Collaring Price Volatility in a Carbon Offset Market

A major concern with cap-and-trade proposals to reduce greenhouse gas emissions is the uncertainty over the future price of emissions allowances. Many factors could cause allowance prices to spike, such as rapid economic growth and obstacles to the expansion of renewable and nuclear power. This commentary discusses various price control mechanisms and to what extent they might be complementary with another cost control possibility, namely, provisions for emissions offset credits.

One of the key issues in the design of a domestic cap-and-trade program to control greenhouse gas emissions is how to limit volatility in allowance prices. Such volatility can significantly raise the overall costs of abatement over time and deter clean technology innovation.

Price volatility arises because in a traditional cap-and-trade program, annual emissions levels are fixed, but other price determining factors can vary from year to year. In order to minimize the costs of meeting long-term emissions goals, some year-to-year deviations from the annual limits, which in turn reduce allowance price volatility, should be allowed. Ideally, the discounted cost of reducing the last ton of emissions should be the same across different years as would occur under a carbon tax that rises over time at the rate of interest.

In contrast, the cost of the last ton reduced at different points in time (under a traditional cap-and-trade system) can be quite different. For example, allowance prices can be relatively high during times of rapid economic expansion or when new low-carbon technologies are slow in entering the marketplace, and vice versa during periods of depressed economic activity and low prices for clean fuels. In this way, forcing emissions to meet a fixed cap each year, regardless of the costs of meeting that cap, raises the expected cumulative costs of the policy over time.

Prior experience indicates that allowance prices in cap-and-trade systems can be quite volatile—two examples bear this out. During California’s energy supply crisis 10 years ago, the price of nitrogen oxides allowances under the Regional Clean Air Incentives Market program varied between $400 and almost $100,000 per ton. And in the European Trading System, allowance prices have varied between $0 and $37 per ton of carbon dioxide.

We find that a traditional cap-and-trade program of the stringency envisioned in recent U.S. climate bills might raise the expected annual costs of the program by around 15 percent a year, compared to the type of price collar (described below) that could achieve the same expected emissions reductions over an extended period of time. More generally, price volatility might also be an impediment to investment in research to develop clean energy technologies.

What can be done to reduce the potential for price volatility that seems inherent in cap-and-trade systems?

Banking and Borrowing

Harrison Fell is a fellow at Resources for the Future.    Richard Morgenstern is a senior fellow at Resources for the Future.

One option is to allow firms to bank emissions allowances during periods when the allowance price is low and borrow allowances, or run down previously banked allowances, when the price is high. Such provisions do not entirely eliminate volatility, however, not least because proposals typically embody penalties for, or limits on, allowance borrowing in order to limit default risk. Our research suggests that the limited banking and borrowing provisions in recent bills might reduce expected costs by around 10 percent.

Another option is a safety valve, where the government steps in to supply additional allowances to the market if the allowance price hits a ceiling or trigger level. The safety valve is a hybrid, in that it gives cap and trade, which fixes the quantity of emissions, a feature of a carbon tax, which fixes the price for emissions. As critics have noted, unless the cap is lowered in future years to recoup any additional allowance sales, total cumulative emissions could be higher with a safety valve.

A price collar, another quantity-price hybrid, is intended to restrain price swings by creating a price floor as well as a price ceiling. As with the safety valve, the price ceiling is achieved by providing additional allowances at a predetermined price. The price floor could, most likely, be implemented in one of two ways. If emissions allowances are auctioned, the regulator could set a minimum reserve price, which would serve as the floor. Alternatively, the regulator could promise to always buy and retire allowances at a predetermined floor price.

Both of these mechanisms, particularly a price collar with a relatively narrow band, could significantly limit the potential for additional price volatility. A price collar may also encourage greater investment in new technologies by allaying concerns about a price collapse.

Price Volatility and Emissions Offset Provisions

Another cost-containment feature of many proposed climate bills is to allow firms to offset their obligations by paying for emissions reduction projects in other countries or other sectors of the domestic economy (such as forestry) not formally covered by the program. Although offsets ostensibly help firms meet emissions reduction goals, huge uncertainty surrounds both the quantity of offsets that will be verifiable at any given time period, as well as the costs of such offsets.

To examine these issues, we set up a model that includes uncertainty in future abatement costs, due to uncertainty in future baseline emissions levels, and uncertainty in the future availability of offsets. And, we allowed these sources of uncertainty to be correlated. We focused on a quantity-based regulation with banking and limited borrowing, with offset limitations similar to the climate legislation passed by the House last year (H.R. 2454), and looked at how such a system would function with the introduction of a price collar.

We found that as the sources of uncertainty  become more negatively correlated (that is, the higher the abatement costs are associated with a more limited supply of offsets), the expected total cost of compliance (calculated as the net present value of abatement costs plus the offset purchase costs) increases. Additionally, we found that the range of possible costs, cumulative emissions, and price paths could be quite large in situations where this negative correlation is significant. With respect to the price paths, for the cases without price limitations, we found potential for substantial price variability, especially in the early periods. This result would appear to be particularly worrisome for policymakers and market participants, as great price variability early on will surely open the system to considerable criticism.

With the introduction of a price collar, the expected total cost of compliance was lower and the range of cost outcomes was narrower, yet expected emissions increased only slightly. Importantly, we also found that the range of possible cumulative emissions outcomes could actually be smaller with a price collar compared to a no-collar policy if both the offset supply shock was highly persistent and the negative correlation between offset uncertainty and abatement cost uncertainty was large.  As expected, the range of possible price paths was smaller than with no-collar policies.

The fact that price collars can reduce the variance in emissions outcomes contradicts previous analyses that did not consider offsets. The explanation is straightforward. When offset supply is low, emissions targets must be met with additional abatement. With price collars, the additional abatement can be such that it will trigger the price ceiling and release more allowances into the system. Because a price collar will release more allowances into the system when the supply of offsets is sufficiently limited, the cumulative emissions results for these limited offset cases will be larger with collars than without. Alternatively, when offsets are plentiful, a larger share of compliance will be met with offsets and reliance on the price ceiling will be diminished, so the upper bound of cumulative emissions outcomes will be nearly equivalent for cases with or without collars. These two cases combined lead to a narrower band of cumulative emissions outcomes with a collar than without.

At the end of the day, price collars and offsets should not necessarily be considered cost containment policy substitutes, but rather as potentially effective complements. Indeed, the case for a price collar is even stronger with a heavy reliance on offsets than without it.

Harrison Fell is a fellow at Resources for the Future. His research interests lie in quantitative analysis of marine resource issues, particularly those related to fishery rationalization, industrial organization, the economics of property rights, and game theory.

Richard Morgenstern is a senior fellow at Resources for the Future. His research focuses on the economic analysis of environmental issues with an emphasis on the costs, benefits, evaluation, and design of environmental policies, especially economic incentive measures.

​