Summary
-
- Estimates for 2023 demand growth continue to get revised higher.
-
- We find that a supply deficit remains the most likely outcome in H2, supporting our bullish view.
-
- Changes in speculative flows look to be driving the oil price in recent weeks. We find that managed money net shorts are now larger than net longs, showing the market has become increasingly pessimistic.
This article is only available to Macro Hive subscribers. Sign-up to receive world-class macro analysis with a daily curated newsletter, podcast, original content from award-winning researchers, cross market strategy, equity insights, trade ideas, crypto flow frameworks, academic paper summaries, explanation and analysis of market-moving events, community investor chat room, and more.
Summary
- Estimates for 2023 demand growth continue to get revised higher.
- We find that a supply deficit remains the most likely outcome in H2, supporting our bullish view.
- Changes in speculative flows look to be driving the oil price in recent weeks. We find that managed money net shorts are now larger than net longs, showing the market has become increasingly pessimistic.
Consensus Expects Rising Demand Into Q4
The US Energy Information Agency (EIA), OPEC, and the International Energy Agency (IEA) all updated their 2023 oil forecasts. Here are the key takeaways:
- Global demand growth for 2023 continues to be revised higher. All three parties revised their outlook for global demand higher versus last month, the IEA by most. It revised global oil demand growth from 0.2mn b/d versus last month up to 2.2mn b/d. This brings their estimate of total demand this year to 102mn b/d versus just 101mn barrels for the EIA, whose estimates remain the most conservative (Chart 1).
- Expectations of demand growth are tilted towards H2. Both the IEA and OPEC expect global oil demand in Q4 to breach 103mn b/d, with Chinese demand rising above 16mn b/d – the highest on record. Both parties also expect demand growth for the year to peak in Q3, then decelerate slightly into Q4. In contrast, the EIA estimates global demand will slow after Q2, which is when they see Chinese demand peaking (Chart 2). We think these expectations of oil demand continuing to grow strongly into Q4 could be optimistic.
- Global demand growth rests on Chinese reopening prospects. Estimates of Chinese demand growth over 2023 also continue to get revised higher, with both the EIA and OPEC now expecting demand growth of around 0.8mn barrels in 2023, from 0.5mn barrels in January. The IEA is bullish still, forecasting China will contribute 60% of the 2.2mn barrel increase in global demand for 2023 (Chart 3).
- The oil market will likely be undersupplied later this year. Given these forecasts, the modal outcome remains that the oil market will be undersupplied in Q3 and particularly Q4. 2023 global supply is forecast between 101.1 and 101.3mn barrels a day, resulting in a deficit of c. 1.6mn b/d (Chart 4).
Why Undersupply in Q4 Is Our Base Case
There appears to be a narrow distribution for global supply estimates at 101.1-101.3mn b/d a day. We see risks skewed to the downside given the potential for US supply growth to disappoint – however, this is unlikely to change the equation materially.
Therefore, we can model a few scenarios for demand to indicate how large a supply deficit could get this year and the necessary conditions.
Here is a breakdown of each scenario.
- Scenario 1: Chinese demand stays weak – only growing by 0.5mn b/d while OECD demand falls by 0.5mn b/d.
- Scenario 2: Chinese demand normalises, growing by 0.7mn b/d while OECD demand stays the same.
- Scenario 3: Chinese demand normalises (as above), while OECD demand grows by 0.5mn b/d.
- Scenario 4: Chinese demand recovers strongly – increasing by 1mn b/d, while OECD demand grows by 0.5mn b/d.
Only Scenario 1 would see a net surplus this year. For Scenario 1 to materialise, we would need to see weaker demand in H2 given the data so far. Namely, our latest estimates show Chinese demand has increased by c. 0.75 b/d in Q1, while US demand also is likely to be 0.2mn b/d higher.
Scenario 4, which is closer to both OPEC and IEA estimates, seems unlikely without an upturn in the industrial cycle. Based on the latest Chinese credit and global manufacturing PMI data, we think this is unlikely.
Scenarios 2 and 3 would both result in a deficit while being less optimistic than both OPEC and IEA’s estimates. An outcome between these two would be our base case for this year.
Market implication: We continue to be bullish on oil to $85 given the likelihood of a supply deficit later this year.
Speculative Positioning Is Bearish
Looking at the Commitment of Traders (CoT) data via ICE and CFTC, we continue to see large shifts in speculative positioning from week to week, which looks to be driving the oil price.
We find that the change in managed money (MM) net longs has been correlated with the oil price in recent weeks – with its most recent eight-week correlation being 0.7 (Chart 6).
On speculative activity, net shorts at 51% now exceed net longs as a percentage of total managed money positions. Net longs as a percentage of total open interest (OI) stand at 5.1%, the lowest since the end of March and just before OPEC announced their voluntary supply cuts (Chart 7).
Further, the two-month 25 delta risk reversal for WTI indicates market sentiment. The risk reversal compares the implied volatility traders are willing to pay for upside versus downside protection. It continues to show an increasing bias towards downside volatility, though by less than in March (Chart 8).
Notably, the bearish positioning has come during US driving season when oil typically performs well.
Which markets has oil been most correlated with?
Three-month correlations show that oil has been most correlated with US regional banks and US stocks. This speaks to the market being more concerned about US hard landing concerns which could supersede improving oil market fundamentals. We also see oil has returned to its more usual negative correlation versus the dollar index (DXY), after being positively correlated for most of last year (Chart 9).