Fossil Fuels

In a post earlier this month, I showed that by some measures crude oil price volatility is nearing a low ebb historically, though looking at historical volatility in isolation can mask the magnitude of recent years’ price changes in absolute terms and relative to the size of the broader economy. Bob McNally, president of Rapidan Group, suggested a follow-up post modifying the calculus slightly. What would the graph look like if volatility times spot price were divided by U.S. GDP? Framing it that way might help provide a very rough sense of the scale of the movement relative to the size of the domestic economy over time. See the result in Figure 1 in log scale below. Source: Bloomberg; U.S. Bureau of Economic Analysis One point to keep in mind when thinking through the macroeconomic implications of  oil price movements is that consumers, and hence government officials, experience these fluctuations much differently than volatility traders, for instance, do. Yes, implied and historical volatility may be historically low, but normal people live in a world where short-term swings in nominal prices have outsized importance. That said, the degree to which U.S. consumers are coping with retail fuel prices at today’s levels (including by driving less) would have defied the predictions of many economic models a few years ago.

In an exchange about the Keystone XL pipeline earlier today, NASA’s Gavin Schmidt made an important point: “Many things can raise the effective carbon price: tax, cap-and-trade, regulatory action (mercury standards, pipeline decisions etc)”. (I’ve taken the liberty to expand some twitter abbreviations.) That’s true. So what carbon price would blocking the Keystone XL pipeline be equivalent to? It’s a trickier question than one might think; there are actually no clean analogies. But here are two useful ways to think the issue through. (And a preemptive note: I understand that one can’t boil the Keystone fight down to a little cost-benefit calculation. I’m just picking up on an issue that others have raised.) I’d be eager to hear from others who have issues with these approaches, or who have other ideas -- the broader challenge of thinking through "carbon price equivalents" for regulatory and other non-price measures is an important one for domestic and international climate policy alike. What level of carbon price would be required to have the same impact on pipeline development as simply rejecting the thing? It’s easiest to look at this by first focusing on a concrete price.  Take $100/tCO2, which is many times in excess of the shadow price that the U.S. government applies in its regulatory decisions. Producing and refining a barrel of oil sands crude entails about 100 kilograms of carbon dioxide emissions. That translates to $10 a barrel in extra charges. At the same time, applying the same carbon charge to U.S. oil use would add about $40 to the cost of using a barrel of oil. Demand elasticity for U.S. oil is low. Increasing the price of U.S. oil by this increment (let’s assume a base price of $120) would eventually lower U.S. demand by perhaps 15 percent (again I’m being generous). U.S. oil is about a fifth of the global market, so that would lower total global oil demand by about 3 percent. The IMF estimates that elasticity of world oil demand is around -0.072 in the long run; the net result should be about a 34 percent decline in the oil price. (This is a very, very conservative lower bound, since OPEC members would cut back investment, and other countries would pick up some consumption, but let’s run with it.) Assuming that we start $120/bbl oil, that takes us down to $79 oil. (Again, in reality, probably not nearly so far.) Most oil sands producers can make their investments work – even after paying the extra $10 – at this price. How high a carbon cost would be required to negate other positive externalities from oil sands development? Oil production doesn’t just cause damages that aren’t accounted for by market prices – it also brings benefits. Let’s focus on just one: more production pushes down oil prices. That creates benefits for all U.S. oil consumers. What level of carbon damage would be required to fully offset this? Imagine that you boost net world production by one million barrels of oil a day. (This could be one million additional Canadian barrels; it could be five million additional Canadian barrels and four million barrels of offsetting OPEC cuts. The absolute magnitude of the number doesn’t matter – all of what I’m about to do scales linearly. That’s one of the things I particularly like about this line of analysis.) Let’s stick with the same IMF estimates of demand elasticity. The resulting price hit is a bit less than $20 a barrel. Now apply that to the roughly 7 million barrels of oil the United States imports each day. You’re talking a total cost of about $130 million a day. Climate damages need to be around $130 dollars a barrel – approaching $300/tCO2 – just to get even.

The State of the Union address last night was notable for the prominent placement of energy and climate and for its recommencement to what President Obama has called an all of the above strategy. I was particularly struck by the inclusion of two new efforts that would aim to concretely bridge the gap between fossil fuel backers and clean energy enthusiasts: the Energy Security Trust Fund and a new prize for development of natural gas with carbon capture and storage. The Energy Security Trust Fund has been pitched, in different variations, by Securing America’s Energy Future (SAFE) and Senator Lisa Murkowski (R-AK). These proposals have envisioned earmarking revenues from new oil and gas drilling for spending on clean energy innovation. The White House fact sheets on the speech are ambiguous as to whether they see the Trust Fund getting money from existing or new drilling. But the political reality is that this proposal only has legs if it has something for everyone – and that means it needs to mix new oil and gas development with more investment in clean energy innovation. Indeed that is pretty much the point. There is no technical reason that money can’t be taken from general funds to support innovation, and there is no reason that lands can’t be opened for drilling without spending the revenues on clean energy. To me the biggest virtue of this approach is that it starts to tie the fortunes of various combatants together: oil and gas supporters can only cut clean energy funding by blocking drilling; clean energy backers suffer if oil and gas development is curtailed. The proposal to award a $25 million prize to the first developer to implement carbon capture and storage (CCS) on a natural gas combined cycle power plant is in a similar vein. The most profitable way to do CCS on a gas-fired power plant is to inject the carbon dioxide that’s captured in order to enhance production of oil. If the administration can find ways to jump-start this effort, the result would be development and cost-reductions of a critical low-carbon technology together with activity that could, as the Natural Resource Defense Council has pointed out approvingly, simultaneously give a big boost to oil production. My only quibble with the proposal is that it’s too small – I’m not sure the $25 million will do the trick. Why not propose that a slice of the roughly $10 billion (over the next decade) that’s currently slated to go to the percentage depletion tax credit currently enjoyed by oil producers be redirected to support projects that combine oil, gas, and CCS? It would be a win for zero-carbon energy and for many oil and gas producers at the same time. That gets back to what I really like about these two ideas. (What I dislike is that they’re both in the final chapter of my forthcoming book; so much for originality.) Neither of these ideas alone is likely to be directly transformative for American energy – that requires much bigger moves on both oil and gas and on zero-carbon fuels. But both could help build some of the trust between long-warring parties that will be required for large changes to eventually happen. As someone wise reminded me recently, once people agree on one thing, they often end up agreeing on other things too.

Big-name economists have been lining up to show their support for yet another target-based approach to monetary policy making: nominal gross domestic product level (NGDP) targeting. The basic idea is that a central bank should aim to stabilize GDP, unadjusted for inflation, at around 4.5% as a means of stabilizing aggregate demand and avoiding recessions. NGDP targeting having once been the intellectual stomping ground of economists on the right (notably Scott Sumner), its newest supporters come overwhelmingly from the left (such as Christy Romer). After the collapse of Bretton Woods in the 1970s, targeting of the money supply became the monetary Holy Grail. In the 1990s, as money supply targeting became operationally too problematic, the world shifted to the targeting of consumer price inflation. But after 2008, when July U.S. CPI hit 5.6% in the midst of a financial crisis, support for inflation targeting – which had become as close to global monetary orthodoxy as the gold standard had been in the late 19th century – melted away. Credible justification was needed for loosening policy at a time of elevated inflation. A year later, with CPI at -2.1%, such justification was no longer necessary. But those fearing a too-early tightening in policy turned to other targets. Targeting the price level, rather than price inflation, became popular, as it required the Fed to tolerate more inflation in the future to compensate for deflation and under-inflation in the past. The Fed itself has now turned to a temporary unemployment-level target. But NGDP targeting is truly the new intellectual rage. New Bank of England governor Mark Carney is the most prominent advocate in policy-making circles. We think the rage will be short-lived. The reason is that NGDP targeting’s newest supporters are bad-weather fans. That is, they like it now, when NGDP is well below its 2007 “trend” line, meaning that the policy implies extended and more aggressive monetary loosening. But what happens when NGDP goes above its target, as it eventually will? NGDP targeting then requires tightening, even if inflation is low – it may even require a deliberately deflationary policy stance. In this week’s Geo-Graphic, we identify in yellow 11 periods between 1983 and 2003 when the Fed was loosening policy but where a 4.5% NGDP target would have prescribed tightening.* This suggests strongly that NGDP targeting has no legs: when it tells the Fed to tighten, its prominent new supporters will abandon it even more quickly than they embraced it. Indeed, two noted monetary economists have even called pre-emptively for the abandonment of NGDP targeting once it’s done its job of justifying looser policy today. “Once the nominal GDP growth shortfall has been eliminated,” Michael Woodford and Frederic Mishkin wrote in the Wall Street Journal on January 6, “ it will be appropriate to again conduct policy much as was done before the crisis.” Yet since the rationale behind both inflation targeting and NGDP targeting is that they anchor public expectations for the long-term, adopting them opportunistically is a particularly bad idea. * We look at the annual rate of NGDP growth, rather than NGDP levels—using levels would suggest continuous tightening throughout the period and many more yellow bars. Sumner: The Money Illusion Romer: Dear Ben: It’s Time for Your Volcker Moment Steil: The Battle of Bretton Woods Mishkin and Woodford: In Defense of the Fed's New Interest-Rate Policy

Dan Ahn and I have a new energy brief out that takes a fresh look at oil taxes. From the introduction: "Policymakers are confronting difficult choices [regarding tax hikes and spending cuts].... In this context, it might be possible to reconsider oil taxes not only as an unwelcome burden, but as an alternative to something worse. We have modeled the potential consequences of substituting taxes on oil consumption for either higher non-oil taxes or reduced government spending, both as part of a larger deficit reduction package. [We show that] doing so can improve economic performance while reducing oil consumption if done right." The paper goes on to quantitatively explore the growth, employment, and oil consumption impacts of different ways of modifying deficit reduction packages using oil taxes. The paper is the first to look at oil taxes in the context of broader deficit packages; it’s also novel in that it looks at how oil taxes might perform in a weak economy. In a Bloomberg View op-ed today, we explain some of the basic results, and provide some simple intuitive explanation for the paper’s conclusions that goes beyond what’s in the paper itself. Take a look.

Okay, so the title of this post is tongue-in-cheek: the U.S. president has far less power to influence gasoline prices than campaign-season banter would lead you to believe. But I figure if a sitting president can take the blame for high and volatile oil prices, maybe the White House should take a little pride in the fact that, by some measures, oil prices reflect the lowest volatility in years. (There’s a bit more to the story, though.) Figure 1 shows the annualized 30-day historical volatility of spot WTI prices since 1984, the first full year after the benchmark launched based on that crude stream. Since the start of 2013, volatility by this metric hasn’t been so low on average since 1995. Were it to drop below 15 percent and stay there consistently, it would be the first time since 1992. Figure 1. Annualized 30-day historical volatility of spot WTI prices (1984 – present) Source: Bloomberg Note implied volatility as well. The so-called “Oil VIX,” a barometer of market expectations of 30-day volatility of crude prices, dropped to 21.67 percent on January 22, the lowest reading since the index began tracking the market in 2007. The index is an oil-specific version of the VIX, a measure of equity market volatility, derived from volatility skew from a range of option strike prices on the United States Oil Fund. Figure 2. CBOE Oil ETF VIX Index (2007 - present) Source: Bloomberg Looking at historical volatility only in percentage terms, though, can mask the magnitude of the changes in absolute prices and thereby gloss over the potentially harmful macroeconomic effects of large price swings. If you were to look only at Figure 1, you couldn’t tell that the oscillation in flat price has been of an order of magnitude higher between 2010 – 2012 than it was from 1996 - 1998. These absolute changes are not always quantitatively discernible when viewed in relative terms, but they matter to net effects on the macroeconomy. Figure 3. WTI spot prices (1983 – present) Source: Bloomberg One way to see how noisy today’s oil prices are in absolute terms is to multiply prices by volatility. Figure 4 shows WTI spot prices since 1984 times the same rolling 30-day volatility depicted in Figure 1. The result is striking: the market’s gotten a whole lot rowdier when it comes to the magnitude of the price swings in dollar terms, even if volatility by other measures has fallen. Figure 4. Annualized 30-day historical volatility of spot WTI prices * spot WTI prices (1984 - present) Source: Bloomberg

A piece in Slate by Ray Pierrehumbert arguing that tight oil abundance is a myth is making the rounds. The essay makes some fair warnings against irrational exuberance when it comes to hundred year supplies, claims of endless energy independence, and complacency on climate change as a result of abundant natural gas. But the piece does at least as much to confuse as illuminate. Fortunately, that provides a good opportunity to look at a few important misunderstandings that frequently arise in discussions about U.S. oil. The Slate essay starts with an attack on a paper published last year by Leonardo Maugeri that had a distinctly cornucopian bent. Many, myself included, have argued that that paper was way over the top. But Pierrehumbert makes a big mistake when he claims that all of the other excitement – from the EIA, IEA, and others – have flowed from bandwagoning on the Maugeri report. Those of us who spend decent parts of our professional lives involved in this stuff know quite well that this isn’t what happened. For starters, the first prominent and enthusiastic projections weren’t from Maugeri; they were from Ed Morse at Citigroup. The EIA and IEA reports used bottom-up analyses that were independent of the Maugeri work. If people want to pick apart these studies, that’s fine, but cutting down one largely unrelated outlier won’t do the trick. The next big problem with the Slate essay – again one that many others make too – is that it appears to assume that tight oil will need to deliver all U.S. oil production. That allows it to claim things like this: “At the high end of the estimates, predicted production from Bakken and Eagle Ford together amounts to perhaps a two-year oil supply for the United States at 2011 consumption rates.... Even if it were to prove possible to achieve production rates comparable to those of Saudi Arabia, that would only mean that we would deplete the resource faster and bring on an oil crash sooner.” On top of this, while Pierrehumbert is right that some people are ignoring the fact that current tight oil prospects will peak and then decline, he errs in presenting this as a critique of mainstream estimates, like those by the IEA and EIA, despite the fact that those projections show precisely that same decline. The essay then launches in an oft-heard discussion about high decline rates and large capital costs. Geologists’ focus on this as an argument for why production will be low continues to baffle me. Do people think that the models used by government agencies and industry forecasters don’t incorporate this? Of course they do. They just find that, even when they include this, economic incentives still push things toward higher production, at least through the end of the decade. There is no law of nature that says it’s impossible to produce a lot of oil from a field whose wells are expensive and decline quickly. The Slate essay also manages to bring in one of my favorite bugaboos: energy return on investment (EROI). It is taking ever more energy, Pierrehumbert points out, to produce a barrel of oil. This is supposed to herald the disastrous coming of a day when we need to put more energy in than we get out. But not all energy is the same, and it can make very good sense to put in large amounts of energy in a relatively low-value form (e.g. gas) to get a smaller amount of high-value energy (e.g. oil) out. Once again, geology and physics are important, but economics need to be factored in. One last point: the Slate essay repeats the misleading juxtaposition of the amount oil in a massive resource (this time the Green River shale formation) with plausible emissions limits in a carbon-constrained world, in order to warn about the climate consequences of extracting the new fuels. But this suffers from the same problem that the “game over” claims for the tar sands have: it pays no attention to time scales. There is no plausible scenario in which we’ll spend the next thousand years with a totally decarbonized economy – except that we’ll burn everything in Colorado or Alberta or some other discrete carbon pool. The causal arrow runs the other way: these big pools of oil will be burned if we choose to cook ourselves; they will mostly remain in the ground if we don’t. It’s how much fossil fuels we use, not where they come from, that matters most to the planet.

In my last post I discussed how trading volumes show a migration into ICE Brent from NYMEX West Texas Intermediate (WTI), two of the world’s most watched crude oil benchmarks. The trend is part of Brent’s broader rise as the preeminent world price of oil. Here I’d like to show graphically part of the reason why some financial market participants are opting to trade the North Sea crude instead of its American cousin. It has to do with recent trends in the futures prices for both crudes. Figure 1 shows the forward curves—i.e., a series of sequential prices for future delivery— for NYMEX WTI and ICE Brent crude oil. Prices for Brent, in red, stretch out to 2019, whereas WTI, in green, go through 2021. I noted in my last post the remarkable divergence between front-month contracts for Brent and WTI, a significant shift from the historical norm of WTI’s slight premium. Note in Figure 1 the difference in the shape of the two curves for near-dated contracts. WTI contracts through August 2013 were in a state of contango, with prices rising into the future. For backwardated Brent, on the other hand, prices are progressively declining over the entire curve. Figure 1. Forward curves for NYMEX WTI and ICE Brent as of Jan. 24 Source: Bloomberg This difference in structure may seem arcane, but it has proven an important driver of the shift among some financial market participants into Brent contracts at the expense of WTI. Commodity index investors, for example, are interested in reaping roll yield, or the normalized difference between a nearby and a deferred futures contract, since they roll their exposure as futures contracts approach expiration. This type of yield is one of the primary sources of long-term returns for commodity index investors. As Figure 1 shows, though, the Brent curve offers a positive roll yield for contracts at the front of the curve—in contrast to WTI, where that yield would be negative, all else equal. That differential, which tilts the balances toward the North Sea benchmark insofar as the roll-related return potential, has been one factor pushing trading volumes in its direction. Figure 2 shows this differential in another way. It depicts the price spread between futures contracts for the two crudes one and six months in the future since 2000. The close correlation between the spreads for the two markers were tightly positive up until around 2010, when the glut of light-sweet crude in the U.S. midcontinent sent WTI into a prolonged state of contango, whereas Brent’s structure reflects relative physical tightness in the underlying North Sea streams. Figure 2. 1 - 6 month price spread for NYMEX WTI and ICE Brent (2000-2013) Source: Bloomberg Because the market expects Cushing’s supply woes to begin to be amerliorated later this year, the result of additional pipeline capacity draining excess crude inventories, WTI’s contangoed structure extends out only through September as of yesterday’s market close. After that delivery date, as Figure 1 suggests, the slopes the two curves come closer into alignment, minimizing Brent’s roll yield advantage over the American benchmark. As Figure 3 makes plain, though, near-dated contracts for the crudes are traded much more heavily than those further out along the forward curve. The differential in the term structure of the two crudes, though by market expectations impermanent, appears sufficient over the most heavily traded contract months to have helped incline some commodity traders toward Brent over WTI. Figure 3. NYMEX WTI 20-day average trading volume over the length of the curve as of Jan. 24 Source: Bloomberg

The failure of ailing Venezuelan president Hugo Chavez to return from Cuba, where he is recovering from another round of surgery, to Caracas for his inauguration underscores the uncertainty of the South American country’s future as a critical oil supplier. Chavez, first elected in 1998 and inaugurated in 1999, rode ultra-low oil prices to power, promising a tougher stance against the majors and a more hawkish voice within OPEC. So how’s the country’s oil industry faring today versus when he entered office? Venezuela was the third-largest producer in OPEC when Chavez took office, its roughly 3.5 million barrels per day (mb/d) surpassed only by Saudi Arabia and Iran (see Figure 1). Output was on the upswing, +1 mb/d since the start of the decade. But the country’s production has trended steadily downward under Chavez—now 30 percent lower than it was in 1998—falling victim to the mismanagement of PDVSA (Venezuela’s national oil company) to finance other state projects, hostility toward foreign investment, and a mature production base where decline rates at existing fields are as high as 25 percent, according to the U.S. Energy Information Administration (EIA). Figure 1. Oil Production Among Select OPEC Suppliers (1998-3Q2012) Source: EIA. Includes crude, NGLs, and other liquids. The woeful production record under Chavez isn’t for lack of oil in the ground. Venezuela sits on more proved reserves, according to BP’s estimate, than any other, at 297 billion barrels (Figure 2). Saudi Arabia comes closest, at 265 billion barrels, though the kingdom is also less heavily explored and has a generally higher-quality resource base. Of Venezuela’s proved reserves, most (some 220 billion barrels, per BP) are extra-heavy crude and bitumen in the Orinoco Belt, but industry estimates suggest that even this low-quality oil can be produced at as low as one-third the cost of its Canadian cousin, due to favorable fluid and reservoir conditions that make for better per-well production rates. Figure 2. Proved Oil Reserves by Country Source: BP Statistical Review of World Energy 2012 While production has fallen under Chavez, consumption has risen (Figure 3)—up from about 490 thousand barrels per day (kb/d) in 1998 to 850 kb/d today—biting into net exports, which poses a problem for the country’s future fiscal health. Crude exports have collapsed since Chavez took power, down nearly 40 percent to roughly 1.5 mb/d (Figure 4). Refined product export patterns are looking increasingly shaky as well. Last September saw a sharp jump in U.S. gasoline and other refined product exports to the South American country, some 196 kb/d, and some industry sources estimate a reliance on net product imports as high as 300 kb/d. The proximate causes of the September jump were accidents at the Amuay and El Palito refineries, which knocked out a substantial portion of the country’s refining capacity. But the more troublesome underlying factor is the slow deterioration of the country’s refining complex and oil-specific technical prowess, causing a string of outages and unplanned stoppages in recent years. Figure 3. Venezuelan Oil Consumption by Major Product Category Figure 4. Venezuela Crude Oil and Natural Gas Liquids Exports (1998-2011) Source: IEA Part of what underpins the climb in Venezuela’s consumption is deeply subsidized oil; drivers there enjoy the cheapest gasoline in the world (Figure 5). Figure 5. Price per Gallon of Premium Gasoline Ranked by Country Price per gallon of premium gasoline, Aug. 12 Country Rank, most expensive (out of 60) $10.12 Norway 1 $9.41 Turkey 2 $9.28 Israel 3 $8.61 Hong Kong 4 $8.20 Denmark 6 $7.87 United Kingdom 10 $3.75 United States 49 $1.89 United Arab Emirates 56 $1.73 Egypt 57 $0.89 Kuwait 58 $0.61 Saudi Arabia 59 $0.09 Venezuela 60 Source: Bloomberg News, Aug. 13, 2012 Whoever takes over for Chavez will face unenviable challenges: righting an industry that accounts for 95 percent of the country’s export earnings and 40 percent of government revenue after years of mismanagement; restoring the once-venerable PDVSA, whose debt as a percentage of GDP rose from 7 percent to 11 percent between 2000 and 2012; taming inflation running at 26 percent; and dealing with a host of other economic woes. It’s a full inbox.

Constant chatter about an impending oil release from the U.S. Strategic Petroleum Reserve (SPR) was a prominent feature of the oil market last year. Much of the speculation was driven by the ongoing loss of crude from Iran, due to sanctions, and the possibility of a confrontation with Tehran over its nuclear program, which could have cut off traffic through the vital Strait of Hormuz. The market’s SPR talk has died down, but Washington is likely to face some important questions in the near future about the country’s emergency oil reserves, driven by evolving domestic supply-demand conditions. Here are five: -- Size Some analysts are calling for the SPR to be downsized in light of the country’s declining net import levels, thanks to growing domestic production and declining consumption. The 695 million barrels of crude oil in the SPR are currently around 80 days’ worth of net imports (at 2012 net petroleum imports of 8.72 mb/d). If current trends hold, net imports could fall to roughly 6 mb/d within 3-4 years. That would mean that trimming down the SPR to the International Energy Agency (IEA)-mandated 90-days of net imports (using government stocks alone) could free up around 155 million barrels of SPR oil, which could be sold on the open market to generate substantial public revenue. -- Location It may make sense to consider adding or transferring emergency inventories to the East and West coasts, rather than having them confined to the U.S. Gulf Coast. Locating the SPR in the Gulf was a natural decision when the SPR was first created. But the turnaround in light sweet crude production in the greater Gulf region and the Upper Mississippi Valley, where SPR oil was designed to be shipped via pipeline, point to a declining need for that grade in that part of the country. In the case of a disruption in imports, the East and West coasts may benefit much more from extra oil on hand (and potentially in the form of refined products like gasoline and diesel). -- Composition As it stands, 38 percent of the SPR is made up of sweet crude oil, much medium API in gravity. But with imports of light sweet crude into the Gulf Coast in what may be a terminal decline, displaced by indigenously-produced tight oil, and ongoing pipeline reconfigurations, the region’s crude mix is quickly changing. It may be worth considering the types of crude held in the SPR to better reflect refiners’ needs in that part of the country, which is home to roughly half of U.S. operable refining capacity. Moreover, adding refined products may be worthwhile. Hurricanes Katrina and Rita in 2005 crippled refineries in the Gulf (Katrina alone shut in 8 percent of total U.S. refinery output), and crude isn’t any good for drivers if it can’t be turned into gasoline when they need it. -- Criteria for release Only broad criteria govern when the president can release oil from the SPR. When the White House should pull the trigger is a highly subjective decision as a result, which has caused (usually partisan) bickering in the past. Presidents Clinton and Obama both took heat for their SPR decisions, justified or not. The 2011 IEA-coordinated release seemed to up the intensity of calls for greater transparency and predictability in the agency’s decisions, which some energy experts support. Criteria could be based on a variety of variables, including the absolute or relative physical supply shortage or price increase. Estabilishing strict criteria would have trade-offs, though. If the market knows that an SPR release will be triggered at X dollars, for instance, traders may try to test Washington’s commitment, leading to suboptimal auto-releases. -- Inclusion of newer oil heavyweights China is now one of the world’s largest strategic stockpilers of oil, not to mention the second largest consumer, and is actively growing its emergency crude holdings. Yet it remains outside of the IEA. That leaves open the risk that, in a future release, Washington could sell oil in the open market only to be absorbed into Chinese public stocks. In that scenario, the net effect on global oil supply would still be better than if the United States hadn’t released anything, but not as good as if Beijing had held off on bidding for its own account—or better yet, released its own stocks simultaneously. Finding a reliable way to bring China into future talks about coordinated stockpile drawdowns could benefit Washington and Beijing alike. The IEA and China have discussed impending releases before, as they did prior to the 2011 Libya-related release. But establishing a system for joint action, even if China remains outside IEA membership, could be a worthwhile goal, partly because the story won’t end with China. The growth of India’s strategic stocks, as well as those of other non-IEA countries,  are sure to raise similar questions down the road.

There’s a lot of buzz today about new projections for U.S. oil imports showing that imports are poised to continue diving. The Financial Times captures the essence well with the headline “U.S. oil imports to fall to 25-year low”, referring to projections through 2014. I’ve written before about the risks of focusing on imports rather than consumption. If you want to focus on imports, though, the number to drill down on isn’t the volume of imports – it’s spending on imports as a fraction of GDP. Alas, by that measure, despite a positive trend and strong improvements over the last decade, the United States will remain in worse shape next year than in any year between 1983 and 2003. Let’s start by taking a quick step back. To the extent that U.S. dependence on imported oil is consequential, that’s either because spending on imports bleeds the U.S. economy, or because volatile import bills hurt the United States. The first problem is measured directly by U.S. spending on imported oil relative to U.S. GDP. The second problem can be measured indirectly by the same figure: everything else being equal, the higher the baseline for U.S. import spending is, the greater the economic impact of a given oil shock will be. The chart below shows U.S. spending on oil imports as a fraction of U.S. GDP. It incorporates the new projections for next two years by assuming that one-quarter of U.S. imports (i.e. imports from Canada) are priced at WTI and the rest are priced at Brent, and assumes 2.2 percent GDP growth (the results aren’t sensitive to this choice). The underlying data comes from the EIA and FRED. The result is clear: imports measured in value relative to the size of the economy aren’t anywhere close to their 25-year lows. (This is because lower import volumes have been substantially offset by higher oil prices.) The result is that projected import spending as a fraction of the economy is higher than import spending was in 1973, the year of the first modern oil crisis. It isn’t far below the figure for 1978, the year before the second crisis hit. And it is double its level in 1988, the year before Saddam Hussein invaded Kuwait and touched off an oil-centered crisis. Reaching the low mark for the past 25-five hears, achieved in 1998 at 0.45 percent of GDP, would require U.S. imports or prices to be slashed by a factor of four from their 2012 levels. There should be no question that the decline in U.S. imports – and, more fundamentally, the production gains and consumption curbs behind it – is good news. But, with oil prices appearing fairly steady at historic highs, it’s important to keep that in perspective. [Note: This post has been updated to include projections for 2014; its qualitative conclusions are unchanged.]

One interesting feature of the U.S. hydrocarbon boom is the widening gap between the industry’s interest in drilling for oil and other liquids versus dry natural gas. It’s all about economics: the disparity in prevailing market prices and outlook between these commodities is dictating companies’ willingness to sink money into, and bear the risk of, trying to produce them. The two graphs below show industry expenditures on exploration and production (E&P) in North America relative to crude oil and natural gas prices, courtesy of the U.S. oil services and drilling equity research team at Barclays Capital. Surging prices for both goods starting around 2003 sparked a boom in spending, which roughly tripled between 2002 and 2012. Figures 1 and 2: North American E&P spending vs. benchmark U.S. crude oil and natural gas prices As the two figures show, the surge in gas production from the investment boom helped swamp benchmark natural gas prices at Henry Hub, Louisiana, reflecting a glutted market. Benchmark West Texas Intermediate (WTI) oil prices, in contrast, quickly rebounded to around triple-digits after their epic collapse in 2008-9, and remain far above long-term inflation-adjusted historical averages (despite having to contend with a glut of their own at the WTI pricing hub of Cushing, Oklahoma). Looking at oil- versus gas-directed rotary rig counts in the United States makes it clear just how much more drilling activity is occurring here right now relative to the 1990s (Figure 3). There are about twice as many rigs deployed today as there were a decade ago. In absolute terms, drillers favored gas between 2002 and 2009, moving hundreds of rigs into production. But with natural gas prices down for the count, and with the rapid resurgence of the price of oil, North American operators are overwhelmingly choosing to channel their investment into the hunt for liquids. Figure 3. Rotary rigs in operation in the United States since 1989 (including oil- vs. gas-directed) The substitution of gas for oil-directed drilling activity has been a defining reversal in U.S. hydrocarbon production over the last three years. Whereas just a few years ago nearly 90 percent of rigs were looking for dry gas, that figure’s plummeted to 24 percent—and oil-directed rigs now make up three-quarters of the total. Figure 4. Percentage of oil- versus gas-directed rotary rigs in operation in the United States This picture could change if gas prices were to move high enough to justify companies increasing their budgets for gas drilling again, which could set off a scramble for those contractors able to quickly get rigs back into gas plays. But for the time being, it’s oil, not dry gas, that operators are interested in, not surprisingly, given the price differential. Either way, the aggregate picture is of a country where drilling is at full tilt. Even Hollywood wants in on the action.

I have a new paper (PDF) in Climatic Change that explores the climate consequences of natural gas as a bridge fuel. [Update: The article is now behind a paywall. If you don’t have access, you can download an unformatted pre-print version here.] Here’s the abstract (followed by a discussion): Many have recently speculated that natural gas might become a “bridge fuel”, smoothing a transition of the global energy system from fossil fuels to zero carbon energy by temporarily offsetting the decline in coal use. Others have contended that such a bridge is incompatible with oft-discussed climate objectives and that methane leakage from natural gas system may eliminate any advantage that natural gas has over coal. Yet global climate stabilization scenarios where natural gas provides a substantial bridge are generally absent from the literature, making study of gas as a bridge fuel difficult. Here we construct a family of such scenarios and study some of their properties. In the context of the most ambitious stabilization objectives (450 ppm CO2), and absent carbon capture and sequestration, a natural gas bridge is of limited direct emissions-reducing value, since that bridge must be short. Natural gas can, however, play a more important role in the context of more modest but still stringent objectives (550 ppm CO2), which are compatible with longer natural gas bridges. Further, contrary to recent claims, methane leakage from natural gas operations is unlikely to strongly undermine the climate benefits of substituting gas for coal in the context of bridge fuel scenarios. I’m not going to go through the details of the paper, but I want to discuss some of the physical intuition that underlies it, and add some explicit comparisons with a couple other papers that have garnered a lot of attention (and that motivated this work). The underlying explanation for the results on methane is intuitively straightforward. When one models mitigation scenarios, peak temperatures are typically realized many decades after greenhouse gas emissions (and intensive natural gas use) have fallen deeply. That’s because the climate system has a lot of inertia. This means that it’s the long-term impact of methane -- known to be much smaller than its short-term impact -- that really influences peak temperatures. That weakens the ultimate impact of methane. In particular, gas is never worse than coal for peak temperatures, even with 5 percent leakage, regardless of the choice of emissions target. I explore a wide range of scenario pairs that differ only in their relative use of coal and gas. In every pair, peak temperatures are higher in the cases that feature coal than in those that feature gas. This is a consequence of the phenomenon that I just mentioned: because peak temperatures lag the decline of conventional fossil fuel combustion by several decades, the effect of methane leakage largely dies out (loosely speaking) before it can influence peak temperatures much. All of this is compounded by the fact that, if one wants to keep to an aggressive emissions target, a natural gas bridge can’t last long. A short bridge means relatively little in the way of methane leakage, and a relatively small impact on peak temperatures as a result. This corollary of this result, though, is that using gas as a bridge instead of keeping coal around a bit longer (assuming the same path for zero-carbon energy in both cases) doesn’t make much of a difference to carbon dioxide emissions if you’re trying to stabilize concentrations near 450 ppm. The bridge is simply too short for the distinction to be large. Some of these results change a bit when you’re looking at scenarios that stabilize carbon dioxide concentrations around 550 parts per million. Extreme methane leakage can now be more consequential for peak temperatures, because the natural gas bridge is longer, allowing for more methane to be emitted. (Lower leakage rates of 1-2 percent, consistent with mainstream estimates, are still of only minor consequence.) At least as important is that substituting gas for coal in the context of such targets can be far more consequential (because fossil fuels without CCS can stick around longer). The upshot is that, even with an aspiration to keep carbon dioxide concentrations below 450 parts per million, transitioning from coal to gas may be valuable as hedge in case an ultimate transition to zero-carbon energy occurs late. Comparisons with Howarth et al. and Wigley These results differ from those in two papers on natural gas and methane that have garnered particularly widespread attention for their alarming results. Robert Howarth and colleagues combined high estimates of methane leakage with a focus on 20-year warming potentials to conclude that natural gas is worse for climate change than coal. The new Climatic Change paper shows that the 20-year horizon is completely inappropriate for discerning the impact of methane leaks on peak temperatures. Tom Wigley raised a similar concern about Howarth et al. in a paper published in 2011. (He kindly helped me replicate the results in his paper.) To avoid Howarth’s reliance on global warming potentials, he constructed a scenario in which natural gas use rises strongly through 2100 and then declines through 2200, ultimately ending at approximately present levels. He then estimated the impact of methane emissions on temperature profiles over the course of his scenario, rather than on a particular time horizon, finding that methane negated any warming benefits for many decades. But there is an important limitation to that paper: natural gas use is never phased out in its scenarios. (They are not stabilization scenarios.) That makes it impossible for that paper to discern the impact of methane leakage on peak temperatures. (Temperatures never peak in the paper’s scenarios.) My new paper was originally motivated by a desire to address this issue. The result should cool down some of the alarm that the earlier paper generated. Limits and Directions for Future Work My new paper looks strictly at the climate consequences of bridge fuel scenarios. It does not dive into two other critical questions: Are such scenarios technologically, economically, or politically plausible? And what are their economic, security, and environmental costs and benefits? Both questions are massive and are essential to address. The paper says nothing about whether pushing into natural gas in the short run would make it more or less likely for the world to make a timely transition to zero-carbon energy after that; in-depth study of the plausibility of different pathways is essential to addressing that. Moreover, peak temperatures are only one criterion by which scenarios should be judged. Comprehensive assessments need to take issues like economic cost and local environmental consequences into account. I can’t stress this strongly enough: My paper does not say that any particular pathway is "better" or "worse" or "preferable". It explores some important properties of theoretical paths that have been widely discussed but poorly investigated. In doing that, it shows that recent studies have tended to overestimate the importance of methane, but that, at the same time, some commentators have given too much credit to the potential value of natural gas as a bridge fuel for achieving stringent climate goals. Taking things to the next level, and understanding how a near-term shift to gas might affect long-term trends and outcomes, will require considerably more in-depth work on how gas fits into economic and political systems.