In the last essay I introduced Joe: 55, $117,000 saved, needs $50,000 a year beyond Social Security and a small pension. Traditional math told him to work into his seventies. The different equation — Anchor & Ladder — got him there in ten years instead of fourteen to sixteen, funded by dividends instead of by selling shares.
That essay made a promise at the end: the mechanics. What actually qualifies a stock to earn a place in the portfolio? When do you buy? When do you sell? What do you do when a company cuts its dividend?
This is the answer. It's the entire stock-selection method, and it fits on an index card.
The buy rule: two gates, both required
Every stock purchase has to pass two tests. Not one. Both.
Gate one: buy it cheap. You want the stock trading at a meaningful discount — at least 20% below a fair estimate of what the business is actually worth. Morningstar publishes fair value estimates for free; that's a reasonable starting reference. Buying dollar bills for eighty cents is what makes a position safe and profitable. If there's no discount, there's no margin for error, and you pass.
Gate two: the future income has to justify the homework. Take the stock's current yield and project it five years out: current yield × (1 + annual dividend growth)⁵. A 4% yield growing at 8% a year becomes roughly a 5.4% yield in five years. That five-year number has to beat what you'd get from a plain dividend index fund with zero research. If it doesn't beat the no-homework alternative, buy the index fund instead. No shame in it — that's the alternative doing its job.
In plain terms: buy it cheap, and only if the future income earns the effort. If either gate fails, walk away. There is always another candidate.
Why both? Because each gate guards against a different expensive mistake. Gate one without gate two gets you cheap stocks whose dividends never grow — value traps that pay you to wait for a payday that never comes. Gate two without gate one gets you wonderful growers bought at prices that guarantee mediocre returns. The two gates together are the whole trick.
Size by conviction, and keep the roster closed
Not every stock that passes deserves the same dollars. I run three tiers:
- 8% position size — highest conviction: an exceptional business bought at a clear discount. The core of the portfolio.
- 4% — solid growers, good businesses at fair prices.
- 2% — promising but early: you're still building conviction, or the story is still proving itself.
(Percentages are of the total portfolio.) The tiers do two jobs at once: they concentrate your money in your best ideas — concentration in high-conviction names is a feature, not a bug — and they cap how much damage any single company can do. No stock, however beloved, exceeds its tier.
And one more rule that prevents the most common beginner disease: the roster stays closed. A new company has to beat a current holding to get in. If it can't displace something you own, it doesn't belong. Without this rule, portfolios drift toward forty mediocre positions and the average quietly becomes the enemy.
Why individual stocks instead of just a fund
Fair question — a dividend-growth fund needs no research at all. Three reasons I own the companies directly:
- You keep every raise. A fund's payout is an average — your best growers' raises get diluted by the laggards. Own the company, and every raise flows to you in full.
- No fee drag. Fund managers charge every year, forever. Your broker charges you once, at purchase.
- You learn the businesses. Following a dozen companies you own teaches you more about how the economy works than any fact sheet.
The honest tradeoff: homework and single-company risk. A company can cut its dividend — funds rarely do outright. So you select against that risk before you buy: a record of raises through rough patches, management that treats the dividend as a promise rather than a leftover. You don't hope the cut won't happen. You buy companies where it would be out of character.
The sell discipline: the raises are the thesis
You own a dividend stock for one reason: the raises. So the sell rule writes itself.
If the raises stop, the thesis is broken.
A dividend cut is an automatic sale — no debate, no waiting to see, no hoping it recovers. A multi-year freeze in a company you bought for growth is a sale candidate. A grower whose raises shrink from double digits to token increases is telling you the story changed, even if the payout technically went up. It may not be broken — but it's no longer doing the job you hired it for.
And here's the part beginners get backwards: a falling stock price is not a sell signal. You sell on dividend behavior, not price behavior. A price drop on an intact dividend is the market offering you a richer yield on new dollars — that's an opportunity, not an emergency. This discipline is what makes the method work over decades: every holding is either growing its payout or it's gone.
What it looks like over time
$10,000 in a single holding: 4% starting yield ($400 a year), 8% annual dividend growth, no new money added:
| Year | Annual income | Yield on your $10,000 |
|---|---|---|
| 1 | $400 | 4.0% |
| 5 | $544 | 5.4% |
| 10 | $800 | 8.0% |
| 15 | $1,175 | 11.7% |
| 20 | $1,732 | 17.3% |
By year ten the income has doubled and the yield on your original money is 8%. By year twenty it's paying 17% a year on what you put in — and still accelerating. Now imagine eight to twelve such holdings, in tiers, each raising most years. That's the ladder working.
And remember the anchors from the last essay — the high-yield half (a blend like SCHD for dividend growth plus SPYI/QQQI for income, roughly 8.1% starting yield). The anchors fund the early years so the ladders never have to be sold before their time. The anchor holds; the ladder works.
Back to Joe
So what does Joe actually do? He starts with the anchors — the income floor, paying from month one. Then, one holding at a time, he runs the buy rule: cheap enough? Five-year income beats the index fund? Pass both, size it by conviction, add it to the closed roster. Every year he checks one thing per holding: did the raise come? Cut means sold, freeze means on notice, price drop means nothing.
Same $117,000, same $30,000 a year going in. Ten years to $50,000 of annual dividend income — because every dollar was hired for a job and the ones that stopped working got fired.
The method, written down
This is the complete stock-selection method: two gates to buy, three tiers to size, one rule to sell, a closed roster to keep it honest. It fits on an index card because it has to — a method you can't explain simply is a method you won't follow for twenty years.
The full system — anchors, ladders, the buy rule, account placement, the monitoring routine — is what I teach step by step in Anchor & Ladder™. If you want the whole method rather than the essays, join the waitlist and you'll hear about it first.
Next in Pillar 2: why I run options trading like a business — and what "like a business" actually means when the business is you.