Westwood Insight: Rig count doesn't tell the whole story...
Every Friday at 1pm Eastern, the oil market pauses to digest Baker Hughes' weekly rig count. Traders move positions, analysts write notes, and journalists file stories about what the number means for supply, prices, and the health of the US oil patch.
The problem is that the rig count, as currently reported, is an increasingly poor proxy for what actually matters: how many wells are being drilled, how productive those wells are, and how much capital is being deployed.
A rig is not a rig
The Baker Hughes count treats all rigs equally. A walking rig drilling four Permian wells per month from a single pad counts the same as a conventional rig drilling one vertical well in the Williston Basin over six weeks. The output of the first rig, in terms of both wells drilled and barrels produced, might be five to eight times greater than the second.
This matters because the US rig fleet has undergone a dramatic compositional shift since 2014. Today, over 85% of active rigs are horizontal, up from about 60% at the pre-downturn peak. The average rig is more powerful, more mobile, and more efficient than its 2014 counterpart. Operators retired their worst equipment during the downturn and invested in high-spec replacements.
The result: at 920 active rigs in November 2017, the US is drilling roughly the same number of wells per month as it was at 1,600 rigs in mid-2014. Wells per rig per year has increased from approximately 12 to over 20 in the most active basins.
Completions matter more
If the rig count is a lagging and distorted indicator of drilling activity, the completion count is an even worse gap in public data. There is no weekly equivalent of the Baker Hughes count for frac spreads or completion crews. Yet completions expenditure — the cost of fracturing, perforating, and bringing a well to production — accounts for 55-65% of total well cost in unconventional plays.
The DUC (drilled but uncompleted) well inventory, which stood at approximately 7,700 wells in September 2017, represents a massive overhang of deferred completions spending. As these wells are worked through, completion activity could run 15-20% above the level implied by the rig count alone.
What to watch instead
DW tracks several metrics that we believe give a more accurate picture of US activity than the headline rig count: wells spudded per week (from state permit data), lateral feet drilled per month, frac spread count (from primary research), and proppant consumption. Together, these paint a picture of an industry that is considerably more active than the rig count alone would suggest.
The rig count isn't useless — it captures directional trends and remains a useful cross-check. But it has been elevated to an importance in oil market commentary that it no longer deserves. The US drilling machine has changed. Our metrics for measuring it should change too.
Thom Payne, Oilfield Services Analyst, Douglas-Westwood