What to do with surplus money each month, as a rule your forecast follows

Cash routing is a rule the forecast engine applies as it walks forward through time. Every month, after your income lands and your bills are paid, it looks at what is left in your current account and does something deliberate with it instead of letting it pile up. It is the newest capability in the app and it sits alongside the rest of what CrestCast forecasts.

Three stages run every month, always in the same order. The diagram below is a picture of how the rules run, not a screenshot of the app.

Each waterfall step can point at a savings or investment pot, at a debt as an overpayment, or at another cash account. Each step can carry a limit: an annual allowance bucketed from the 6 April UK tax year start, a target balance to stop at, or a fixed amount per month.

On an ISA step the allowance is filled in for you, and it is the right one: £4,000 for a Lifetime ISA, £9,000 for a Junior ISA, £20,000 for an ordinary one. A cash ISA follows the published schedule rather than one flat figure, so a forecast that runs past 6 April 2027 already steps down to £12,000 for under-65s, while anyone aged 65 or over keeps the full £20,000. You can still type your own number over any of it.

One step, in the app

Figures shown are the sample profile, not a real household.

This used to be the least discoverable thing in the app: one row in Settings labelled Surplus and savings rules, a name that used neither of the words anyone went looking for. It is now called Spare cash waterfall and it appears in two places: inside the Cash balance card on the input screen, and still in Settings. The Cash balance card is collapsed by default, so on the input screen it is one tap in rather than sitting out on the landing. Wherever you meet it, the card does not just carry a label: it reads back the rule you have actually saved, in the order your own money goes in.

Each account keeps its own waterfall, so the rules you write while looking at the household fill the household's pots and pay down the household's debts, and a personal account keeps a separate set. Where a profile has more than one account, pills at the top of the screen say which one you are editing. That matters more than it sounds: if your savings and much of your debt are joint, a rule written against your personal account has nothing to act on.

Where it lives, and what it says when you get there

Figures shown are the sample profile, not a real household.

It is opt-in, and switching it off is a genuine no-op. With routing off, the forecast behaves exactly as it did before the feature existed. Nothing quietly changes underneath you, and nothing gets moved around because the app thought it knew better.

The interesting part of a savings rule is not the good months. It is what happens when the money is not there, and this is where most homemade versions fall over.

• Short months do not depend on data entry order. When contributions have to be trimmed, they are trimmed proportionally across every pot. A spreadsheet that pays pots top to bottom gives a different answer depending on which row you happened to type first.
• A skipped contribution stays skipped. It is not silently doubled up next month, which would flatter the forecast and tell you the year was fine when it was not.
• Allowances reset on 6 April, not in January. A step with an annual allowance buckets against the UK tax year, so a rule set in February gets its full allowance back two months later rather than eleven.
• A pot that reaches its target stops taking money. The overflow continues down the waterfall to the next step instead of stopping there.
• Going below the buffer is handled, not ignored. Drawdown pulls from the pots you nominated, in your order, rather than leaving an impossible negative balance sitting in the forecast.

Say the household takes home £5,400 a month and the bills, the mortgage and everything else come to £4,300. The minimum is set at £1,500 and the current account target balance at £2,500. The waterfall is three steps: the stocks and shares ISA up to a £20,000 annual allowance, then the credit card as an overpayment, then the easy access savings account up to a £15,000 target.

In a normal month, roughly £1,100 accumulates. Once the current account climbs past £2,500, the excess starts going to the ISA. Say the card is sitting at £4,200 on a rate that hurts, and you reorder the waterfall to put the card first: the sweep now lands on the card each month, the balance falls, the interest charged next month falls with it, and the month after that there is slightly more surplus to sweep. That feedback is the whole point.

Then a £2,900 boiler replacement lands in November. Contributions that month are trimmed back to the buffer, cash still ends short, and drawdown pulls the shortfall out of the easy access pot rather than the ISA, because that is the order you set. December carries on as normal. The forecast shows you all of that before you decide whether the rule is one you want.

A routed overpayment is not a cosmetic transfer between two boxes. Mortgages and loans are genuinely re-amortised with the routed money, so the balance, the interest and the payoff date all move. Credit cards take the routed amount as extra payments on top of the scheduled one.

Student loans are deliberately left out of the overpayment options. UK student loan repayment is income-contingent: you repay a percentage of income above a threshold until the balance clears or the loan is written off, so throwing spare cash at it does not reduce what you pay each month, and for many borrowers it never comes back at all. Excluding it from a surplus waterfall is the correct default.

Because routing is computed inside the forecast itself rather than bolted on afterwards, the cashflow, profit and loss and net worth views all reflect it. The cashflow view goes one step further on the savings side: on the cash balance breakdown, with savings included, the routed money gets its own band reading Swept to savings in the months it goes out and Drawn from savings in the months it comes back. The money routed onto debt has no band of its own, so you read that side off the falling balances rather than off a line naming it.

A blank waterfall is a chore to fill in, so there are four one-tap presets. They are not generic templates: each one is assembled from the actual accounts, pots and debts already in your profile.

• Safety first. A £2,000 minimum held back, then everything above it goes to savings.
• Max ISA. A £1,000 minimum, then the ISA filled to its allowance, with the overflow running on into ordinary savings.
• Clear debt first. A £1,000 minimum, then the surplus at your credit cards, priciest first, with any loans beneath them.
• Balanced. A £1,000 minimum, then the same ISA-then-savings order as Max ISA. It is not a split between saving and debt, and as the app currently builds them the two produce the same rule, so treat it as a second name for the same starting point rather than a different strategy.

The four presets

Figures shown are the sample profile, not a real household.

A preset is a starting point that you then edit. Reordering the steps, changing a limit or removing a step entirely takes seconds, and the forecast redraws against the new order.

Plenty of what CrestCast does could be reproduced in a spreadsheet by someone patient. This one genuinely fights back, because it is a forward pass with feedback. Each month's sweep changes the debt balance. The changed balance changes next month's interest. The changed interest changes next month's surplus. The changed surplus changes the next sweep. Every month depends on the one before it and feeds the one after it.

You can build that. It takes either a circular reference with iterative calculation switched on, or a full month-by-month grid where each row rebuilds the amortisation from the previous row. Both work, and both break the first time you insert a row, add a fourth waterfall step, or copy the sheet to try a different order. Then you are debugging a formula chain instead of answering the question you sat down with.

There is a second thing a spreadsheet will not do, and it matters more than the feedback loop. It has no view about what you would do when the money runs short, so it will happily project a current account sinking further into the red every month for a decade while a savings account sits untouched beside it. One household, run twice with the waterfall off and on shows what that difference is worth, and what the held balance costs.

The question was never hard. It was, if I put the spare £600 at the card instead of the ISA, where am I in three years. See also overpay the mortgage or invest, which is the same shape of question with a longer time horizon.

Four edges worth knowing before you set the rule up, so the forecast does exactly what you expect.

• It is not a snowball or avalanche engine. The Clear debt first preset sorts your credit cards by rate. It does not rate-sort loans and it leaves mortgages out entirely. CrestCast has neither strategy as a feature. For the ordering itself, the debt payoff planner page is the better read.
• It caps, but it does not warn. The waterfall will not route more into a pot than its annual allowance takes, and it knows the ISA limits by pot type and by tax year. But nothing raises a flag to say a plan would have over-subscribed one, and a contribution you set directly on a pot sits outside the waterfall and is not capped by it.
• It does not show you how much of a pot is routed money. The balances themselves are right: an asset's projected value already has the routed money in it. What no screen breaks out is the split, so you cannot open a savings pot and read off how much of it arrived through the waterfall rather than through a standing order you set on the pot itself.
• No bank connection, and no money moves. CrestCast does not link to your accounts and does not transfer a penny. It shows you what a rule would do so you can go and set it up yourself.