Up 3.1% on a rotation week — three sells, eight buys

A rotation week, not a quiet one

Last week I wrote about three sells and four buys and called it a busy one. This week was busier. Three more names came off the books and eight new lines went on, with a couple of top-ups in the middle. The portfolio finished the week up +3.14% — the same headline number two weeks running, but the work underneath was different. Two weeks ago I was clearing names that had quietly broken. This week I was building.

How a stock ends up on my "sell" list — again

I keep a watchlist inside my own dashboard, and every stock I own carries one of three labels — Hold, Watch, or Sell. A stock gets the Sell label when it slips below a set of minimum quality rules I use to decide what's worth owning: is the company actually profitable, is the debt getting out of hand, does the price still make sense for what you get. Once the Sell label is set, the decision is made. The only question left is when to push the button.

The "why" of a sell has to be one of three things:

1. The company has fundamentally changed for the worse.
2. I was wrong to buy it in the first place.
3. I've found something clearly better for the cash.

All three of this week's sells were reason number three. I had names I wanted to own more than the ones I was holding, and the cash had to come from somewhere.

What I sold

These three names had been sitting on the Sell list for weeks. Tuesday afternoon I cleared them out:

• PBH — Premium Brands, a Canadian specialty foods company. The position was small and the conviction had been thinner than what was on offer in the screener.
• SJ — Stella-Jones, a wood-products company. I trimmed about a quarter of the position rather than exit it — I still like the business, just less than where the cash ended up.
• PPC — Pilgrim's Pride, a chicken producer. Cyclical, smaller than most of what I hold, and the screener wasn't ranking it anywhere I could justify the slot.

The exits closed Tuesday at PBH $58.61, SJ $82.57, PPC $32.82.

What I bought, and how I found them

Once a week — on Saturday morning, never Friday close — I run a stock screener: a tool that takes the thousands of stocks listed in the U.S. and Canada and filters them down to a short list of names that pass a set of quality and value rules I've written. The one I rely on most is the one I call the Bedrock — it surfaces large, durable businesses I'd happily own without checking every quarter. Names that show up on Bedrock are usually already familiar to most people: brands, networks, the kind of business that's hard to copy.

This week Bedrock surfaced five names I'd been waiting for at the right price. Tuesday afternoon I took them, then added two more on Wednesday morning:

These are not exciting names. They are boring in the specific way I want a portfolio to be boring — high-quality businesses that compound quietly. MA runs a payment network most of the world has to use; COST sells warehouse memberships that renew like clockwork; ADP runs payroll for a third of the U.S. workforce; DIS owns more recognised stories than any other company on the planet; CRM is the customer database most of the Fortune 500 runs on. The screener doesn't pick them because they're famous. It picks them because they keep earning more money every year and the market isn't asking too much for the shares.

I also topped up two existing positions on Tuesday afternoon:

• NVDA — added 2 shares at $201.29. Already a Bedrock name; this was just sizing it closer to where I want it.
• INTU — added a full share at $412.80, on top of a tiny fractional drip I had running.

And one small gold add on Tuesday — about GOLD.WS 0.15 units at the unit price the market quoted, which is a chunky one. Gold ran hard in March and I'm letting the position size up slowly, not all at once.

What moved this week

The portfolio's two big stories had nothing to do with my trades.

AMD is the kind of weekly winner you can't plan around: the position is too small to matter to the portfolio's headline number even at a 25% move. LMT is the more useful read — a single-week 13% drop in a defence name is the market saying something about the next earnings cycle, not just this one. I'm not selling. I am paying attention.

How the portfolio did

Up +3.14% on the week, Friday-close to Friday-close. The line walked up Monday through Wednesday on the buys landing into a rising market, gave a little back Thursday, and finished Friday near the week's high. Eight new positions; three closed; two topped up. Not a quiet week.

What I'm watching

Whether the Bedrock adds (DIS, MA, CRM, COST, ADP) earn their keep over the next month. These are not trades — they're positions I intend to hold for years — but the next 30 days will tell me whether the entry prices were honest or whether the screener was a step late.

Whether LMT's drop is one-week noise or the start of a sector rotation out of defence. The position is too small to hurt the portfolio, but it's a signal worth tracking either way.

Three sells. Eight buys. One quiet thesis.

The thesis behind a week like this isn't loud. It's just: when the screener finally lets a name through that you've been waiting for, you buy it. And when the cash has to come from somewhere, you take it from the names you'd already decided to part with. Most of the work of investing isn't picking what to buy. It's clearing the bench so the buy fits.

— Mark

— Mark