How the numbers work

Four rules decide what a figure includes, and almost every mismatch is one of them rather than an error.

1. THE PERIOD. A month view and a year view of the same category are supposed to differ. Check the selector at the top of each screen before anything else.
2. THE SCOPE. With business mode on, personal figures exclude business money entirely, and business figures exclude personal. The same category can hold both.
3. SPLITS. A transaction split across categories contributes only its SHARE to each one — a $400 receipt at 25% business moves the business figure by $100, not $400.
4. TRANSFERS. Money moved between your own accounts is neither income nor spending, so paying off a card doesn't appear as a purchase.

If all four match and the figures still disagree, that is worth reporting — include both screens and both numbers.

The date on the transaction — the one you can see and edit — not the date you entered it and not the date it cleared your bank.

This is why an imported charge can appear in last month after you import today, and it is the correct behavior: a dinner on the 31st belongs to that month's dining budget whatever day the bank got around to posting it.

If something is in the wrong month, edit its date. Everything downstream — budgets, reports, rollover, the trend chart — follows.

Income is money arriving from outside your own accounts. A paycheque, a refund, a gift, rent from a tenant.

Deliberately NOT income:
• a TRANSFER between two of your own accounts — moving savings to checking does not make you richer;
• a credit card PAYMENT — that is a transfer to a debt;
• the business side, when business mode is on and you are looking at personal figures.

Getting this right matters more than it looks: your savings rate, your health score and the cash-flow forecast are all built on it, so one transfer miscoded as income quietly flatters all three.

It shouldn't — pace is bill-aware.

A naive pace draws a straight line from the 1st to the 31st, which means rent leaving on the 1st reads as “600% of pace” every single month, and the reader learns to ignore the number. Instead, pace knows WHEN the money in a category is actually expected: a category whose recurring bill falls on the 1st is supposed to be near its limit on the 2nd.

So “over pace” means over what this category normally looks like at this point in the month — not over a straight line. One rule feeds the number, the tick mark on the bar and the curve, so they cannot disagree.

Assets minus liabilities, right now: account balances, real-estate equity (value minus what is owed on it), vehicles, other assets, and investment holdings at their latest price — less credit cards, loans and mortgages.

The trend chart is built from SNAPSHOTS — a stored figure with a date, so the past does not silently rewrite itself when today's prices move. That is why a chart point from March still says what March said.

One consequence worth knowing: a holding whose price hasn't refreshed carries its last known price into today's figure. It is stale rather than wrong, and refreshing prices is what settles it.

They answer different questions and are computed from different starting points.

• TODAY'S CHANGE measures from the PREVIOUS CLOSE. Before the market opens there is no session yet, so what you see belongs to the last completed one — the row says “at close” when that is what it is showing.
• TOTAL RETURN measures from what you PAID. It needs a cost basis; without one the row says “No cost basis” rather than showing zero.

And there are two units in play. Beside the price, a figure is PER SHARE. In the POSITION figure, it is the whole holding. “+$1.29” and “+$372” can be the same fact about 286 shares.

It walks forward day by day from your liquid balance — checking and cash, not investments — subtracting each active recurring bill on the day it is due and adding recurring income on the day it arrives. The lowest point that walk reaches is the number.

So it is only as good as your recurring items. A forecast that looks suspiciously healthy usually means a bill isn't in the app; one that looks alarming usually means income isn't.

It deliberately does NOT model your variable spending as a straight line. Guessing what you will spend on dining would make the figure look precise without making it true.

By WHO ADDED the transaction, not by whose money it was. Those are different questions and the app used to run one through the other.

Every row added since the app started recording authorship carries its author, and the feed, the split bar and the Mine / Theirs / Joint filter all read that.

Rows entered BEFORE that have no author at all, and they say “Joint”. That is honest rather than a guess — the information does not exist for those rows and inventing it would put words in somebody's mouth.

Three legitimate reasons before it is a fault:

1. KEPT PRIVATE. Anything either of you held back is not in the other's figures — it was never sent, so it cannot be counted.
2. FRESHNESS. A summary is published by the sender's device while the app is open. If they haven't opened it since yesterday, yesterday is the newest figure that exists anywhere.
3. WHEN YOU JOINED. Years tracked before you shared are yours unless you chose to bring the history across, and that choice is deliberately opt-in.

In a full-access household with nothing kept private and both apps recently open, the totals should match. If they don't, that one is worth sending in.

Because a credit-card balance is what you OWE, not what you have.

A $150 charge increases what you owe, so the balance rises and your net worth falls by $150. Paying the card is a transfer: money leaves checking, the card balance falls, and your net worth doesn't move at all — you swapped cash for a smaller debt.

That last part is why a card payment is not spending. The spending already happened on the day of the charge, and counting the payment too would count it twice.

As a transfer between the two accounts, not as an expense.

The purchases on the card are already in your budget, dated when you made them. If the payment were an expense as well, a $600 card payment would show up as $600 of spending on top of the $600 of purchases it settles.

This is the single most common way a budget total ends up looking twice as bad as the month really was.