This is the fourth post in five-part series on the details required to get carbon policy right. See also parts one, two, and three.

We now get into an issue that will seem a bit arcane, because no one’s talking about it, at least not explicitly. But it’s a real choice, and in many conversations about carbon policy we are implicitly getting it wrong.

Should we price carbon in spots, or strips? Or, to take it out of financial jargon, should we:

  1. set up markets such that people who are selling or buying emissions credits have to go to the market with each incremental ton to determine what the price will be (a “spot” market), or
  2. set up markets such that buyers and sellers can enter into long-term contracts for the emissions they will produce/reduce (a “strip” market)?

Before talking about carbon, we need to take a brief foray into electricity deregulation, specifically back to the early 1990s, when academics were developing rules for how a deregulated market would work, based on broader theories of how markets work. As a reminder, the central test of a good carbon policy is whether or not it encourages investment in carbon-reducing technologies. Replace the word “carbon” with “electricity” and you’ve got a good test for whether the attempt at electric deregulation worked. It didn’t.

In those states which elected to deregulate their wholesale power markets, the driving theory was that if you provide a spot price for power, and allow any buyer or seller to trade at that spot price, you will liberate the power of markets.

While there are many complicated details to these transactions, that is basically still the way all these markets work, whether at PJM, ISO-New England, NYISO, or any number of foreign markets. There is a fair amount of regular trading volume on these markets today, and so it may sound odd to hear someone say they don’t work. But they don’t, for the simple reason that it’s really hard to get comfortable spending billions of dollars of capital based on a gamble about where future prices are going to go.*

A thought experiment may be helpful. Imagine that you have a billion dollars — about enough to build a modestly sized coal plant (or, if you prefer, around 400 MW worth of wind turbines). You have the good fortune to be planning your project in a deregulated state, so you don’t need to get a bunch of permission slips from the local utility to build your plant. You just need to get permits, connect to the transmission system, and build. Now, for the sake of argument, we will stipulate that the average price on those power markets for the last year has been sufficiently high that — if your plant was operating last year, and selling into those markets — you would have earned $200 million/year in net revenue. That sounds pretty good, right? Five year payback, or an 18 percent return on your investment over the next 15 years. Would you build the plant?

This pretty quickly raises an obvious question: what’s the price on the market going to be next year? The answer: You have no idea. Sure, you can predict. You can hire consultants to make forward price forecasts. But when all is said and done, if you build this plant, you are placing a bet with risks you cannot control. What if the economy slumps, demand for power falls, and the spot price collapses? You lose. What if spot prices go up? You win. Which is more likely? You don’t know.

This is not to say that no one will build the plant. Some people have a higher tolerance for risk and will build anyway, in the belief that they have some underlying knowledge of where power markets are going. Some of them will be right. Some of them will be terribly wrong (witness Calpine’s stunning $16 billion bankruptcy which, at core, was the result of a bad bet on the way that gas and electric prices were going to move on those spot markets).

But here’s the rub: we didn’t have to design the market that way. Indeed, there is no real market that works that way. (And by “real,” I mean a market that traces its origins back to one dude who had something another dude wanted and negotiated a price, as opposed to ones whose rules were crafted by academics and regulators.) If you live in New England and buy fuel oil to heat your house, you have a whole choice of payment options. You can pay each month based on the price of oil. You can prepay and lock in your price for the year. You can pay a little more on delivery in exchange for a predictable fixed contract. And so on. Ditto for any number of other volatile commodities that we all buy on a regular basis.

And, ironically, the old, regulated utilities never build plants on spot prices. When they build a plant, they hold a rate case and then lock in their price. (They will also then buy long-term “strips” on their fuel contracts to ensure that they don’t get pinched between future fuel and electric price volatility.)

Note that this is not to say that strips are always better than spots — simply that if you want people to invest capital, you need to provide the option to sell on a long-term strip, short-term spot, and any crazy hybrid of the two. And those options naturally emerge whenever governments simply allow buyers and sellers to meet up, negotiate deals, and get out of the way. Risk-averse buyers will naturally gravitate towards risk-tolerant sellers and vice versa, ultimately creating a mélange of spots, strips, futures, swaps, hedges, derivatives, and all those exotic-sounding increments that are the hallmark of a functioning market. (For those not familiar with the terms, don’t worry — they can all be thought of as different ways to bundle risks, with higher risk “flavors” offering potentially greater — but more volatile — returns and lower risk options offering stabler — but generally lower — returns.)

Relevance to carbon markets

You can probably see where this is going. How are we going to price carbon?

As a tax? Notwithstanding my prior post, this approach isn’t even as good as spot, since it is a price set from a regulator on high, stipulating a fixed price with no market correction to capture the vagaries of supply and demand.

As an auction, with periodic re-auctions to reset the price, as Gar suggested? That gives us something like a spot but has all the problems that we’ve seen in the last decade in the electric sector.

As a government-established trading floor, modeled on the electricity sector? That simply repeats the problems we’ve already faced in the electric sector.

Here’s the salient point. As of today, the only way you can buy or sell carbon (in the U.S.) in voluntary markets is on a spot basis. And the majority of the trading structures that are being considered in Congress are implicitly spot markets. And spot markets will have little impact on the decision-making process of those who want to invest money in projects to reduce greenhouse-gas emissions.

To the extent that there is a simple solution, it is this: trust markets. Let buyers and sellers meet up and trade in whatever fashion they like, and you’ll get what you need. (With suitable government oversight, of course.) And if regulators cannot trust markets, then at least take advantage of the government’s balance sheet to build in long-term contracts from the get-go. But don’t assume that a spot price alone is sufficient.

*Point of candor: In the clearing prices for power (PJM, ISO-NE, etc.) that were set up as purely spot markets, one can now find other parties — primarily financial traders — who will buy and sell long-dated contracts, finally giving some semblance of “strips” to these markets. However, those transactions remain external to the regulated market, and it bears noting that it has taken about a decade for these players to arrive.