DVN · Devon Energy Corporation

conf 4/10 partial

enthusiasm:27.0 · trend:8 · quantifies:5 · impact:0 · under_radar:5 · credibility:12 · business_impact:4 · disruption:0 · commitment:6 · confirmation:3

Enthusiasm latest 9 / prev 6 (rising)

Devon's AI thesis rose sharply from pilot-scale production and drilling analytics in Q4 FY2025 to an explicit three-wave AI framework in Q1 FY2026. The most credible evidence is operational: AI-enabled gas lift moved from pilots to more than 850 wells, with a disclosed 2% to 3% pilot uplift and management saying current uplift is higher. Management still does not isolate AI's dollar contribution, so the quantified financial impact remains indirect.

Grounded on actual base — revenue $17.2B · net income $2.6B · net margin 15.4% · diluted EPS 4.2

These are next-fiscal-year annual uplift estimates, not next-quarter numbers.

Aggregate next-FY est. rev uplift: % · next-FY EPS uplift: 4.5% · vs analysts: inline · priced in: medium (model's call-read: high; verdict above is the hard-data one used for ranking) · confidence: 4/10

Assumptions: Percentages reported as percentage points. Tax rate 21% on cost savings (after_tax = saving*0.79). Cost/FCF program treated as bottom-line only (~0 topline). Phasing: 85% of the $1B already in the FY2025 net-income base, so only the remaining ~$150M is counted as FY2026 forward incremental; full $1B (29.9% EPS) shown as gross reference. The $1B is a BLENDED optimization target — AI is one of several named catalysts ($50M interest savings is non-AI) — so the AI-attributable slice is a fraction of the 4.5% and is not separately disclosed. Production-uplift claims (2-3%, 850 wells, 1,500 target) are real percentages/counts but the piloted-well share of total production is undisclosed, so none can be anchored to the $17.188B revenue base; we decline to apply the per-well uplift to total revenue (as Answer Y did) since that invents the base the rule forbids.

Top line: No sizable topline impact can be quantified. Every production claim (2-3% pilot uplift, 850 wells on autonomous/Smart Gas Lift, 1,500-well target) is a genuine adopter-side efficiency gain but is expressed as a per-well uplift or well count with no disclosure of those wells' share of total production, so it cannot be applied to the $17.188B revenue base. est_rev_uplift_pct = null (unanchored, not zero).

Bottom line: The only anchored figure is the $1B optimization/FCF target. Full run-rate after-tax = $790M = 29.9% of $2,642M net income, but that overstates AI: it is a blended target ($50M of it is non-AI term-loan interest savings, AI is one of 'several catalysts'), and 85% ($850M) is already captured inside the FY2025 base. The clean FY2026-forward incremental is the remaining ~$150M -> $118.5M after-tax = 4.5% EPS, of which the AI-only portion is an undisclosed subset. So the bottom-line AI lift is positive but modest and largely already in the run-rate.

Consensus already models ~+37-44% revenue ($17.188B->$23.567B) and ~+35-42% EPS for FY2026 — that growth is production/acquisition-driven, not AI. The quantifiable AI/optimization lift is ~4.5% EPS forward (and less for the pure-AI slice), which is dwarfed by and well within that consensus EPS jump; 85% of the $1B program is already in the FY2025 base analysts started from. AI therefore does not push numbers beyond consensus.

• After years of cleaning and organizing our data, we have a fully firewalled internal tool called ChatDVN that has been up and running for three years and is today a standard part of our daily workflow.
• We piloted this back in 2025, and we saw about a 2% to 3% uplift.
• We've seen those advantages already in the drilling results that we've had.

• We are now deep into seeing the benefits of wave two, where the AI is doing the heavy lifting of complicated calculations and time-consuming work.
• Wave two value is showing up in cutting-edge drilling and completion time, directly translating into lower capital costs.
• We are also having very significant wins in production, leveraging AI-created tools to do real-time artificial lift optimization.
• We now have over 850 wells on fully autonomous artificial lift optimization with a very impressive productivity improvement.
• We have now moved into full implementation in the Delaware Basin.
• We are over 850 wells at this point, and we have seen uplift in excess of what we saw in the pilot phase.
• What I can confidently say and what we're really excited about at the team level is a lot of the investments we've made in artificial intelligence and in the platforms that we've built over the last year are really coming to fruition in the production space.

• We are now moving into wave three, where we are redesigning internal processes from the ground up with AI at the center.
• We are on our way to 1.5 thousand wells across the portfolio.
• We are already taking similar AI-derived models to look at other forms of artificial lift—ESPs and rod pumps.
• Much like Smart Gas Lift, these are programs we will be able to scale.
• We're going to see a lot of benefits flow through on the production side, ultimately resulting in kind of our ability to lower capital long term, and we'll see improvements on the LOE.

85/100 track record   delivers  6 calls reviewed

Devon made few explicit AI-only promises, but its quantified technology/AI-enabled optimization targets — chiefly the $1B business-optimization free-cash-flow goal — were tracked transparently and delivered ahead of schedule, alongside concrete scaling of autonomous artificial-lift optimization. The track record is credible; the newer AI-specific downtime and merger-synergy targets remain too early to judge.

Est. revisions rising  ·  Fwd P/E 11.6  ·  EV/Sales 2.1x

AI claim maps to N G L Product Sales

AI enthusiasm across 6 calls — trend ↗ rising

AI moved from absent to operational agents, then scaled into autonomous lift optimization, ChatDVN workflows, and explicit productivity gains.

• 2026-05-15 · Devon Energy: Still Cheap With FQ2'26 & H2'26 Commodity Tailwinds - Reiterate Buy

6/10 qualitative impact   moderate  near-term · mixed evidence

Where AI matters: production optimization, drilling efficiency, LOE/capital costs

Devon has moved beyond pilots: autonomous artificial-lift optimization is on 850+ wells with a disclosed 2%-3% pilot uplift and rollout toward 1,500 wells, plus AI-assisted drilling/completion workflows and internal ChatDVN productivity. Still, the AI-only dollar contribution is not isolated and the anchored forward EPS impact is only an undisclosed subset of the remaining optimization savings, so this is meaningful operational leverage rather than company-transforming upside.

Caveats: AI-specific contribution is blended into broader optimization and synergy programs; Per-well uplift is not tied to Devon-wide production or revenue share; Commodity prices and reservoir quality can overwhelm AI productivity gains; Efficiency tools may diffuse to peers and reduce relative advantage

AI DISRUPTION / CANNIBALIZATION RISK  tailwind · 1/10

AI does not commoditize Devon's core product: the company sells hydrocarbons, not labor hours, content, search traffic, or software functionality that AI can directly replace. The main AI effect is industry-wide efficiency, which may eventually lower marginal costs and supply curves, but that is not a direct cannibalization of Devon's revenue model.

option liquidity: good

proxy inputs — dollar-ADV $734M · beta 0.478 · px $46.22

source: proxy (no options chain on FMP)
FMP /stable/ exposes no options-chain endpoint on this key, so ATM IV, bid-ask spread and open interest are unavailable. Liquidity below is a PROXY from dollar-ADV, beta and price level (a stand-in for option depth), not measured option-market data.