Financial scenario comparison: two versions of your plan, side by side

Most tools that offer what-if analysis let you nudge one field and watch one number move. CrestCast works the other way round. A version is a complete copy of everything: every income, every bill, the mortgage, the savings, the pension contributions, the planned changes you have already scheduled. You give it a name and a colour, and it lives alongside your live plan.

Once you switch into a version, you edit it exactly as you would edit your real finances. Add a debt, change a salary, sell the house, stop an income. Every one of those edits writes to that version and never to your live plan. That is the property worth leading on: you can take your real finances completely apart and put them back without touching the real thing. If the result is depressing, delete the version and nothing has happened.

There is no cap on how many you keep, and a new version branches from whatever you are currently looking at, so you can branch a branch. Start from the version where you downsize, then branch that one again to see what downsizing plus clearing the car loan does. Versions can be renamed, recoloured and deleted at any point. This is the feature the trying it before you commit part of CrestCast is built on.

Compare Versions puts exactly two versions side by side. Not three, not five. Two, because a variance figure is a difference between two things and stops meaning anything the moment you add a third. If you have three plans to weigh up, you compare A with B and then compare the winner with C.

What you get back is not a headline number. It is the charts and the complete statements. You choose:

• one statement at a time, cashflow, profit and loss, or net worth
• monthly or annual periods
• whose money you are looking at, through the account perspective picker

Then every period renders with the left version, the right version and the variance between them. That is what turns “the overpayment version looks better” into “the overpayment version is £3,400 better on net profit and £16,200 ahead on net worth by the last year both plans reach”. It also lets you find the exact period the two plans cross over, which is usually the number the decision actually turns on.

Milestones are drawn on the comparison charts by default and are worked out for you rather than typed in: a debt clearing, a promotional rate ending, a mortgage deal ending, a student loan write-off date, and every dated change you have scheduled. So when one bar jumps, you can normally see why without leaving the chart.

This is where comparing versions beats comparing figures. A change to one input almost never stays in one place. Put a £6,000 pay rise into a version and CrestCast recalculates the income tax on it, the National Insurance, the student loan repayment if you have one, and the pension contribution taken as a percentage of the new gross. What lands in your account is what is left, and that is what then flows into the cash balance, the surplus, and the net worth line years out.

The same is true in the other direction. A version where one income stops does not just remove a salary row. It changes the tax position, changes what is left after the bills, and changes how fast the savings are drawn down. The comparison shows you the second-order effects because it is running the same engine twice on two complete sets of facts, rather than subtracting one number from another.

Say the mortgage is £220,000 at 4.6% with 22 years to run, and there is £300 a month spare. The question is whether to throw it at the mortgage or leave it in savings.

Make a version called Overpay £300/mo, set the recurring overpayment inside it, and compare it against your live plan. The mortgage is re-amortised with the overpayments in it, so the balance falls faster, the interest charged falls with it, and the payoff milestone moves earlier on the chart. The interesting part is that the three statements do not agree. On cashflow the overpayment version is behind, because £300 a month is genuinely leaving your account. On net worth it is ahead, because the debt is shrinking faster than the cash is. Both readings are correct, and they are answers to different questions, which is exactly why the comparison hands you all three statements rather than picking one for you.

Then you can ask the harder question the same way: which year does the gap actually become worth having, and what does the cash balance look like in the meantime. Our overpay versus invest page walks through that trade-off in full.

Nothing about Compare Versions is mortgage specific. The versions people build most often are the ones where the numbers are large and the intuition is unreliable:

• Downsizing. Enter the sale as a dated event inside the version. The proceeds settle the linked mortgage, what is left arrives as cash, and the smaller property goes on the balance sheet. Both statements move at once.
• Upsizing. The mirror image, with a bigger mortgage, a bigger monthly payment and a different answer on cashflow than on net worth.
• A pay rise. Schedule it on the date it actually starts rather than assuming it applies from today, and let the tax, National Insurance and pension come out of it.
• Redundancy, or one income stopping. Set the income to stop on a date inside the version and read the cash balance. The useful output is not the gap in income, it is the month the balance crosses zero.

Each of those is built the same way: branch, edit inside the branch, compare. See should I downsize my home for a worked version of the first one.

CrestCast is built for two people planning together, and a household shares the joint forecast. Versions are the deliberate exception. They belong to your account and are never published to your partner, even when you are sharing everything else.

That matters more than it sounds. Some of the scenarios worth modelling are uncomfortable ones: leaving a job, one of you going part time, what the numbers look like if an income stops. Being able to work out whether it is even viable, quietly, before you raise it, means you can bring a considered position to the conversation instead of a worry. When you do want to talk about it, you can rebuild the change in the live plan and you are both reading from the same numbers.

The spreadsheet answer to scenario comparison is Save As. It works for about a week. From the moment you press it you have two files that diverge, and the divergence is not the interesting part of the model, it is your actual life. Your mortgage rate changes and only one file knows. You add a standing order in one and forget the other. Six weeks later you cannot tell whether the difference between the two files is the decision you were testing or the maintenance you skipped.

Then there is the comparison itself. A spreadsheet has no variance column unless you build one, and building one means every row has to line up in both files, in the same order, forever. Insert a row in the copy and every formula in the comparison tab is quietly wrong. Most people give up and read the two tabs side by side, flicking back and forth, which is precisely the mental arithmetic the tool was supposed to remove.

Every CrestCast projection is a single deterministic line calculated at the rates and amounts you enter. There is no Monte Carlo simulation, no probability attached to any outcome and no confidence band around any line. A comparison answers “what happens if both of these sets of assumptions hold”. It does not answer “how likely is either”, and any tool that claims to answer the second question is telling you more than it knows.

It compares two versions at a time. Three plans means running two comparisons, and that is a deliberate limit rather than a missing feature.

And there is no bank connection. CrestCast does not link to your accounts and nothing is pulled in automatically, so a version is built from figures you type, or from a bank statement CSV you export yourself and feed in. That is more work up front. It is also the reason the model can describe a future that has not happened, which no amount of transaction history can do. If you want the reasoning in full, we wrote it up in why CrestCast does not connect to your bank.

Versions are the branching layer. Underneath them sits the same engine that runs your live plan: the cashflow forecast, the profit and loss statement and the net worth projection, all built from the incomes, bills, debts and assets you have entered once. A version inherits all of it, which is why creating one takes seconds.

They also work with dated changes rather than against them. Anything you can schedule in your live plan, a pay rise, a rate change, an income stopping, an asset sale, you can schedule inside a version, at a different date or a different size. Comparing two versions that each carry their own schedule of planned changes is the closest the app gets to comparing two lives.