EOG Resources (EOG)

The AI said 27. The records said 8.

Every week we run the same research questions through an AI assistant to pull the raw numbers for the five boxes. This week it came back confident: EOG has "grown its regular annualized dividend every calendar year since 1999," a 27-year streak. That would have been a perfect 20 on the Raises box and a score of 84.

It was wrong. We pulled EOG's actual quarterly dividend record, every payment back to 1995, and found the payout parked at $0.1675 a share for fourteen straight quarters, from late 2014 through early 2018. That is the oil crash. EOG held the dividend flat for three full years instead of cutting it, which is admirable, but flat is not a raise. The streak restarts in 2018 and today stands at eight years.

So the Raises box gets a 12, not a 20, and EOG scores 76, Strong, not 84. If we had taken the AI's number at face value we would have published a score eight points too high, and somebody would have caught it eventually. We would rather catch it ourselves. That is the whole reason the five boxes come with the numbers attached.

The rest of the scorecard:

• Pays: 28 years of dividends since EOG became an independent company in 1999, never cut, never suspended. The company's own CFO confirmed it on the last earnings call. 20/20
• Raises: eight straight years, as above. 12/20
• Safe: 31.75% of earnings paid out. Covered three times over. 20/20
• Undervalued: today's 2.72% yield sits about 1% above its own five-year average. Fairly priced, not a bargain. 12/20
• Yield that matters: 2.72%. 12/20

Why we bought a 76

Our previous low was NextEra at 80 back in week four. Every score before this one was 80 or better, so a 76 deserves an explanation.

First, EOG passes every box. None of the five is broken, and a broken box is the thing we refuse to buy. Clorox had a 104% payout. Campbell's had a frozen dividend. Cisco failed the yield floor. EOG's weak boxes are weak the honest way: a middling yield and a fair price. That is a different kind of 76 than a 92 with a rotten plank in it.

Second, the business behind the dividend is unusually sturdy for an oil company. Morningstar rates EOG's capital allocation Exemplary, one of very few energy names to earn that. Its breakeven oil price is under $35 a barrel, it carries essentially no net debt, and last quarter it returned two-thirds of its free cash flow to shareholders. In May management told investors its own stock was fairly valued, which is a sentence almost no management team says out loud. When oil crashed in 2015 and 2016, ConocoPhillips cut its dividend by two-thirds. EOG held. That freeze cost it a box today, and it is also the reason we trust the Pays box.

Third, the value gate actually did its job here. Back in July, an earlier screen flagged EOG as trading richer than its own norm and we passed. Two months later the yield is back to its average and it clears the gate, barely. We didn't chase it. We waited for the price to come to the rule.

The honest part

This is a commodity business. EOG does not set the price of oil; the world does, and this year the world has been volatile, with tanker traffic through the Strait of Hormuz disrupted and the UAE walking out of OPEC. Morningstar gives EOG only a narrow moat, because a cost advantage is the only kind of moat an oil producer can have. We are not buying a forecast for oil prices. We are buying a company that has paid through every oil cycle since 1999.

Dividend growth is slowing. EOG's raises ran about 10% a year in 2023 and 2024, then 7%, then 4.6% last October. We stored 4.6% in our income projection, not the flattering 24% ten-year average that the catch-up years of 2018 to 2022 produce.

The special dividends don't count. EOG paid out a string of one-time special dividends in 2021, 2022 and 2023, some as large as $2.00 a share. It has paid none since, though Morningstar expects them to resume. Our rules count specials as income received when they arrive and exclude them from forward income, yield, and the projection, because a one-off is not a rate. If you see EOG quoted at a 4.6% yield somewhere, that number includes specials the company has not declared. Ours is 2.72%.

The three we said no to

The screen this week returned six energy and technology names. Three were worth a look.

Accenture, again. It topped the screen a second week running, with a 3.55% yield against a historical norm near 1.5%. We said no last week for a reason the five boxes cannot see, and nothing changed: it sells hours of skilled analytical work, and that is the work AI compresses first. A yield at two and a half times its norm is the market agreeing.

Qualcomm. A 23-year raiser with a 43% payout, and it failed our 2.5% yield floor at 2.02%. The research report was a bigger warning than the yield: Morningstar cut its fair value from $200 to $170 in July, Apple's modem business goes to zero by fiscal 2028, and the case now rests on a data-center chip business that barely exists yet.

Archrock, a natural-gas compression company yielding 2.8%. It cut its dividend in 2016 under its old name, Exterran, and only resumed raises in 2023. Its yield today is a fraction of what it was in 2021, which means the stock is far richer than its own norm, and long-term debt is 60% of its capital. Two broken boxes and a weak balance sheet.

The numbers on this week's buy

• Invested: $99.99
• Bought: 0.6651 shares at $150.34
• New annual dividend income added: about $2.71 per year
• The portfolio after 14 weeks: $1,399.99 contributed, $2.55 reinvested, worth about $1,375.85, throwing off about $49.59 a year, roughly $4.13 a month
• Thirteen holdings across ten sectors. Energy is the newest. Technology is the last one we don't own.

Two housekeeping notes, because we publish everything.

Tractor Supply paid us $0.78 on September 8, and the brokerage reinvested it straight back into Tractor Supply, buying 0.022 more shares. That was not supposed to happen. Our published rule is that every dividend pools into the next weekly pick, and reinvestment was meant to be switched off on every holding. It was still on for this one. The ledger records it exactly as it happened, a $0.78 dividend received and $0.78 reinvested, and the setting is being corrected. Lifetime dividends collected now stand at $2.55, every cent of it back in the portfolio.

And for the first time since we started, the portfolio is worth less than we put in, about $27 under, or 1.9%. Prices dipped across the board last week. We mention it because we publish everything, and because it is the part we care about least. The income line went up again this week. That is the line this portfolio is built on.

The running totals are on the portfolio page, and the slider lets you drag through time to watch the income grow.

That's week fourteen. We bought a 76, we told you why, and we told you the AI got the streak wrong before anyone else could. See you next week for pick number fifteen.

This is not financial advice. We're individual investors sharing our own approach and the reasoning behind our own real decisions, not licensed advisors. We own, or are buying, the stocks we write about in this series. Do your own research and consider speaking with a financial professional before investing. Dividend figures, yields, share counts, and prices are accurate to the best of our knowledge as of the publish date and will change over time.