The screener ran without me

I stopped writing. I did not stop trading.

Summer got busy. The weekly posts fell off the list. But the screener kept running, the signals kept coming, and three positions got added — one of them the day before the screener it was tracking dropped that name from the list. That one stings a little. More on that in a minute.

This post covers June 20 to today. Three weeks, a rough patch in late June, a recovery through July, and a portfolio that came through up +1.92% on the adjusted curve.

What I bought

Three real decisions this period, not counting the auto-invest drips that run in the background every week.

CF Industries (CF), July 1 — 13 shares at $108.00.

CF had been sitting in the Quantum screener for multiple cycles: NMS 6.0, F-Score 8 or 9 depending on the cycle, Margin of Safety around 35–48%, EV/EBITDA below 6, beta −0.75. That negative beta is the thing. In a portfolio that's already heavy on tech and quality growth, a fertilizer and chemicals company that tends to zig when equities zag earns its slot. I'd been watching it for four weeks. July 1 I bought it. It's up +13.6% since.

EBAY (EBAY), July 2 — 5 shares at $113.15.

EBAY showed up in Bedrock at Rank 10 this week. NMS 6.0, F-Score decent, a free-cash-flow business that's been quietly compounding while nobody's looking at it. It's not a screamer and that's the point. The buy is small — a starter. I'll add if the screener keeps it at this rank and the thesis gets clearer with a closer look.

TechnipFMC (FTI), July 9 — 14 shares at $70.13.

FTI had been one of the steadiest names in the Quantum screener: NMS 6.0, F-Score 9, ROIC near 25%, consistently at Quantum + Prospector for months. I bought it Thursday July 9. Two days later — this week's screener — FTI was gone from Quantum. Dropped out entirely.

I want to be clear about what that means and what it doesn't. FTI dropping from Quantum is a signal to monitor, not a sell trigger. The buy thesis was based on the fundamentals the screener captured — not on a promise that the screener rank would hold. If the fundamentals are intact (which I haven't confirmed yet — I need to run a Layer 5 check), the position stays. If they've broken, it goes. I'm not going to exit a name two days after buying it because a screener snapshot changed. But I am flagging this for a closer look next week.

What I watched and didn't buy

Two signals ran hot for most of the period and neither crossed the threshold for a buy.

ADBE — eight consecutive weeks at Bedrock R1.

Adobe hit its sixth, seventh, and eighth week at the top of Bedrock during this stretch. RSI spent most of late June below 30 — the most extreme oversold readings in this screener run. Shareholder Yield above 10%, Margin of Safety above 30%, F-Score 8. The case has been clean for two months.

I haven't bought it. The sticking point has been NMS — it sat at 5.0 when I wanted to see 6.0 for a Buy-Ready signal on a name I don't already hold. This cycle it's still NMS 5.0, RSI 54.9 — the window from the June lows has closed. If RSI pulls back toward 35 again and NMS recovers to 6.0, the math changes. Until then, I'm watching.

NFLX — RSI 18.2 at the June 26 cycle.

Netflix hit 18.2 on the Relative Strength Index in the June 26 screener. That was the lowest reading in the entire screener universe that cycle. Cross-screener (Bedrock and Prospector), held position, extreme oversold.

I didn't add. NMS 4.0 (Warming Up) and negative Margin of Safety at −9% are the reasons. The screener says the business is Warming Up — not Buy-Ready. The technical pressure was real but the quantitative entry case wasn't there. The RSI bounced to 47.8 by the July 7 cycle and has since slid back to 37.6. No add.

The gold thesis took a hit

Every buy, every sell, every loss — and the gold cluster had a rough three weeks.

AEM (Agnico Eagle) was the hardest: −11.9% over the period. B (Barrick Mining) was down −9.1%. NEM (Newmont) down −8.2%.

The fundamental case for all three is unchanged. F-Scores remain 8–9 across the cluster. Barrick is the one I'm watching more closely: its NMS score dropped from 5.0 to 2.0 on the July 7 cycle and has stayed there this week. Two consecutive cycles at NMS 2.0 is the trigger, per my own rules, for a Layer 5 council check — a proper thesis review with news context, not just the numbers. That's on the list for next week.

The gold thesis itself is now in its seventh consecutive week confirmed across the screeners. NEM, AEM, Barrick, KGC (Kinross), AU (AngloGold Ashanti), and ATAT (Atour Lifestyle) all remain in the screener universe. The RSI across the gold cluster drifted lower this week. No name is at an oversold extreme yet. The thesis is intact. Barrick's NMS anomaly is the one thing I need to investigate.

The tech side ran while I looked away

META was up +15.9% over the three weeks. I hold a small position, haven't added, and the RSI is now at 67.6 — getting into overbought territory. Not a sell signal; just not the time to add.

The rest of the tech-adjacent held names had a quieter version of the same story. INTU was down in the first two weeks of the period — RSI touched 29.8 on June 26, its first sub-30 reading this cycle — and then recovered to neutral at 45.5. Same with MSFT: hit 28.2 on June 26, recovered to 47.7.

Both of those sub-30 readings happened while I wasn't writing about them. The process flagged the technical stress. I saw it in the screener reports. I held.

How the portfolio did

The Blue Portfolio finished the period up +1.92% on the adjusted curve — price-only movement with the trade cash stripped out so the CF, EBAY, and FTI buys don't read as fake price gains.

The shape: down through late June, flat around June 26–30, then a step up in early July as markets recovered and the new positions started performing. The gold names were the drag the whole time.

Best of the period: META at +15.9%. Worst: AEM at −11.9%. The gold-vs-tech divergence in the same portfolio is worth sitting with. Both clusters pass the screener. They just don't move together.

One thing I'd do differently

Write the post on Friday whether or not I feel like it.

Three weeks of not writing didn't hurt the trading. But it created a gap in the reasoning record — three screener cycles where the signals ran and nothing got filed. If FTI's fundamentals have slipped, I don't have a clean week-by-week record of when the signal started changing. The posts aren't just for readers. They're the trail I use to audit my own decisions.

The summer excuse is real. The cost of it is also real.

What's next

The Barrick council check is the top action item — NMS 2.0 for two cycles is the rule I set, and now I have to run the play. ADBE is in its eighth week at Bedrock R1 with a closed entry window; if RSI pulls back, I want to be ready to move rather than watch another three weeks go by. And CF, EBAY, FTI all need a proper Layer 5 review — I bought all three on quantitative signals and haven't done the qualitative work yet.

The process ran fine while I was away. The writing is what fell behind.

— Mark