Documentation

The Trades tab

Company ratings, explicit targets and a clear reason for every planned or skipped trade. Free on every plan.

Illustrative only. Check suggestions and tax estimates against your records. No broker orders are placed.

The simplified planner

Targets decide the split; one company rating orders the stocks. In Settings → Trades, choose Company ratings for stock selection alone, or Portfolio targets to maintain a mix of selected funds, crypto, individual stocks and cash. Existing allocation-monitoring targets do not activate this plan.

The rating is 0–100, with 50 neutral. Buys start with the highest-rated eligible companies; sales start with the lowest-rated holdings, after required company-limit or target reductions. Funds and crypto follow explicit targets and have no company rating.

Maximum per company is unset by default. Set it to limit a company’s share of the portfolio after the plan. The limit can reduce a buy or exclude it; known share classes count together. It does not apply to funds or crypto.

Each candidate keeps its rating and shows its plan result beside its rating drivers: In plan, Company limit reached, Capped by company limit, Order limit reached, or another short reason. Constraints change the order, not the rating.

Example: Invest 30,000 and Sell 15,000 plans up to 30,000 of purchases and 15,000 of sales. The 30,000 is new money; sale proceeds stay cash. Limits, targets and minimum orders can leave money unspent.

The decision ribbon

A compact view of the current plan and tax-lot timing. Open the related card for the orders or lots behind it.

The trade plan

One divider separates Buy and Sell. When the planner has a note on an order (for example “Capped by holding size”), an ⓘ sits right after that order’s amount; point at it, or tap it on a phone, to read the note at the top of the row’s tooltip. Orders without a note have no ⓘ, and “In plan” is never repeated.

Hover any asset on this tab — a plan order’s ticker, a Recent trades row’s asset, a candidate’s ticker or name — for its asset card: the week’s mini chart, the price and today’s and the week’s move in the asset’s own trading currency, and, when you hold it, the position’s value and share of your portfolio. The card opens only on the asset itself — its logo, ticker or company name, which underline while you point at them — never on the empty rest of the row, and a small spinner turns in its chart’s place while the week is still loading. Clicking still opens the asset.

Every order in the plan, and every row of the Buy and Sell candidate tables below, starts with the asset’s logo left of its ticker — the same marks as in Holdings and Recent trades — so a company is found at a glance; a listing without a logo shows its company’s initials instead.

Each list labels its columns: Amount, Share and Rating on both sides, and Tax-free now as well on the Sell side. Amount and Tax-free now name your display currency in the label, for example “Amount [CZK]”, so each row prints just the number. Rating is the company rating the plan ranked by, rounded and coloured as in the candidate tables: green from 50, red below, a dash when the company has no rating. Tax-free now is the most of that holding you could sell today without income tax (see Reading the table). When nothing has matured yet, the cell says when the next lot does — “matures Nov 2027” — with the exact date and the quantity behind it in the row’s tooltip. A dash is left only where there is nothing to date: no tracked lots, matured lots with no current price, or lots the configured rules apply no time test to. The ⓘ next to the Tax-free now label carries the rule. Buy and Sell sit side by side for as long as both lists fit one line per order; as the window narrows Buy gives up its spare width first, and only when both no longer fit does Sell move under Buy instead of wrapping. Only on a phone, too narrow for five columns at all, does Tax-free now take its own line under the row.

Hover a row for its one tooltip: the planner’s note, what this order means for tax (estimated tax, and how much of it is tax-free or taxable) and, on a Sell row, the holding’s Tax-free now sentence. Like every explaining tooltip it carries an Ask AI button, and the pointer can move onto it. The ticker keeps the asset card. The two plan buttons explain themselves the same way, each with its own Ask AI: Simulate tries the whole plan in the What if I trade? card, and Run historical backtest opens the Simulation tab and replays your saved plan settings over the past year of market data, rebuilding the plan at each trading interval from what was known that day, and compares the result with doing nothing (keeping your starting holdings) and with what your portfolio actually did.

  1. Set the money. Invest is a new contribution, added once to the portfolio value used for limits and targets. Sell is independent; its proceeds stay cash.
  2. Choose the split. Targets mode reserves the deficits of explicit fund/crypto targets and the individual-stock bucket, using the post-contribution portfolio value. Cash and unfillable buckets stay cash.
  3. Choose eligible orders. Buy list, data coverage, company capacity, available price, order count and minimum size all apply.
  4. Size stocks. Split purchases proportionally to each rating and sales to 100 minus the rating, with a minimum weight of 1. Company capacity and minimum orders still apply.
  5. Plan sales. Sell 0 disables sales. Otherwise, company-limit and target reductions take priority, followed by the lowest ratings, within the sell budget and holdings.

Tax estimates and lot coverage are informational; they do not block or postpone sales. Sale-selected assets, including known share classes, are excluded from buys: no asset is bought and sold in the same plan.

Each order shows its amount and any cap. Unspent money and fewer orders are explained. These are suggestions; the app does not place broker orders.

The card’s heading holds its actions at the right: Simulate, Run historical backtest and Edit settings. Simulate hands the whole plan to the Simulation tab’s What if I trade? card — every buy and sell row with its amount, one line each — and shows their combined effect, sells applied first so their cash funds the buys. Nothing is traded. Its last line sums the plan up — what is bought and sold and, when you sell, how much of it your lots cover tax-free — with Copy plan at the right, which puts the whole plan on the clipboard as plain text.

Backtest the plan with historical data.

Net amount

Net amount is purchases minus sales; it does not measure new money. Company ratings are never reduced by tax or an opposing sell score.

Recent trades

Directly under the trade plan, the Recent trades card lists your executed buys and sells across all accounts, newest first. Each row shows the asset, a Buy or Sell chip (green for a buy, red for a sale), the price in the trade’s own currency, Since trade, the units, the amount in your display currency, the broker’s fee where the export reports one, the date and the account. Click a row to open the asset’s details — for an asset you have since sold out of, too, with its trades still marked on the charts.

The amount is what the trade cost or raised, converted at that day’s exchange rate and including fees where the broker reports them. Sales are money coming in, shown in green with a “+”; purchases are money going out, in black (white in dark mode). Since trade is how the asset’s price has moved from that execution to today, in the asset’s own currency, so exchange rates do not affect it. Later splits are accounted for. For a sale it shows what the sold units would have done since. A dash means there is no current price.

The card shows 10 rows a page. The ⚙ in its corner sets 10, 15, 20, 50 or 100, and this browser remembers the choice. The ‹ and › buttons under the table turn the pages, as in the Explore tables, and the card’s title shows which trades you are looking at, for example “11–20 of 1396”. A new page size starts again from the newest trades. Broker exports record the date of a trade but not the time. Transfers between your own accounts, ledger corrections, staking credits and corporate actions are not trades, so they are left out.

Sorting. Click any column header to sort by it, and click it again to reverse the order. Figures start with the biggest and words run alphabetically; the table starts newest first. The sort covers your whole trade history, not just the page on screen: sorted by Amount, the first row is your biggest trade ever. Trades with no value in the column, such as no fee or no current price, stay at the end either way.

Filters. The filters are folded away under Filters › next to the card’s title. Click it to open them. They are the same filters as the Trade recommendations tables: a box for a name or ticker (type an idea, like “dividend payers”, and when no name or ticker matches it the assistant picks the matching assets; quotes ask the assistant straight away), the Type chips, a Date range from one day to another (both days included), and under More filters buys or sells only and one or more accounts. Like the sort, they apply to your whole history. The title and the page buttons then count only the matching trades, a line under the title says how many match, and Clear filters removes them all. While the filters are folded, the label shows how many are on, for example “Filters › · 2 active”. Changing the sort or a filter goes back to the first page. The sort and the filters stay when you switch to another tab and come back, and when the page reloads; a new browser tab starts from the newest trade, unfiltered.

Money flow

Under Recent trades, the Money flow chart shows how much went into purchases and how much came back from sales in each week, month or year. Use the Weekly | Monthly | Yearly switch to choose. The chart covers the last 26 weeks (starting on Mondays), 24 months or 10 calendar years (the current year counts up to today); the ⚙ in the card’s corner sets 12, 26, 52 or 104 weeks, 6, 12, 24, 36 or 60 months, or 3, 5, 10 or 20 years, and this browser remembers the choice for each. Amounts are in your display currency, converted at each trade’s own exchange rate. A period without trades shows as zero.

Sales rise above the zero line in green: cash inflows, what your sales brought in. Purchases hang below it in black (white in dark mode): cash outflows into positions, what your buys cost. Both are the broker’s cash amounts, with fees included where the broker reports them. The thin line with dots is the net: sales minus purchases, above zero when the period sold more than it bought. Deposits, withdrawals, transfers and dividends are not counted. Hover a legend key for its definition, or a period to see its sales, purchases, net and number of trades.

Allocation monitoring

Asset, sector and class targets show how your holdings compare with your chosen mix. These monitoring targets are separate from the planner’s explicit fund, crypto, stock and cash targets.

Rated candidates

Companies share one rating across buy and sell views. Each row separates the rating’s drivers from the reason it was included, omitted or capped. Refreshes use current data and show progress.

Buy candidates

Choose Owned + favourites, Owned only, or Favourites only, optionally limited to one watchlist. The choice is the buy table itself: only those companies are rated and listed under Buy candidates. The table’s title names it in parentheses — Buy candidates (all favourites) — so you can always see which universe you are looking at. Under Favourites only a company you own but have not starred gets no row and no rating — it is not a purchase candidate. Holdings remain the full Sell candidates list in every mode. Explicit fund/crypto targets are a separate buy source; the planner does not add example ETFs on its own.

Reading the table

Rating is the company opinion; Owned is your current position; Tax-free now is the most of that holding you could sell today without income tax — the open lots already past the holding-period time test, at the current price. It is the same figure the Taxes tab shows as Tax-free value, and the trade plan’s Sell rows print it too. A dash means no lot has matured, or the asset has no tracked lots — on the trade plan’s Sell rows a holding with nothing matured yet says when its next lot does instead. The year’s small-sale exemption is a separate limit on total proceeds and lives on the Taxes tab. Illustrative, not tax advice. The leading criteria explain the rating. Plan reasons include Order limit reached, Insufficient data, Below minimum order and No budget. A fund/crypto target says Target allocation.

Hovering the ticker or name underlines it and opens the asset card, which names the company in full — and nothing else pops up beside it; on a narrow screen, where the name is hidden, the row’s breakdown is headed by it. Click anywhere on a row for its criterion values — except its ticker or company name, which open that asset’s own details instead, the same panel the heatmap, the search and your holdings open.

Opening a row. An opened row shows, above its table, the company’s earnings stage and Where the N points come from: a bar with one segment per criterion, as wide as the points it adds to the rating, and the points it missed left bare. The table lists Criterion, Raw (the figure the criterion scores), Score (0–100), Criterion weight (the weight you set in Criterion weights, or the part of it this company uses, such as “25 of 40”) and Total points, which add up to the rating.

One tip at a time. Point at a criterion’s name for what it measures and how its 0–100 scale runs — the same explanation the Criterion weights sliders give. Point at a value for this company’s own reading only (for Fair value, for example, “+18% vs fair value · 5 models”), and at a column header for what the column holds. Tabbing to a criterion’s name opens its tip from the keyboard.

The headers explain themselves. Hover any column label — Asset, Rating, Owned, Tax-free now — and it says what the column holds and that clicking it sorts the table by that column. Click again to reverse the direction (Rating excepted — see below); the sorted header shows ▲ or ▼, and screen readers hear the same through aria-sort. Rows with nothing to compare — a candidate you do not own, or one the lot engine has no answer for — sink to the bottom in both directions rather than reading as zeroes. Drivers & plan explains itself too, and is the one column that does not sort.

Each table shows its top ten — the first ten rows of whichever column you sorted by, and nothing else. show all N opens the full list, show top 10 folds it back. A candidate the trade plan funded is marked In plan on its own row; being in the plan never adds a row from below the tenth. Rating sorts one way only — the way its list means: Buy candidates best-rated first, Sell candidates weakest-rated first. Clicking it again leaves the table as it is. Asset, Owned and Tax-free now sort both ways.

How the rating works

The rating is the weighted mean of available, positively weighted inputs. An actionable company rating needs at least three inputs, including Quality and either Valuation or Fair value (a question judged on both its readings counts once). Missing inputs are shown, never counted as zero or as 50 — they are left out of the mean. All weights off means no rating. Only individual stocks are rated; funds and crypto get no company rating — and one of the curated gold ETCs the market data lists as a share (Xetra-Gold, SGLD, IGLN…) counts as a fund here, its row typed Commodities like a gold ETF’s.

Every criterion is a score from 0 to 100, where 50 is neutral. Each figure below maps onto that scale on a straight line between the two ends shown and is clamped beyond them. Hover a criterion in Criterion weights for the same explanation as a formatted tooltip.

CriterionMeaning
QualityHow strong the business is. The average of four reported figures: return on equity (0% or less → 0, 30% or more → 100), profit margin (0% → 0, 25% → 100), net debt ÷ EBITDA (4× or more → 0, no net debt or net cash → 100) and revenue growth (0% → 0, 25% → 100). A figure that is not reported is skipped; with none, Quality is left out. The Raw column shows the return on equity.
Quality · growth readingHow strong a business is that its profits do not explain yet. The average of the Rule of 40 — revenue growth % + free-cash-flow margin % (0 or less → 0, 40 or more → 100) — gross margin (20% → 0, 70% → 100) and cash runway, cash ÷ yearly cash burn (1 year → 0, 4 years or self-funding → 100). Growth above 100% counts as 100%. Takes the growth share of the Quality weight (earnings stage); with none of the three figures, the profit reading takes the whole weight. The Raw column shows the Rule of 40 as a percentage.
ValuationHow cheap the share is against its earnings and cash. The average of forward P/E (40 or more → 0, 10 or less → 100), PEG — forward P/E ÷ earnings growth in % (3 or more → 0, 0.5 or less → 100) — and free-cash-flow yield (0% or less → 0, 8% or more → 100). An expected loss and a PEG without positive growth are skipped. The Raw column shows the forward P/E (a dash when a loss is expected).
Valuation · sales readingHow the price compares with sales and their growth. The average of EV/Sales — market cap plus debt, less cash, ÷ yearly sales (30× or more → 0, 2× or less → 100) — and EV/Sales per point of growth (1.0 or more → 0, 0.2 or less → 100; shrinking sales → 0). Takes the growth share of the Valuation weight; without the figures, the earnings reading takes the whole weight.
Fair valueWhat textbook valuation models say the share is worth, against its price. Six models (DCF, Graham, Lynch, EV/EBITDA, dividend discount, sector P/E) are averaged with fixed weights, DCF the heaviest; a model that cannot run, or lands over 2.5× off the models’ median, is left out. Model value ÷ price − 1: −20% or less → 0, 0% → 50, +20% or more → 100. Not used in the historical backtest.
Analyst upsideWhere analysts’ price targets sit against today’s price. Mean analyst price target ÷ price − 1: −30% or less → 0, 0% → 50, +30% or more → 100. Left out without a published target. Off by default; not used in the historical backtest.
MomentumThe 3–6 month price trend: buy what is rising. The 3- and 6-month price returns, averaged (dividends excluded; one window alone when the history is shorter): −20% or worse → 0, unchanged → 50, +20% or better → 100. A rising price lifts the rating and a falling one lowers it — trend following, the mirror image of Drawdown. Left out with under three months of history. Measured in the currency the asset trades in — an exchange-rate move is not part of it.
DrawdownThe dip from the 52-week high: buy what got cheaper. Today’s close against the highest close of the past year: at the high → 50, 35% or more below → 100. It never goes below 50 — buying the dip, the mirror image of Momentum. Pair it with Quality (a good business on sale) rather than with Momentum: the two read the same price in opposite directions and largely cancel. Needs at least five closes in the year. Measured in the currency the asset trades in. Off by default.
DividendThe dividend yield, annual dividend ÷ share price: 0% → 50, 5% or more → 100. It never goes below 50, so it can only favour payers; with no reported yield it is left out. Off by default.
Group consensusHow many of the funds you follow hold it. The share of your followed Super Investors whose latest 13F filing includes the company: none → 50, all → 100. 13F filings are delayed (up to 45 days after each quarter). Needs Pro, Social and a followed fund with filings. Off by default; not used in the historical backtest.

Criterion weights

Open the weights above the candidates; the breakdown’s Criterion weight column prints what they set. A weight is how much a criterion counts, relative to the others: 40 counts twice as much as 20, and 0 leaves a criterion out. A weight is not a score: a low weight is not “below neutral”, it only counts less. The rating is the average of the criteria’s 0–100 scores (50 is neutral), each counted by its weight. Defaults: Quality 40, Valuation 20, Fair value 20, Momentum 20; other criteria 0. The Quality and Valuation weights also cover their growth readings (earnings stage). Apply changes or restore defaults.

Early-stage companies: the earnings stage

A company that is not profitable yet cannot score on profit questions — return on equity, margins, P/E, free-cash-flow yield, the fair-value models. So each company gets an earnings stage: how much of its price its profits already explain, 40 ÷ P/E, using the lower of last year’s and next year’s P/E, capped at 100%. At a P/E of 40 or less it is judged on profits only, exactly as before; at a P/E of 80, Quality and Valuation split half-and-half between their two readings and Fair value counts half; with no profit last year or next, the growth readings only, and no Fair value. The bar above an opened row shows the split, and its tooltip the arithmetic.

The growth part uses two readings in place of Quality and Valuation: the Rule of 40 (a common benchmark for software companies), gross margin and cash runway (see burn rate); and enterprise value ÷ sales (sales multiple), alone and per point of revenue growth. The stage only splits each weight between its two readings and scales Fair value; it never adds weight. When one reading of a question has no data, the other takes the whole weight. Financial-services and real-estate companies stay on profits, and so does a company without comparable sales: under $50M a year, sales in a currency that could not be converted, or a pence-quoted London listing. Momentum and the other market signals keep their weights at every stage (with Fair value out, their share of a growth company’s rating is larger). Not used in the historical backtest, which has no archived P/E.

Filters

The filters are folded away under Filters › next to the card’s title, as on the Recent trades card. Click it to open them. Search names or tickers and filter by type, size or sector. Filters narrow the displayed rows; they do not change the saved plan. While the filters are folded, the label shows how many are on, for example “Filters › · 2 active”, and the line above the tables still says how many candidates are shown, with Clear filters. The filters and each table’s sort stay when you switch to another tab and come back, and when the page reloads (saving a setting reloads it); a new browser tab starts from the defaults.

Recommendation history

The Recommendation history card has been removed from the Trades tab for now. Your recommendations are still recorded once a day in the background, so the history will be complete if the card returns.

Trading activity

Expand the activity card for your executed buys and sells by month.

Using the assistant

Ask “Why was Delta skipped?”, “Limit each company to 10%”, or “Remove the company limit”. The assistant reads the same rating, decision reason and amount, and can change planner settings and targets. Assistant actions.

Group consensus

Available with Pro and followed fund filings. These delayed holdings are an optional preference input, not a forecast; missing filings leave the input unavailable.

Large exact optimizations

Exact tax optimization has workload limits. A refusal names the limit; choose FIFO explicitly or reduce the calculation. Spreadsheet and PDF imports likewise report processing limits.

Expanded buy and sell candidate lists keep show top 10 at the bottom of the visible list, above phone controls. Collapsing returns to that list’s start.