First appeared in Q3 2021 commentary
Global oil markets could not have looked grimmer than back in the summer of 2020. COVID cases were increasing again and further global economic lockdowns were being threatened. Only two months before, oil prices had crashed into negative territory for the first time in history as global inventories had surged and threatened to overflow. Between the never-ending pandemic and electric vehicles, common investment wisdom believed oil demand was in secular decline, bloated inventories would remain elevated forever, and oil prices would never recover.
At Goehring & Rozencwajg, we love to undertake in-depth insightful research that identifies newly developing investment trends and often comes up with conclusions that differ vastly from consensus opinion. Our goal is to share these results with our investment clients and partners at least twelve months before they become headline news in the financial press. This is an ambitious goal and we don’t always get it right. However, the satisfaction in recognizing trends long before the general investment community not only brings large profits to our investors, but a huge amount of professional pleasure to us as well.
To that end, we titled our 2Q2020 letter “On the Verge of an Energy Crisis”. At the time, no one agreed with us. Now headline after headline talks of an “unforeseen” energy panic. It looks like the energy crisis we discussed 14 months ago is now here.
If the energy crisis has arrived, where does Goehring & Rozencwajg see things 12 months from now? By the end of 2022, we believe global oil demand will have exceeded pumping capability for the first time in history. Just as no one agreed with our assessment of an emerging energy crisis this time last year, almost no one agrees us today either. Instead, conventional wisdom strongly believes OPEC spare capacity will be returned, eventually throwing the market into huge surplus in 2022.
When we say demand will exceed pumping capability, what do we mean and why is this so unprecedented? While it is true that demand has exceeded supply many times in the past, it has never -- according to our models -- exceeded pumping capability. What is the difference? Supply is the amount of crude produced and sold into the market. Pumping capability, on the other hand, is a more subjective term that refers to total supply that could, in principle, be produced quickly with minimal additional capital spending. Therefore, while demand has exceeded actual production – most recently in 2006/2008 and 2020 – there has always been the availability of additional spare capacity. This will not be the case as soon as the end of next year. The consequences could be dire.
Even during the twin oil crises of the 1970s, when OECD countries were forced to implement gasoline rationing, demand never actually exceeded pumping capability. OPEC production fell in 1973 in retaliation for the United States’ involvement in the Yom Kippur War and again in 1979 following the Iranian Revolution. Demand exceeded supply in both cases and prices spiked, however pumping capability was unchanged – production had been curtailed far below pumping capability. Our models estimate that pumping capability exceeded demand by nearly 5 mm b/d following both the first and second oil crises.
More recently, oil markets were in a severe deficit between the summer of 2007 and 2008, caused primarily by disappointing non-OPEC production trends. Demand exceeded supply by approximately 500,000 b/d, causing inventories to fall and prices to surge from $55 to a record-setting $145 per barrel. However, even then OPEC maintained over 3 m b/d of spare capacity, according to the IEA.
Last year, COVID-19-related shutdowns led to a widespread reduction in crude demand. Inventories built and US shale producers actively shut-in as much production as possible. OPEC+ agreed to a nearly 8 mm b/d supply cut to help balance markets. Inventories built at first but then immediately drew, as demand rebounded while production remained subdued. Prices rose as the market realized it was in near-term deficit by as much as 1.5 m b/d. Even though the market was in severe deficit, global demand did not exceed pumping capability. US shut-in production could quickly be reinstated and OPEC+ maintained ample spare capacity. We estimate that even as inventories were drawing more than 1 m b/d, pumping capacity exceeded demand by as much as 5 m b/d.
By the end of next year, our models tell us this cushion will have eroded completely. To understand why, let us explain the current crude balances and our projections for 2022. According to International Energy Agency’s (IEA) Oil Market Report (OMR), 4Q21 demand will average 98.9 mm b/d. Although this is an incredible 5 m b/d above last year’s level, it remains 1.7 m b/d below the pre-COVID 4Q19 level. Global liquids production averaged 96 mm b/d in September, pointing to a market nearly 3 mm b/d in deficit. OPEC+ production reached 42.2 mm b/d in September and based upon their current agreement (last reconfirmed on November 3rd), production quotas will grow by 400,000 b/d per month. The IEA estimates that OPEC+ sustainable capacity – what we are calling pumping capability – totals 50.1 mm b/d, nearly 8 mm b/d higher than current production.
Given these figures, why do we believe demand will exceed global pumping capacity by the end of 2022? The primary reason is demand.
We have long written that emerging market demand has been much stronger than investors (and the IEA) expect. Few energy analysts seem to make this adjustment and their models end up chronically underestimating non-OECD demand. Every demand dip over the last 20 years has been less severe, and rebounded more quickly, than forecast. This was true following the 2008 global financial crisis, when oil demand exceeded its pre-crisis high within a mere 18 months. It was true following the European debt crisis, when global demand actually grew by 2.5 mm b/d despite EU demand falling by 700,000 b/d. And it was true during COVID-19.
As recently as last fall, analysts predicted demand might never again reach 2019 levels; once again these estimates have proven too pessimistic. While total global demand has not yet surpassed 2019 highs due to ongoing travel restrictions, several countries have exceeded their pre-COVID levels. The three largest sources of demand, China, India and the United States, all registered record monthly demand in 2021. Moreover, widespread global travel restrictions mean demand reached record levels in these countries despite jet fuel demand still well below pre-COVID readings. The US dropped its remaining travel restrictions on November 8th, and we expect demand will surge as other countries follow suit.
The IEA expects global demand to reach 99.6 mm b/d next year, with the highest reading in Q4 at 100.2 mm b/d. Our models tell us the IEA is again dramatically underestimating demand by as much as 1.5 m b/d. We carefully monitor so-called “missing barrels” which occur when the IEA’s own figures do not balance. For example, the change in inventory should equal the difference between supply and demand but many times this is not the case. We euphemistically refer to “missing” barrels as those that were allegedly produced but neither consumed nor stored in inventory. Either consumption is understated or production is overstated -- inventories are usually accurate. Our S-curve models suggest demand is the problem. Based upon the “missing barrels,” we believe demand is currently running at least 500,000 b/d ahead of expectations and that this will continue into 2022.
Furthermore, the ongoing natural gas crisis will dramatically increase oil demand by another 500,000 b/d as utilities switch from gas to oil wherever possible. The IEA has already started accounting for this by revising 2022 demand higher across the first three quarters by 300,000 b/d on average. However, they have offset much of this increase with a huge 400,000 b/d downward revision to 4Q22 demand without offering any explanation. This is a technique we have observed in the past: the IEA will revise near-term demand higher as actual results come in while revising longer-term demand lower, keeping the full year projection unchanged.
This year is no different. Over the last 90 days, the IEA has revised first half 2021 higher by 300,000 b/d as actual results have come in ahead of expectations, and offset the increase by lowering second half demand by the same amount. We believe this is a prime example of “kicking the can down the road” and will result in a flurry of upward demand revisions both this year and next.
Taken together, these two revisions will likely leave 2022 demand much higher than expectations, notably in Q4. Instead of averaging 99.6 mm b/d, we believe 2022 demand could exceed 100.6 m b/d – an all-time high. Q4 demand could reach 101.6 m b/d.
Turning to supply, the only source of non-OPEC+ growth over the past decade -- the US – remains stubbornly 1.8 m b/d below its late 2019 peak and does not appear to be improving. Over the past eight months, US production has grown by only 7,000 b/d per month compared with 150,000 b/d per month on average in 2012-2014 and 2017-2019. Higher prices have not led to materially higher drilling budgets for 2021 or 2022 and the oil rig count remains depressed at 450 rigs – up 280 rigs from the bottom but still half the pre-COVID level.
It is no surprise the US is not seeing a strong rebound in production: the shales are suffering from depletion, a topic we first discussed in late 2019. Every shale basin except the Permian is experiencing outright decline. Over the last 12 months, the Eagle Ford and Bakken have declined by 4,000 b/d per month on average compared with monthly growth of 20,000 b/d as recently as late 2018. The Permian is the least developed of the major basins and we have often predicted it will still be able to grow, albeit at a much lower rate than in years past. Over the past 12 months, Permian growth averaged 35,000 b/d per month – 65% less than it grew in 2018. The remaining shale basins are declining by a total of 13,000 b/d per month compared with 25,000 b/d monthly growth a few years ago. Taken together, shale production is only growing 14,000 b/d per month compared with 160,000 b/d in late 2018 – a slowdown of 90%.
Even this meager growth is being distorted and will likely drop in the coming months. The distortion is being caused by the harvesting of drilled-but-uncompleted wells (DUCs). During last year’s oil collapse, companies elected to postpone completing a well where possible to save on capital expenditures (completion is half of total well costs). This caused the DUC inventory to swell to unprecedented levels. As prices recovered, completing the DUCs generated extremely high incremental returns on investment given the drilling capital was already “sunk.” Since June 2020, twice the number of wells were completed than drilled in the Eagle Ford and Bakken. In the Permian, 66% more wells were completed than drilled. Across all the shale basins, 60% more wells were completed than drilled. Our neural network estimates that without the DUCs, shale production would be 800,000 b/d lower than it is presently and that every basin, including the Permian, would be in sustained decline.
Clearly this trend cannot persist indefinitely. A certain number of DUCs are required for normal shale development. Historically, the industry operated with a DUC inventory equivalent to five months of drilling activity. Using this metric, “excess” DUC inventory (i.e., over and above five months of drilling) peaked at 5,000 locations across the Eagle Ford, Bakken, and Permian in June 2020. As of September 2021, the “excess” inventory had fallen to 1,300 locations – a decline of 75%. At the current rate, we estimate DUC inventory will reach the equivalent of five months of drilling activity by mid-2022, at which point it will be difficult to draw down DUC inventory any further. For shale production to grow, drilling activity will need to increase dramatically – unlikely given the ESG capital pressures faced by publicly traded oil companies.
Furthermore, depletion problems across the shales persist. Our neural network initially pointed to the fact that once a basin has developed 50% of its Tier 1 wells, total production begins to plateau and then decline. We used this to correctly predict the Eagle Ford and Bakken would peak in late 2019. At the time, we stated that Permian production would still be able to grow for the next few years given only 35% of Tier 1 wells had been developed. Our models now tell us that 45% of Tier 1 Permian wells have been developed, implying we are much closer to its inevitable plateau and decline as well. By the end of 2022, we believe the final US shale basin will cease to growth.
The IEA currently estimates US production will grow by 1 mm b/d in 2022 and by 900,000 b/d from the last monthly reading, but our models suggest this is too optimistic. We admit that we underestimated US growth for most of 2021, but this was largely due to accelerated DUC harvesting. Given the vast majority of excess DUC inventory has now been completed, we believe our initial predictions will take hold starting in mid-2022. Instead of growing by 900,000 b/d from here, we believe US production might only grow 300,000 b/d.
It is unlikely that non-OPEC+ production outside the US will be able to grow next year. This bloc of production has been challenged for years and has chronically disappointed. A dearth of new discoveries has translated into less growth for more than a decade. Since December 2020, the IEA has been forced to revise 1H21 production dramatically lower from this group as actual results have come in far below expectations. Over the last ten months, the IEA has revised 1H21 non-OPEC+ ex US production lower by 800,000 b/d, led mostly by Brazil and Norway. However, just as with demand, the IEA has taken to offsetting revisions in the near months with equal but opposite revisions to later months. At the same time as the IEA revised 1H21 production lower by 800,000 b/d, they revised 2H21 production higher by 200,000 b/d leaving their full-year projection lower by less than 300,000 b/d.
The IEA remains equally optimistic for non-OPEC+ ex US growth in 2022. In their most recent report, the IEA projects this group will grow production next year by 900,000 b/d. In aggregate, they believe non-OPEC+ ex US will average 30.6 mm b/d next year whereas our models predict this could be as low as 30 mm b/d.
After making the adjustments discussed above, we believe demand will exceed pumping capacity by 4Q22. The IEA believes 4Q22 demand will reach 100.2 mm b/d, while US production will average 17.7 mm b/d, and non-OPEC+ ex US will average 30.9 mm b/d. OPEC+ NGLs are expected to average 7.9 mm b/d, leaving the call on OPEC+ crude at 43.7 mm b/d – 1.6 mm b/d higher than today and well within their pumping capability.
Our models suggest 4Q22 demand will reach 101.6 m b/d while US production may only average 17 m b/d. Assuming non-OPEC+ ex US pumps at 30.0 m b/d in 4Q22 and OPEC+ NGLs average 7.7 mm b/d, the call on OPEC+ crude jumps to 46.9 mm b/d or 4.7 m b/d above current rates.
Can OPEC+ meet this increased demand? We believe it will be difficult.
The only countries with material remaining spare capacity are Saudi Arabia, UAE, Kuwait, Iran, and Russia. Iraq has spare capacity, but given the security considerations, it is extremely unlikely production will grow in the near or medium term. We have discussed our skepticism regarding Saudi spare capacity in the past and intend to revisit the important topic next quarter. Ultimately, we believe Saudi Arabia can produce between 10 and 10.5 mm b/d – well below the stated 12.2 m b/d capacity. Saudi has only produced above 10 mm b/d on two occasions and both times it was for only a brief period and the fields had to subsequently be rested. Assuming Saudi has pumping capacity for 10.5 m b/d (a big if), we believe total OPEC+ crude capacity to be 46.9 m b/d – not enough to meet global demand by 4Q22.
Twelve months ago, few people listened when we predicted an energy crisis was imminent. Now, our models suggest that we could be entering a new period in the history of oil – a period without any excess global pumping capability. The ramifications could be huge. Investors today have hardly any exposure to oil producing companies at all. After having averaged 10-15% of the S&P 500 for decades (and reaching a maximum of 30%), energy stocks today stand at less than 3% of index.
Just as few investors saw the energy crisis, fewer believe an oil crisis is looming. Position yourselves accordingly.
Intrigued? To read more on this subject, we encourage you to download the full commentary, available below.