A portfolio isn’t a list. It’s a discipline.
Owning good stocks isn’t the same as having a good portfolio. Here’s how we think about the difference — and why doing nothing is often the right call.
We don’t optimize for the best stock. We optimize for your best next decision.
Most tools are built to answer one question: what’s the highest-ranked stock right now? That question ignores something important — what you already own.
Our process asks a different question: given your current portfolio, what single change — or deliberate decision to change nothing — most improves your expected outcome, once trading costs, added risk, and complexity are weighed against the benefit? We call this Expected Portfolio Improvement, and it’s the actual target our process is built around — not raw rank or raw expected return.
Sometimes the highest-ranked stock genuinely is worth buying. Often, the better decision is to leave a solid holding alone rather than trade into a marginally-better-ranked one — because a small, uncertain edge rarely outweighs the real cost of acting on it.
Not every holding should be managed the same way.
Our process treats two kinds of opportunity differently, on purpose.
Core
Chosen for durable business quality — the kind of company built to compound value over years, not weeks. These are held with real patience: a meaningful, sustained edge is required before we’d recommend replacing one, not just a slightly better-ranked alternative.
Satellite
Chosen for acceleration — businesses showing early signs of a shift, often before that shift is broadly recognized. These carry more inherent uncertainty, and our process treats them accordingly: more responsive to genuine deterioration, but never swapped out just because a new name briefly tops a daily list.
The two are never merged into a single score. Keeping them separate is what lets each one do its job — a blended, average approach would blur exactly the distinction that makes both useful.
Why around 20 holdings?
Twenty is a deliberate number, not an arbitrary one.
Fewer holdings — five or ten — means each position carries outsized weight. A single disappointing outcome does more damage to the whole portfolio than it should, and genuine diversification — owning businesses that don’t all move together — becomes hard to achieve with so few names.
More holdings — a hundred or more — solves the concentration problem, but creates a different one: quality dilution. Once a portfolio is that large, it inevitably includes marginal names included mainly to hit a target size, not because they cleared a genuine bar. At that scale, a portfolio starts to resemble an index fund with extra steps, without an index fund’s low cost or a concentrated portfolio’s ability to be genuinely selective.
Around 20 holdings is where those two pressures balance: enough names that no single disappointing outcome defines the whole portfolio, few enough that every holding still had to clear a real bar to earn its place. It’s also small enough that you can actually hold the whole portfolio in your head — meaningful transparency, not just a number on a page.
“No action” is a decision, not an absence of one.
If you use BSI for a while, you’ll notice something: most days, nothing changes. That’s not the system being quiet — it’s the system working as intended.
A process that recommends a change every time a ranking shifts isn’t more diligent. It’s more reactive. Our process is built to require a real, durable reason before recommending a change — not a marginal, possibly noisy difference between what you hold and what’s briefly ranked slightly higher today.
We tested trading on every ranking change. We chose not to.
Early in building this process, we tested a version that acted on every ranking change — buy the moment something ranks higher, sell the moment something slips. On paper, it looked reasonable. In practice, it meant well over one trade a week, every week, indefinitely.
We chose a different version instead: one that trades far less often, even though it means occasionally leaving some theoretical return on the table. We made that trade-off deliberately, for reasons that don’t show up in a backtest — trading costs compound in the real world in ways a spreadsheet doesn’t fully capture, and a process nobody can actually live with isn’t a good process, regardless of what it looks like on paper.
This is the same discipline behind everything else on this page: portfolio quality, considered patiently, beats chasing whatever ranks highest today.
See this discipline in your daily report
Start free and see exactly how this discipline shows up in what you receive every market day.