Storage withdrawals and falling rig count have been the main sources of hope that U.S. tight oil production will fall and that oil prices will rebound. That hope is fading as it is now clear that recent withdrawals from U.S. crude oil storage are because of price, not falling supply, and that the drop in rig count has stalled.
Figure 1 below shows the relationship between U.S. crude oil storage inventory and WTI price. The thinking around recent withdrawals from storage is that this reflects depleting supply. The data, however, reflects that traders were storing crude oil during the price collapse in order to realize higher prices later. With rising prices over the last month, traders are selling their stored volumes. The recent inventory build correlates almost perfectly with the fall in oil prices and the withdrawals from storage over that last 3 weeks correlate with the 35% increase in oil prices since late March.
Figure 1. Monthly change in U.S. crude oil inventory and WTI oil price (3-month moving average of inventory volumes). Source: EIA and Labyrinth Consulting Services, Inc.
Previous builds and withdrawals from inventory also correlate with price but generally price followed changes in inventory. In the recent case, price lead inventory changes.
The other important point about Figure 1 is that inventory additions and withdrawals are seasonal. A Spring and Summer “de-stocking” is normal. In that sense also, the recent withdrawals from storage say less about oil supply than they do about northern hemisphere summer driving demand and the end of regularly scheduled refinery maintenance in the U.S.
Falling U.S. rig counts have been the main hope for a drop in U.S. oil production that might help balance the global market. This appears to have ended as shown in Figure 2 below.
Figure 2. Tight oil horizontal rig counts for key tight oil plays. Source: Baker Hughes and Labyrinth Consulting Services, Inc.
The Bakken rig count has stabilized between 78 and 80 rigs over the last month. Decrease in Niobrara rig count ended at 28 in mid-April and has been steady at 30 rigs for the last 3 weeks. The Eagle Ford play reached its low in early May at 102 and has been at 104 rigs for the last two weeks. The Permian trend remains lower although one rig was added last week for a total of 172 horizontal rigs. In many ways, these rig count trends reflect the economic attractiveness of the plays.
It is true that these rig counts remain substantially below 2014 highs and the lag from spud date to first production is about 5 months in the Bakken and 3 months in the other plays. In other words, the effects of lower rig counts have not yet been reflected in current production data which lags about 3 months itself. Published EIA production for February and April is an estimate.
The third pillar of hope for decreased U.S. oil supply has been growth in demand because of low price. That support remains strong as March vehicle miles traveled data indicates the highest gasoline demand since 2007, just before the Financial Crisis began.
The recent 35% increase in WTI and 40% increase in Brent prices is based more on sentiment than real evidence and I expect that prices will fall unless tangible data appears to support present prices. A geopolitical risk premium or an OPEC production cut in early June would constitute a “hard” reason for higher prices.
Present data, however, suggests that the global over-supply has gotten worse, not better, that overall demand for liquids remains weak, and the world economic outlook is discouraging. At the same time, market movements are not always based on fundamentals. In the long run, however, fundamentals rule so I maintain my view that the current price surge is at best premature.