Review of Crude Volatility: The History and the Future of Boom-Bust Oil Prices by Robert McNally

It has been almost three years since we recognized the post-pandemic buying opportunity in the energy sector. During that short period of time, WTI crude oil has traded as low as $36/bbl (October 2020) and as high as $124/bbl (March 2022). Over the past twenty years, crude oil has traded as low as $10/bbl (March 2020) and as high as $145/bbl (March 2008). But if you look at a chart of the oil price from the mid-19th century until today, you see that the cycles of volatility are episodic, and there was a particularly long period of stability after World War II.

Why so volatile today, and why not at other times in history? This is the question that Robert McNally attempts to answer in Crude Volatility. (This was published in 2017, so it does not even include the volatile post-pandemic period.) He identifies six eras of price stability and volatility since the Civil War: "Boom-bust I," with volatile prices, the "Rockefeller era" with more stable prices, "Boom-bust II," after the breakup of Standard Oil, "Texas era," with the lowest volatility (thanks to "the most heavy-handed, government-imposed quota regime the world has ever seen"), the "OPEC era," and finally, our current "Boom-bust III" era.

McNally says that "the most important feature of today's oil market is the absence of a swing producer able and willing to adjust supply to keep oil prices stable." His view is that the eras with a very large share of production under the control of a single producer or cartel that was willing to adjust production made for the stable eras. He thinks that stable prices are a good thing in and of themselves, because "sustained oil price volatility reduces planning horizons, deters investment in machinery and equipment, and increases unemployment."

The key feature of the oil market is that both supply and demand are rather inelastic in the short run, unless there is a "swing producer" that is willing to significantly increase or decrease its own production. Absent such a swing producer, it can require very large changes in the price of oil to balance the market. There is some reason to believe that the demand elasticity of oil has fallen over time since it is used much less for generating electricity in the past. On the other hand, the shale producers have increased the supply elasticity to some degree since capital expenditures on shale wells can quickly produce significant quantities of oil.

As we have been discussing recently, there is no better example of capital expenditure cycles than the oil and gas industry. McNally mentions "widespread fear of peak oil ten years earlier had resulted in enormous investments in new fields and a buildup in inventories." What time period do you think he is referring to - 2015? No, he's talking about 1928.

By the way, since the book was written in 2017 it predates the Biden sales of crude oil from the Strategic Petroleum Reserve. The SPR peaked in October 2010 at 727 million barrels. Obama sold 32 million barrels over the remainder of his presidency, leaving it at 695 million in January 2017. Trump sold 57 million barrels during his term, leaving Biden with 638 million. And now Biden has sold 291 million barrels in a two-and-a-half years, a rate of about 300k bbl/d. 

While Obama and Trump's sales were relatively modest, Biden's sales during the summer of 2022 were at a rate of 1 million barrels a day, a rate that could fairly fit under McNally's concept of swing production. The inelasticity of demand means that Biden's sales, which amounted to 1.25% of daily world production at peak rates, could have had a very significant impact on the oil price. As an example, McNally points out that in 1979 the loss of Iran's 2 million barrels per day - which was about three percent of world supply - led to the price of oil more than doubling.

And it is taking not only these sales but also Federal Reserve balance sheet reductions ("tightening," which was initiated at the same time as the oil sales) to keep oil under $80 per barrel. The SPR oil sales are obviously not sustainable in perpetuity - there is less than a year's supply left at a pace of 1 million barrels per day. Nor is tightening likely to be sustainable indefinitely, as we have written in the past. As we have seen, tight money does not really help an over-leveraged country delever since it causes debt/GDP to rise and also threatens the banking system.

It is worth thinking about the longer run supply and demand factors for crude oil. On the bullish side, we have the shale basin depletion in the U.S. (what we have called the Shale Treadmill). We also seem to be coming off of the bottom of the capital expenditure cycle in oil and gas and well as energy, oil services, mining, and natural resources generally. And there is the export land model whereby the developing country oil exporters have rising populations as well as rising oil consumption per capita (and depleting resources), which together can lead to abrupt declines in exports.

Also, as McNally notes, price volatility in oil is self-reinforcing. (And we have been having a lot of price volatility.) A price bust does two things that set up higher prices in the future: it causes capital investment to be delayed or canceled, and it stokes future demand by disincentivizing fuel efficiency.

On the bearish side, more than anything else we should be paying attention to the economics of electric vehicles, since they are a potential substitute for the most important use of oil. The question is whether batteries will get cheap enough for there to be an actual, market-driven (not subsidized) transition from nature's perfect fuels (gasoline and diesel) to batteries for passenger vehicles.

It is certainly true that the price of lithium batteries (in $/kWh) fell substantially over the past decade, from about $1,355 per kWh to $150/kWh in 2022; an order of magnitude decrease. An electric vehicle can go about 3 miles per stored kWh of energy, which means that a 100 kWh battery pack will cost $15,000 and give you 300 miles of range. In California that amount of electricity at retail might cost $25. The gasoline to drive the same distance would be about twice as much but the gasoline motor would be much cheaper than the battery and electric motors.

So the big question is: will the battery "learning curve" continue to result in cost declines (in which case an electric vehicle transition would soon make sense)? Or was a meaningful part of the learning curve and order-of-magnitude cost decline caused by the commodities bear market from 2008-2020? 

We did see last year that EV battery prices rose for the first time due to higher input costs. But the electric vehicle manufacturers responded to this by changing the cathode chemistry of many EV batteries to lithium iron phosphate (LFP), which has a lower energy density than the nickel and cobalt chemistries do, but avoids using those more expensive metals. Iron is obviously not scarce, so the big question going forward will be the price of lithium. (Also, there's a downside of LFP, which is that it seems to do very poorly in cold weather.) We are starting to see the iron versus nickel catalyst price difference show up at retail. The F-150 Lightning standard range (230 miles) is much cheaper than the extended range (320) because the standard range is LFP.

There is a great tension between physics-based pessimism about natural resources and economics-based optimism (some might say cornucopianism) about the ability to respond to higher prices with substitution and invention. The LFP battery seems like a major point in favor of the cornucopian, economist viewpoint. We would not have thought it possible a few years ago to make a battery with just lithium and iron. 

The price of lithium is currently very high (about $20/lb) compared with where it was ($2/lb) before serious electric vehicle production began. One calculation that we found is that 80g of lithium per kWh is the theoretical maximum efficiency for a LFP battery and the current real world efficiency is under 50%. In other words, it takes more like 200g per kWh, which would mean that a 100 kWh battery (that can take you three hundred miles) requires 20 kilos (44 lbs) of lithium, which currently costs around $1,000.

Most of the cost of a LFP battery is therefore not from the lithium metal; perhaps only $1,000 for a $10,000+ pack that holds 100 kWh. That means that the price of the battery will not benefit much from a decline in the cost of lithium (which seems perfectly likely as capacity expands since lithium is not all that rare). On the other hand, when a manufactured item costs much more than its bill of materials, it seems more likely that the final cost will continue to decline due to learning curve effects. As Winfred Hirschmann wrote in the Harvard Business Review in 1964:

Practice makes perfect. A thing can always be done better not only the second time but each succeeding time by trying. This everybody knows. But how many know that the pattern of improvement can be sufficiently regular to be predictive? How many realize that such patterns can characterize, not only individual performance, but also the composite performance of many individuals organized to accomplish a common task?

The industrial learning curve quantifies such performance. It has evolved from experience in airframe manufacture, which found that the number of man-hours spent in building a plane declined at a regular rate over a wide range of production. Such continuing improvement was so common in the aircraft industry that it became the normal expectation in the war time mass production of aircraft; thus, production and other types of performance were customarily scheduled on some basis of progressive betterment. [...]

People do learn, and they learn according to a generally predictable pattern. The learning curve, I believe, is an underlying natural characteristic of organized activity, just as the bell-shaped curve is an accurate depiction of normal, random distribution of anything, from human I.Q.’s to the size of tomatoes. Wherever people strive to do better, improvements result; otherwise, how would progress take place?

By failing to capitalize on this natural phenomenon, managers will not encourage continued efforts once they become convinced that “further improvements are not possible.” Further improvements are always possible over time, so long as people are encouraged, or even ordered, to seek them. Thus, an understanding of the learning curve becomes of crucial importance to the business manager.

So what does this mean for oil investments - have we been disrupted?

It will take time to scale up lithium mining, lithium refining, battery production, mining of other metals, electric vehicle part production, and electric vehicle assembly. It will take time (a long time) for the existing fleet of internal combustion engine vehicles to turn over. In fact, it will be a big day when the number of ICE vehicles in use worldwide actually begins to decline. 

Another issue is that, in addition to the production of the vehicles, an entire, complimentary charging system has to be created as well to replace the distribution system for gasoline. That is everything from the generation of electricity to the transmission and distribution of it to the vehicle charging locations and appliances. The transportation sector uses about 24 quadrillion BTU of energy from petroleum (annually). The electric power sector delivers only 13 quadrillion BTUs, out of a total of 37 quadrillion BTUs consumed from fuels and resources. (The electrical system is very lossy.)

The electric vehicle transition can only happen as fast as the slowest one of these steps. All it takes in a series process (as opposed to parallel) is one bottleneck to delay the whole process. So, it seems like an electric vehicle transition is possible but it will be slow. And, in particular, the bottlenecks mean that it will be unexpectedly slow, which will continually discourage investment in oil production, which in turn should prolong the profitable part of the capital cycle.