Why CrestCast Doesn't Connect to Your Bank Account

A lot of finance apps lead with the same pitch: connect your bank account and everything happens automatically. It's a reasonable thing to expect in 2026. CrestCast deliberately doesn't work that way.

What a bank feed actually gives you

Open banking connections pull your transaction history, which is to say what you've already spent, categorised as best the algorithm can manage. That's genuinely useful for a spending tracker, because tracking is a backwards-looking exercise: what did we spend last month, and on what.

A forecast is a different problem. It needs to know what your mortgage rate becomes when your fixed deal ends in fourteen months, what happens to your take-home pay if you go part-time next year, and what a second child would do to your childcare bill from next September. None of that exists in a transaction history, because it hasn't happened yet. No amount of bank connectivity can hand you your own future plans.

The trust problem, not just the technical one

There's also a simpler reason. A full household forecast is open to both partners, and it holds a complete picture of your financial life: income, debts, assets and plans. Asking a household to hand real bank credentials to a third-party aggregator on top of all that is a bigger ask than most people realise until they stop and think about it.

What you enter instead

CrestCast asks you to set up your income, bills, debts and assets once, as a model of your household, not a live feed you have to police for miscategorised transactions. You update it when something actually changes: a pay rise, a new bill, a mortgage remortgage. In between, the forecast projects forward from what you told it.

• No bank credentials shared with CrestCast or any third-party aggregator
• No live feed to police for miscategorised transactions, so you control exactly what's in the model
• A forecast built from your actual plans, not just your spending history

It's a small amount of setup in exchange for a forecast that can actually answer the forward-looking questions a bank feed never could.

If you want to see what the manual-entry model actually produces, the other posts on this blog walk through real modelled examples. See "Comparing Financial Scenarios Before You Decide" for how the underlying forecast engine works.