For finance managers whose business holds cash across several bank accounts: how to get from eight balances to one reconciled position, forecast 13 weeks from it, and see which balances are sitting idle.
How much cash do we have? In a business with an operating account, a payroll account, a tax reserve, savings, two currency accounts and a company card, that question usually takes a spreadsheet and twenty minutes, and the answer is out of date by Wednesday.
A finance team sees one cash position across several bank accounts by forecasting from the reconciled balances in its accounting system, not from the banks. Every account the team includes adds to one starting balance, transfers between included accounts net to zero, foreign-currency balances are converted at a stated rate, and the 13-week forecast runs from that single figure.
The reasoning is short. The accounting system already holds all eight accounts, reconciled, in one currency, with every receipt and payment assigned to what it belongs to. A balance roll-up built from eight bank logins reproduces that work by hand, once a week, and is right on the morning it is done. Forecasting from the ledger is not a shortcut. It is the only version of the position that every other number in the forecast already agrees with.
Float is built to work this way. It reads every bank account and credit card through your accounting system, so the position it shows is the one the books already agree with, refreshed every day.
Why A Business This Size Ends Up With Eight Bank Accounts
A business of 11 to 50 people opens bank accounts one at a time, each for a good reason. A payroll account so wages are never short. A tax account so VAT and PAYE money is not spent twice. A savings account for the reserve. A dollar account because a customer pays in dollars, a euro account because a supplier invoices in euros, and a company credit card for travel and subscriptions. The tax account has official backing in two of Float's four markets: the Australian Taxation Office and New Zealand's Inland Revenue both recommend putting the tax you collect into a separate bank account.
Together they create a different problem. Nobody can answer the Monday question without logging into several banks, and the cost is not the twenty minutes. It is what the figure leaves out: a transfer counted as cash out, a dollar balance converted at whatever rate was to hand, and a total built from statement balances rather than a reconciled position.
The Common Way: A Weekly Balance Roll-Up
The balance roll-up works like this. Someone logs into each bank on Monday morning, types eight balances into a sheet, converts the two foreign-currency balances at a rate they look up, and totals the column. The forecast is built on that total. It produces a number, and the number is right at the moment it is typed.
It breaks in four places.
It is a snapshot. The balances are true at nine o'clock on Monday and drift from the moment the first payment leaves.
It uses bank balances, which include items the books have not yet dealt with. A customer receipt that landed on Friday afternoon sits in the bank balance but has not been matched to its invoice, so the forecast still carries that invoice as open and expected this week. Count the bank balance and you have counted the receipt twice. A forecast built from the ledger starts from the reconciled balance instead, because an item that has not yet been matched to its invoice, bill or account has no place in the ledger for the forecast to read it from. The rest of the gap between the two figures is the subject of our guide to why your cash position in Xero differs from your forecast.
It records transfers between accounts as cash movements. The payroll top-up appears on the operating account's statement as a payment out, so it goes into the sheet as an outflow, and the arrival in the payroll account is never listed as an inflow.
It belongs to one person. The rate cell, the credit-card convention and the dormant-account tab all sit in the head of whoever built the sheet, and when that person is away the cash position is not produced.
The Better Way: Forecast From The Ledger, Not The Banks
The better way is to forecast from the accounting system, where every account already lives. The ledger holds all eight accounts, reconciled, in one base currency. Starting the forecast from the reconciled balance of the accounts you include gives one figure the rest of the forecast already agrees with, produced the same way every week by anyone on the team. The method has four steps.
Step 1: Reconcile every account, daily or at least weekly. An unreconciled item is money that moved but has not yet been matched to the invoice, bill or account it belongs to, so a forecast built from the ledger cannot place it. The reconciled balance is therefore the one this method starts from. Daily for the operating account and at least weekly for every account, including the savings and foreign-currency accounts that see little traffic, is a workable rhythm; no professional body sets one for a business of this size.
Step 2: Decide which accounts belong in the position. Operating, payroll and tax accounts go in, because they hold cash the business will spend. A savings account goes in if the money is available on demand; a deposit on notice is a different thing and is treated on its terms. A credit card goes in or out depending on how payments to it are recorded: include it and every card transaction appears as spend on the day it happens, with the monthly settlement netting to nothing; exclude it and only the settlement appears, as a payment out. Either is right, doing both counts the spend twice, and the choice is a bookkeeping convention rather than an accounting rule. A dormant account with a few hundred pounds in it is still cash. This method leaves it out of the working position because nothing will be paid from it, and says so, rather than letting it sit in the total as money nobody will use.
Step 3: Treat transfers between included accounts as nothing. Money moving from operating to payroll is not cash in or cash out. It is the same cash in a different account. The accounting standards say the same of the cash flow statement: IAS 7 excludes movements between the items that make up cash from cash flows altogether, which is why a transfer never appears on one. In a ledger-based forecast both sides of the transfer sit in included accounts, so they cancel. The roll-up's error is counting the outgoing side without the incoming side. A transfer is only a cash movement when one side of it sits in an account you have left out of the position.
Step 4: Convert foreign-currency accounts at a stated rate, and keep the original amount visible. The position has to be in one currency, so the dollar and euro balances are converted into the base currency. State the rate against the converted figure and keep the original amount a click away, so anyone reading the position can tell how much of a movement is cash and how much is the rate. The standards draw the same line: under IAS 7 and FRS 102, a change in the translated value of a foreign-currency balance is not a cash flow, and is shown separately from receipts and payments. ICAEW's guidance on reviewing a spreadsheet makes the practical point, that it should be obvious which currency every input and output is in. Rates across several entities and markets belong to our guide to multi-currency cash forecasting.
A Worked Example: Eight Accounts, One Position
Every figure below is invented for the example. The business is a UK company on Xero with a finance team of four, monthly payroll and quarterly VAT, holding eight accounts. The exchange rates are illustrative: £1 to $1.30 and £1 to €1.15 at the stated rate, against £1 to $1.25 and £1 to €1.20 in the spreadsheet, last updated some weeks ago. The first table shows what the Monday roll-up typed into the sheet, beside the reconciled balance in the ledger for the same account on the same morning.
| Account | Balance typed into the sheet | Reconciled balance in the ledger | In the position? |
|---|---|---|---|
| Operating | £219,500 | £212,000 | In |
| Payroll | £48,000 | £48,000 | In |
| VAT reserve | £61,000 | £61,000 | In |
| Savings | £150,000 | £150,000 | In |
| US dollar account, $38,000 | £30,400 at £1 to $1.25 | £29,231 at £1 to $1.30 | In |
| Euro account, €12,000 | £10,000 at £1 to €1.20 | £10,435 at £1 to €1.15 | In |
| Company credit card | −£6,400 | −£6,400 | In |
| Dormant account | £900 | £900 | Out of the working position |
| Total | £513,400 | £504,266 |
The sheet says £513,400. The reconciled position of the seven included accounts is £504,266. The £9,134 between them has three parts: a £7,500 customer receipt that landed on Friday and has not been matched to its invoice, which the forecast still carries as open, so the sheet counts it twice; the dormant account's £900, which is cash the business holds but will not pay anything from, and is left out of the working position for that reason; and £734 of rate, which is not cash.
The transfer is the larger error, and it does not show in the balances at all. On Friday the team moved £48,000 from operating to payroll ahead of the month-end pay run. The sheet's forecast tab lists it as this week's outflow, because that is how it appears on the operating account's statement, so the sheet's closing position is £48,000 lower than the business's. In the ledger-based position both sides sit in included accounts and net to nothing.
The second table runs the 13-week forecast from £504,266, starting on Monday 26 October 2026. Payroll is £128,000 on the last working day of each month. VAT for the quarter to 30 September, £61,000, is paid on Friday 6 November, ahead of the Saturday 7 November deadline. PAYE and National Insurance of £44,000 fall in the weeks of 16 November, 21 December and 18 January. Receipts are dated on when customers usually pay, and December is quiet.
| Week beginning | Receipts | Payments | Closing position |
|---|---|---|---|
| Mon 26 Oct | £118,000 | £206,000 | £416,266 |
| Mon 2 Nov | £141,000 | £133,000 | £424,266 |
| Mon 9 Nov | £97,000 | £80,000 | £441,266 |
| Mon 16 Nov | £133,000 | £120,000 | £454,266 |
| Mon 23 Nov | £104,000 | £79,000 | £479,266 |
| Mon 30 Nov | £116,000 | £202,000 | £393,266 |
| Mon 7 Dec | £139,000 | £83,000 | £449,266 |
| Mon 14 Dec | £108,000 | £77,000 | £480,266 |
| Mon 21 Dec | £51,000 | £113,000 | £418,266 |
| Mon 28 Dec | £23,000 | £180,000 | £261,266 |
| Mon 4 Jan | £64,000 | £81,000 | £244,266 |
| Mon 11 Jan | £147,000 | £79,000 | £312,266 |
| Mon 18 Jan | £136,000 | £120,000 | £328,266 |
The low point is the week of 4 January, at £244,266, after December payroll lands on a month of thin receipts. The position never drops below the £150,000 in the savings account, so the savings balance can stay where it is all quarter. Take savings out of the reading and the rest of the business bottoms at about £94,000 that week. The receipts that lift the position in the week of 11 January are the ones that matter: if they slip into February, that is the week the savings account would be drawn. The forecast does not say which account the January low sits in. That is the team's decision, made from the one position and the list of what each account holds today.
Where Cash Is Sitting Idle
Once the position exists, the next question is how much of it is working. In the example, £150,000 of the £504,266 is a savings balance the forecast never needs, and the £61,000 VAT reserve is spent in the second week. The finance manager can see which balances are doing a job and which are sitting still, and can ask whether the reserve is the right size.
This page does not answer that last question. Sizing a reserve floor is the subject of our guide to how much cash a business should keep in reserve, and reading the position as weeks of cover against dated payments is in our guide to weeks of cash cover. The single position is also one of the four things our guide to improving cash flow visibility in a finance team covers, and it is the starting figure for the weekly cash job that sits between the accounting system and the annual plan, which our guide to the operational cash layer sets out.
How Float Does This Today
Float is cash flow forecasting software for in-house finance teams. It reads your accounting platform once a day at an hour you choose, over a one-way connection that writes nothing back, and you can sync on demand when you have just finished reconciling. The position is worth having because it is current, and keeping it current is two habits: the team reconciles, and Float syncs every day. Each of the four steps maps to something Float does now.
Step 1 in Float. The Balance figure is the sum of the reconciled balances of the accounts you include, the "Balance in Xero" or "In QuickBooks" figure, which is why Float's own guidance is to reconcile daily or at least weekly.
Step 2 in Float. Float imports every bank account and credit card from the ledger and lists each included account with the balance being used for it. You choose which accounts are in the cash flow; excluding an account removes its transactions too. A credit card can be included, so every card transaction appears, or excluded, so only the payments to it appear, as a bank transfer.
Step 3 in Float. Transfers between two included accounts net to zero in a Bank Transfers row. If only one side of a transfer is included, the row shows the net cash in or out of the included account. For QuickBooks Online, Float reads the transfers that are recorded as transfers in QuickBooks; a movement recorded as an expense does not come through.
Step 4 in Float. Foreign-currency accounts, invoices and bills are converted into the accounting platform's base currency, with the original currency and the rate used shown against each item. The dollar account in the example reads in sterling with its $38,000 kept in view.
What stays current without anyone typing it. Once the position is in place, much of the forecast built on it maintains itself. Actuals replace budgets as they land. On Xero, Smart Expected Dates applies each customer's average days late to new invoices as they import, which is how the receipts column in the example gets its dates, and VAT is forecast automatically for UK businesses (GST in Australia and New Zealand), which covers the VAT line in the example. On either platform, an auto budget tracks a cost from last month or a three- or six-month average, and a linked budget sets one line as a percentage of another, with a date offset. QuickBooks Online teams set expected dates and the tax line themselves.
From a position to a decision. The point of one position is the decisions it makes possible. In the example, the team can now answer three questions the spreadsheet could not: whether the savings account needs touching this quarter (no), which week to watch (4 January), and which receipts decide it (the ones due the week of 11 January). A decision can be tested before it is made. Add a hire starting in January as a scenario layer, or move a large receipt back 30 days, and the forecast shows whether the low week still clears the savings balance. Set a threshold at the reserve and that becomes a date on screen.
The forecast runs as a 13-week weekly view, which is the second table above without the spreadsheet. PDF and CSV exports are run by you, and four user roles, including a read-only viewer, let the finance director and the chief executive open the same position the team maintains. Where the business is several companies rather than several accounts, Float consolidates them at company level with a display currency and group or per-company balances, covered in our guide to consolidating cash flow across multiple entities. Float connects to Xero and QuickBooks Online today, with Sage Intacct coming soon; you can join the Sage Intacct waitlist.
Two limits are worth knowing. The position comes through the accounting platform rather than a direct bank connection, so it is as current as the last reconciliation and the daily sync. And Float forecasts the included accounts as one position rather than account by account, and does not move, sweep or pool cash between them.
Float's direction is a cash system that keeps reconciling what was expected with what actually happened, built on a richer record of how cash moves through a business: when customers pay, when suppliers are paid, and how forecasts differ from actuals. The numbers stay traceable to their source and under the finance team's control. Today, that record starts with the reconciled ledger.
Frequently Asked Questions
We hold cash across eight bank accounts, how does a finance team see one cash position and forecast from it?
Forecast from the reconciled balances in the accounting system, not from the banks. Include the accounts that hold cash the business will spend, let transfers between them net to zero, convert foreign-currency balances at a stated rate, and start the forecast from the total. The position is then produced the same way every week.
How do we see how much cash each overseas entity holds and what the group has in total?
Connect each entity's own ledger and consolidate at company level, with a display currency and a toggle between group and per-company balances. The mechanics are in our guide to consolidating cash flow across multiple entities. Reading the group line when exchange rates move is covered in our guide to multi-currency cash forecasting.
Should a cash flow forecast start from the bank balance or the reconciled balance?
For a forecast built from the ledger, the reconciled balance. A bank balance includes receipts and payments the books have not yet matched to an invoice, bill or account, and the forecast usually still carries those items as open, so starting from the bank counts them twice. The reconciled balance is the figure every other number in the forecast agrees with.
How do you handle transfers between bank accounts in a cash flow forecast?
Treat a transfer between two accounts that are both in the position as nothing: the money has moved from one included account to another, and the position is unchanged. Record it as an outflow and the forecast shows the business poorer by its own payroll top-up. A transfer is only a cash movement when one side sits outside the position.
Can you build a cash flow forecast with multiple bank accounts in different currencies?
Yes, provided each foreign-currency balance is converted into the base currency at a stated rate and the original amount stays visible. The forecast then runs in one currency from one starting figure, and anyone reading it can separate a movement in cash from a movement in the rate. The accounting platform's base currency is the natural choice.
Does Float forecast each bank account separately?
No. Float forecasts the accounts you include as one position, from the sum of their reconciled balances, and the Balance panel lists what each account holds today. The 13-week view, the threshold date and the exports all read from that single figure. Which account a future low point sits in is the finance team's decision.
Float is built for in-house finance teams of three or more that hold cash across several bank accounts, entities or currencies. Build the position from your own ledger: start your free 14-day trial, connect your accounting ledger, and the Balance panel will show every account and the one figure they add to.







