Investors don’t have the consistent, comparable and reliable disclosure that they need to efficiently price risk and allocate capital, and issuers likewise are beset by numerous and sometimes conflicting demands for information under the various voluntary frameworks or in the form of bespoke requests from investors, raters and others.” SEC Commissioner Allison Herren Lee, Remarks to Stern Volatility and Risk Institute Conference, April 30, 2021.
Climate change is the topic of the day. The World Meteorological Organization tells us that the 2011-20 decade was the warmest on record. Earlier this year, the U.S. government re-joined the Paris Accord, and is proposing a range of new programs to mitigate the long-run impact of climate change. Now that a warming planet has made the Arctic increasingly navigable, national security specialists are concerned about geopolitical risks there. Thousands of economists have endorsed a carbon tax. Even central banks have joined together to form the Network for the Greening of the Financial System—a forum to discuss how to take account of climate change in assessing financial stability.
Against that background, last month, NYU Stern’s Volatility and Risk Institute (VRI) held a conference on finance and climate change. Speakers addressed issues ranging from the modeling and measurement of climate risk in finance to assessing its impact on the resilience of the financial system. In this post, we primarily focus on one of the central challenges facing policymakers and practitioners: what is the appropriate discount rate for evaluating the relative costs and benefits of investments in climate change mitigation that will not pay off for decades? We also comment briefly on several other issues in the rapidly growing field of climate finance research.
Past responses to the discount-rate question vary widely. Some observers call for a discount rate matching the high expected return on long-lived, risky assets—a number as high as 7%. This would imply a very low present value of benefits from investments to mitigate climate change, consistent with only modest current expenditures. Others postulate that climate change could lead to the extinction of humanity. For plausible discount rates, the specter of a nearly infinite loss means that virtually any level of mitigation investment is warranted (see, for example, Holt).
Recent climate finance research that we summarize here comes to the conclusion that over any reasonable horizon, the appropriate discount rate for computing the net present value of investments in climate change mitigation should be relatively low. For example, the medium-term discount rate over the next few decades probably should be lower than the safe interest rate of 1% or 2%. And, it could even be negative. The very long-term rate—over more than a century—is likely to be somewhat higher, perhaps closer to 2½%, but still well below commonly suggested rates of 4% or higher (see, for example, Nordhaus).
To appreciate the recent advances, we start by going back to basics. The discount rate is a key factor in decisions about virtually all investments with long horizons. The more distant the payoff, the lower the present value. All other things equal, the discount rate determines just how much lower. For example, in the chart below, the present value of a $1 payoff 25 years from now is only $0.16 with a discount rate of 7 percent; but with a discount rate of 1 percent, the same payoff has a present value of $0.78—nearly five times higher. And, as the time to the payoff lengthens, that difference grows.
Present value of a future $1 return from investment today to mitigate climate change
Source: Authors.
Since the benefits of climate mitigation investments accrue mostly over decades, the choice of the discount rate is critical for any decisions. Fortunately, the wave of recent climate finance research—summarized by Giglio, Kelly and Stroebel (GKS, 2021) in their excellent new paper—offers a great deal of useful and intuitive guidance about the appropriate discount rate in that case. Much of what follows is based on GKS (2021) and on Stroebel’s VRI Conference talk.
Finance theory tells us that the discount rate we apply to future cash flows has two parts. It is the sum of the risk-free rate and a premium that compensates the investor for the risks to consumption growth that are associated with those flows. The implication of this logic is that we should place a relatively higher value (and a lower discount rate) on future payoffs that come in states where consumption growth is low. The reason is that the marginal utility of additional consumption in those states will be relatively high.
Put differently, through its impact on the appropriate discount rate, the covariance between an asset’s return and the growth rate of consumption has a decisive impact on the value of an investment. The reason is simple: assets that offer good returns in bad times are valuable because they allow us to hedge consumption risk. They provide insurance that help us stabilize our consumption path. For an asset with such attractive hedging properties, the risk premium (rp) In the following equation would be negative:
Discount rate (r) = risk-free rate (r*) + risk premium (rp)
Importantly, the return on the stock market as a whole (and on many other, long-lived risky assets) typically moves together with the growth of consumption. For example, in economic depressions, both equity returns and consumption growth usually fall sharply. As a result, one might expect the risk premium on the stock market to be positive and substantial. Not surprisingly, economists like Barro argue that the observed combination of high long-run rates of return on equities and low safe interest rates is consistent with rare and difficult-to-predict catastrophes (like depressions).
The same considerations apply when judging the value the distant future returns from investing today to mitigate climate risk. In theory, two different mechanisms push the risk premium in opposite directions. On the one hand, suppose that economic growth adds to pollution: consistent with the work of Nordhaus and others, climate risks would then tend to materialize in prosperous times, so that investments to mitigate these risks would command a positive risk premium. In contrast, in line with the analysis of Barro and Weitzman, realizations of climate risk could be so calamitous that they sharply depress consumption, becoming an independent source of consumption risk. In that case, the benefits to investments that mitigate climate risk accrue in severe states of the world (when consumption is low). As a result, these investments would bear a negative risk premium.
As GKS emphasize, finance teaches us that these differences in the risk premia have a profound impact on the appropriate discount rate. In the former case, the discount rate applied to climate mitigation benefits ought to be higher than the safe interest rate. In the latter case, it ought to be lower than the safe rate, and could even be negative. Determining which of these dominates in practice is both critical and difficult. Ultimately, however, measuring the covariance between climate change and consumption growth is an empirical issue.
Once again, climate finance researchers are on the case. Giglio, Maggiori, Rao, Stroebel, and Weber combine transaction data for U.S. coastal real estate with geo-coded indicators of exposure to sea-level rise and a range of house-quality indicators to measure the differential impact on house values from local changes in climate risk perceptions (say, after a hurricane). Their results are consistent with the disaster-modeling approach that views climate risk abatement as a form of insurance. This leads them to conclude that the appropriate discount rate would be no higher than the risk-free rate.
How high is that? Markets in short-term fixed-income instruments point to safe interest rates in the 1% to 2% range over a few decades. But, since climate mitigation also has benefits over the very long run—say, more than a century—we also need an estimate of the very long-run safe rate. Fortunately, using information from real estate assets Giglio, Maggiori and Stroebel estimate an upper bound for the very long-term discount rate—the sum of the safe rate and the risk premium—of 2.6%. (Highlighting the persistent decline in recent decades of the equilibrium real interest rate—known as r*— Bauer and Rudebusch estimate that the relevant discount rate over both the medium term and the very long run is in the range of 0.5% to 1%.)
Of course, the scope of recent climate finance research cited at the VRI conference goes well beyond estimations of the discount rate. As one example, Engle, Giglio, Kelly, Lee, and Stroebel examine the problem of hedging distant climate risks. In the absence of markets for very-long dated derivatives, they turn to equity markets and develop hedging portfolios. They start by creating an index of climate change news—in this case, based on the frequency of climate-related terms in daily editions of The Wall Street Journal (see following chart)—and then establish a framework for updating risk-hedging portfolios as news arrives. To the extent that their index correctly signals climate risk over time, this dynamic innovation-hedging strategy mimics a security that pays off in the event of a climate catastrophe.
WSJ Climate Change News Index, 1984 to 2017
Notes: The red line is the mean, while the dotted blue lines show a standard deviation above and below the mean. Source: Figure 2 in Engle, Giglio, Kelly, Lee, and Stroebel, “Hedging Climate Change News,” 2020.
At this stage, research in climate finance remains in its infancy. And, there are surely many challenges ahead. For example, there is considerable disagreement among popular indicators regarding what constitutes environmental risk (see Billio, Costola, Hristova, Latino, and Pellizon). The development of tools for assessing climate-related threats to financial stability has just begun (see, for example, Engle and Jung). And, capital markets regulators are now actively seeking to improve disclosure so that investors can make informed judgments about the climate exposure of their portfolios and firms can better manage their climate risk (see the opening citation from SEC Commissioner Lee as well as the Commission’s invitation to comment).
Fortunately, as the VRI’s conference highlights, research in climate finance is making swift progress as it has attracted a wide range of talented experts. Given the likelihood of heightened interest for years to come, we are both hopeful and optimistic that the advancement of knowledge in this critical area will remain rapid.
Acknowledgments: Without implicating him, we are very grateful to our friend and colleague, Johannes Stroebel of NYU Stern, for a valuable discussion regarding climate finance, as well as for sharing the data for the WSJ Climate Change News Index.