Target

A target is the portfolio your blend produces: a list of holdings and weights, plus a cash reserve, plus the names your rules excluded and why.

It is not a recommendation from a black box. It's the arithmetic result of choices you made — which managers you trust, how much, and under what constraints — and every line can be opened to see whose conviction put it there and at what weight.

Explainable by construction

Expand any holding in the target and you see its provenance: which sources contributed it, how much each contributed, and which rules were applied on the way. A name held by three of your sources shows all three.

This matters because the alternative is the thing that makes people distrust automated allocation: a number appears, and you have no way to interrogate it. Here, "why does the plan hold 9.5% of this?" always has an answer, and the answer is made of your own decisions.

The excluded names are shown too, with their reasons in your words — "no international brokerage — permanent exclusion", "cash ballast, not an equity position". A plan that hides what it dropped isn't explainable.

Live preview versus saved snapshot

The target has two states, and the app is explicit about which you're looking at:

• Live — regenerated from your current blend every time you touch it. Free to experiment with; nothing is stored.
• Saved snapshot — a version you deliberately kept, timestamped and named. This is what you reconcile an account against, so the comparison is against a plan that held still, not one shifting under you while you trade.

Effective bets: how many bets are you really making?

A book with 28 positions is not necessarily 28 bets. If eight of those names all rise and fall with the same underlying story, that's closer to one bet in eight pieces.

Effective bets counts your positions after collapsing the names that move together, so you can see concentration you'd otherwise miss. Two details are worth knowing:

• The grouping comes from correlation groups you declare — your own judgment about what constitutes one bet — not from a statistical correlation the system measured. It reports what you told it, which means you can disagree with it by editing the groups.
• It's size-weighted, so a large position in a crowded theme moves the number more than a small one.

Seeing "28 names, 23 bets, largest bet 29.8%" tells you something a position count never will.