US total liquids production averaged 17.26 mb/d in September, up m/m by 0.12 mb/d and higher y/y by 2.89 mb/d. This was slightly higher than our expectations (+73 thousand b/d m/m). After a large increase m/m in August figures (+0.41 mb/d m/m), the September data is more in line with our expectations. Crude production increased m/m by 0.13 mb/d to 11.48 mb/d, up y/y by 1.98 mb/d. This was slightly higher than our model, which projected an increase of 0.10 mb/d m/m. The m/m increase was driven by Texas, up 0.11 mb/d and North Dakota, up 64 thousand b/d m/m as well as Alaska, up 43 thousand b/d m/m. These gains were partially offset by a fall in Gulf of Mexico (GoM) production as the region was affected by outages from Tropical Storm Gordon. NGLs output totaled 4.63 mb/d, higher y/y by 0.9 mb/d, the highest growth on record.
Texas and New Mexico (a proxy for the Permian and Eagle Ford) crude output rose by 0.13 mb/d m/m, higher by 1.41 mb/d y/y to 5.43 mb/d. While these areas continue to grow, growth was lower than the previous print, with m/m growth slowing by 46 thousand b/d. Colorado and Wyoming production (a proxy for the Niobrara) rose by just 11 thousand b/d m/m to 0.73 mb/d after a surge of 56 thousand b/d m/m in August. Production in Oklahoma and Kansas (a proxy for the Anadarko basin) rose by 6 thousand b/d m/m to average 0.66 mb/d (higher by 0.10 mb/d y/y). Bakken output continued to surge, up m/m by 67 thousand b/d to 1.36 mb/d, likely as several pipeline expansions were brought online. GoM production was lower on the month due to outages from Tropical Storm Gordon, down 0.15 mb/d m/m. October numbers should also be depressed due to Hurricane Michael with an estimated 0.16 mb/d offline. However, we expect growth to resume in November as Big Foot and Thunderhorse production ramp up.
With flat price down by more than $12 since the start of November, more attention is being paid to producer breakevens. Most producers need a WTI price between $45 and $55 (especially with wellhead prices trading at discounts to WTI, in some cases quite steeply) in order to profitably drill new wells. Therefore, we would expect completions to stay depressed with a sustained lower flat price. Indeed, our model predicts that for every $5 of flat price drop, roughly 0.10 mb/d of production is lost over a year. With the usual 3-6 month time lag between changes in prices filtering through to production trends, the current price drop will not affect our Q4 18 production forecasts, which are already tempered based on takeaway constraints as well as reports from service companies alerting a slowdown. Should this price environment persist, drilling budgets for 2019 will be reduced, and there is indeed downside risk to our 2019 crude output growth of 1.2 mb/d y/y (with Permian y/y growth at 0.78 mb/d). The fall in WTI has left prices below the original $50–55 budget set out when the industry went into rebuild mode in 2016.