could be a fictitious planner that figures out how to do this take him account the externality. Then we designed policies and the authors consider such a policy in the process. So this type of modeling you know, at the end of the day, this is a quantitative calculation. Very important. It's an engaging kind of quantitative policy analysis. How do we proceed? Now, to proceed, to go forward, modeling this area, you either turn things into enormous black box. This is true in my own work and it is showing the case here. I think it's important. We really understand most applications.
So here there's no explicit climate damage. There's no temperature anomaly, carbon, the atmosphere, not state variables. It was somewhat distant for talking about things like 1.5 or 2 degrees Celsius, targets and thresholds. This makes it difficult to relate to climate science evidence, which is the top one. And this is the and it makes it difficult to talk about temperature anomalies and thresholds. Maybe those are overplayed. It could be the speed that could be the argument as well. Now, the
damage, that there's a disaster damaged probabilities are specified here that depend on these two different types of carbon capital stocks, this is unproductive. The carbonized people, unproductive to the firm. And there's this productive capital. This disaster will depend on the ratios of these. But that doesn't mean environmental externality that triggers damages in the capital stock, that's related to the ratio.
Now, model takes things that commissions as some exogenously specified entity. In the papers, homogeneous of degree one of the two capital stocks. This means that the externality cells, the damages are homogeneous of degree zero, not degree one.
Let me just say that you shift the overall scale of the economy. And to me, that's a rather big tension because there's a lot of climate evidence which is all about how we go from emissions to temperature change, temperature change to potential economic damages. This is where I think there's a bit of a modeling tension in this paper. The one that I struggle with in terms of using for this particular point.
Let me go for a little bit and talk a little bit about the type of evidence that one might want to think about integrating. This is some work we've tried to integrated all the concern. Carbon models, temperature models, and carbon cycle models are highly complex. You couldn't possibly put them inside of the economic dynamic models, delete you have to use complications. Climate scientists themselves have had a device in very interesting ways to do a model of comparison across these different modeling groups. We do things like they run through scenarios, they also do pulse experience. What I'm showing here are the optimal pulse experience. What you do is you put in an emissions pulse in the atmosphere. And then you look across, in this case, a hundred, and I think it's 144 different model.
And these are the responses that come out. This is appropriate adjust units here, which is making where we don't have 5 years. But there's a common pattern on these models is that if you make carbon in the atmosphere, you get a peak effect. You know about a decade, then it flattens out. Now, so what I'm showing you here is the mean effect across 144 model conversations plus the 75th and 25th percentile. And then there's the upper and lower. This is the type of uncertainty that comes out of the model that would seem to want to integrate into thinking about economic and environmental questions. Notice this depends on temperature. I know these are kind of emissions and emissions here we call homogeneous of degree one, not zero. Here's this hypothetical damage function. Suppose somebody got damages that are being triggered after 2 °.
And then there may be some range of different damage response. Some economists argue the very modest responses, some give much steeper responses. As you proceeded over the last several years in the base, where this team, the threshold that would be two, that would be 1.5. It's uncertainty about where that threshold can be triggered, and then the curvature that follows. This is just an example of the type of damage function uncertainty, on which one could counter. Here, we do with this problem, you know, the probability. But these types of inputs are not really possible to map directly into the on the framework of the criminal, in part, because of the modern assumptions.
And I I'm sorry, I worry a little bit that data transmits into results. Now that the paper features in a very interesting way, this decarbonization capital and carbon capture, and I think these are very important things to be thinking about. I wasn't quite sure how to interpret the quantitative magnitude inside the model, or what this kind of carbon neutrality, or the goals might connect to it.
Now, the model itself has a planner's problem. So maybe we shouldn't care about this, once the planner's problem is correct, we could compare to what the optimal just what comes out of that connection. I gather the spirit of this analysis. This thing features mitigation mandates, which is certainly very interesting. I sometimes wonder though, how do we make these enforcements meaningful? We can kind of announce these mandates, but who's going to force them? How do we know we're going to be making productive investments in the decarbonization? Are these going to be mitigation resources that just get handed over to some type of policy maker? Or it is done by the firms itself? If it is to be done by the firms themselves, how do we know they're really accomplishing? It's very serious money. It's a little bit concerned about that. The bat is promising.
One of those interesting parts of this paper, I think, one is that they build this competitive equilibrium in which there are these 2 types of firms that are sustained. Inside this model, investors do not have a direct utility increase for sustainable portfolios. Other papers put it there exogenously and somewhat mechanical. In this paper, consumers and investors care about future consumption, firms have to actually sacrifice output by agreeing to be sustainable. This isn't about the long-term benefit. They're actually sacrificing output.
Now, this leads to heterogeneous exposures that sustain expected return differences. I think this is a very interesting part of the analysis, how this type of comparative equilibrium gets sustained. I wonder under which policy make it supposed to. I think this is actually the most interesting part of the paper. So concluding remarks, I like the models, it's a two capital model, a very important proposal. I think the capital is really what we're thinking about to include things like green capital, green technologies.
For me, I think it's important to include multiple sources of uncertainty, including sources coming from the data economic damages, from environmental damages, from these temperature and carbon dynamics. I think it's very valuable to these quantitative exercises.
But as a collective group of economists, it would be nice if there are ways to do more serious multiple comparisons. The models that I've written down have their own set of shortcuts. It's very easy to pick up into a phenomenon. And I said where they fall short and what question they can address. It's the way, value be looking across models and each researcher building his own model. Where's the research that's really doing the meaningful cost of cross market risk? And I think it's really important for policy analysis.
Finally, I worry about the meaningful implementation of this mitigation. How creditably they can be impressed? I think that author, I think Harrison talked about very important problem, and I think he's put on the table a very interesting policy. Thank you very much.
For more information, please visit Luohan Academy's youtube channel: Luohan Academy