If there's one piece of technical analysis that distinguishes good financial advice from a list of product recommendations, it's cash flow modelling. A cash flow model takes a client's full financial picture (income, expenditure, assets, liabilities, goals) and projects it forward across 20-40 years, testing what happens under different assumptions about returns, inflation, retirement age, and spending. Done well, it turns the whole advice conversation from "what should I invest in?" into "what does my life actually cost, and when can I stop working?"
This guide is the practical one for UK financial advisers: what cash flow modelling actually is, why it matters more than any single product recommendation, the major UK tools (Voyant, CashCalc, Truth, and the newer entrants), how to build a usable model without drowning in assumptions, common pitfalls, and where AI meeting notes fit into the input step that takes most advisers the longest.
What Cash Flow Modelling Is
A cash flow model is a forward-projection of a client's finances across decades. The input is everything you captured in the fact find: current assets, income streams, expenditure patterns, pensions, liabilities, and stated goals. The output is usually a stacked area chart that shows the client's total wealth (or available capital) year by year from today until 30-40 years out, under a set of assumptions you can adjust.
Change the assumptions and the model re-projects. That's the whole trick: a client can see, visually, what happens if they retire five years earlier, buy a second property, fund three university educations, or hit a 2008-style 40% drawdown in year 12. The conversation shifts from abstract product features to specific life choices.
Done well, cash flow modelling is the backbone of a suitability-grade recommendation. Done badly, it's a colourful chart the client nods at and forgets.
Why Cash Flow Modelling Matters More Than Most Product Work
Three reasons it's the highest-leverage technical analysis an adviser does.
It reframes risk. Attitude-to-risk questionnaires are abstract. A cash flow model shows the client, in hard numbers, what their actual capacity for loss is: how much drawdown they can absorb without running out of money in retirement. Suitability conversations grounded in their cash flow model are dramatically more useful than ones grounded in "moderate growth" as a category.
It makes retirement planning concrete. Most clients don't know when they can afford to stop working. A model answers that exactly. It's the single most emotionally charged output financial advice produces, and it's also the one clients most want to see.
It's the compliance spine. Under FCA COBS 9, suitability has to be demonstrably appropriate given the client's circumstances and objectives. A cash flow model tied to documented objectives is the clearest evidence that a recommendation was calibrated to those objectives and not sold into.
The UK Cash Flow Modelling Software Landscape
The UK advisory market has a handful of serious tools. None are perfect; most firms settle on one and live with its idiosyncrasies.
Voyant is the long-established leader, used by many of the larger UK advisory firms. Strong on complex scenarios, multi-generational planning, and deep scenario branching. Steep learning curve. Around £100-£150 per user per month depending on tier.
CashCalc is the modern challenger, bought by Dynamic Planner in 2022. Faster to learn than Voyant, cleaner client-facing output, well-integrated with compliance templates. Mid-market pricing, widely used in smaller IFA firms.
Truth (formerly Prestwood) has a long history in the UK market, particularly strong on retirement planning. Its client-output aesthetic is dated, but the engine is robust.
Dynamic Planner (which owns CashCalc) offers its own integrated cash-flow module alongside its risk profiling. Good option if you already use DP for suitability.
FE analytics, iO, and smaller specialists fill various niches.
New advisers often ask which tool is best. Honest answer: the one your firm already has, unless you're starting from scratch. Switching cash flow tools is expensive, because every model a client has ever seen needs to be rebuilt. Get good at the tool you have before chasing a better one.
How to Build a Usable Cash Flow Model
The skeleton of a good cash flow model, regardless of tool.
- Demographics. Ages, life expectancies, dependants, key dates (children starting university, mortgage end, retirement).
- Income streams. Employment income (with realistic growth assumption), pensions (DB and DC separately), rental income, dividend income, state pension starting at the correct age.
- Expenditure. Essential vs discretionary split. Most clients underestimate both. Tie it to bank statements, not memory.
- Existing assets. Pensions (DB valued as capitalised income, DC at current value), ISAs, GIAs, cash, property (primary and investment), other.
- Liabilities. Mortgages with amortisation schedules, other debt.
- Assumptions. Growth rates by asset class, inflation (CPI vs RPI), salary growth, discount rate if you use one.
- Goals. Target retirement age, spending targets in retirement, specific lump-sum goals (house purchase, university fees, gifts), legacy intent.
Run the base case first. Then stress-test it: 30% drawdown in year 5, early retirement, a year of long-term care at £80,000+, inflation at 5% for a decade. The model earns its keep in the stress tests, not the central case.
Assumptions: The Part Most Advisers Get Wrong
The single biggest source of misleading cash flow models is over-aggressive assumptions. Three specific traps.
Growth rates too high. A 7% nominal growth assumption on a 60% equity portfolio looks fine until inflation eats three points of it. Be honest about real returns. Many UK firms now use 4-5% nominal on a moderate portfolio with 2.5% inflation, yielding ~2-2.5% real growth.
Inflation too low. Historic UK CPI averages 2.5%. Your model should probably assume 2.5-3% as baseline with a stress at 4-5%. Anything below 2% will flatter the result.
Expenditure too static. Real spending changes over decades: higher in early retirement (travel, the 'Go-Go years'), declining in the 'Slow-Go' and 'No-Go' years. Flat expenditure forever is a modelling convenience, not a realistic projection.
The tool won't catch these. You have to set sensible defaults at the firm level and discipline yourself not to move them to flatter the output.
Where the Input Step Usually Goes Wrong
Most advisers we talk to say the same thing: the modelling part is fine, the inputting is what kills Tuesday afternoons. You've had a 90-minute client meeting, you have a fact find that's 60% complete, and now you need to turn it into structured inputs for Voyant or CashCalc with no gaps.
The places it breaks:
- Stated objectives in the fact find are phrased too vaguely to translate into model targets ("retire comfortably" is not a spending figure).
- Income and expenditure numbers from the meeting are approximate; you need to chase the bank statement reconciliation.
- Soft signals (the client mentioned caring for a parent; they think they might downsize in 2030) didn't make it into the structured notes.
- You're rebuilding the fact find from memory two days later.
This is where AI meeting notes have changed the input step materially. A structured, AI-generated fact find that preserves the client's exact language on objectives, captures all the numbers discussed, and surfaces the soft signals gives you the raw material to build a model from without chasing the adviser or guessing.
At Heavenly, we see advisers cutting cash-flow modelling input time by half, and fewer "can you clarify what they meant by X?" emails to the client after the meeting. The modelling still takes however long it takes; the input phase collapses.
See the advisor meeting notes template for the structured fact find we use, and the paraplanner guide for how the model build usually sits with the paraplanner rather than the adviser personally.
How to Present a Cash Flow Model to a Client
The model is built. Now you have to run the conversation.
Three things that consistently work.
- Lead with the base case, not the stress tests. Show them the central projection first so they can orient. Stress tests afterwards demonstrate resilience.
- Use the model's sliders live. Don't prepare five static screenshots; open the tool and move the retirement-age slider in front of them. The live interactivity is what makes it visceral.
- Frame it as "here's what the numbers say, and here are the choices you have." The adviser's job is to show the tradeoffs clearly and name the recommendation; the client's job is to choose.
Things that consistently fail:
- Presenting five scenarios with no narrative. Clients cannot hold that many branches in their head.
- Showing the model on a screen so small they can't read the axis labels. Rehearse the presentation ergonomics.
- Using the stress tests to argue for a specific product. That reads as sales, not advice.
Frequently Asked Questions
What is cash flow modelling in financial planning?
Cash flow modelling is a forward-projection of a client's finances over decades, taking their current assets, income, expenditure, and goals and showing how their wealth evolves under different assumptions. It's the primary tool advisers use to answer "when can I retire?" and "what happens if X?"
What is the best cash flow modelling software for UK advisers?
The biggest UK tools are Voyant, CashCalc, Truth, and Dynamic Planner. The honest answer is that the "best" one is the one your firm already uses; switching is expensive. For new advisers, CashCalc has the lowest learning curve; Voyant is stronger on complex multi-generational scenarios.
How long does it take to build a cash flow model for a new client?
Typically 2-4 hours of adviser or paraplanner time for a first build: 1 hour of input, 1-2 hours of base-case and stress testing, 30-60 minutes of presentation prep. Subsequent reviews of the same client take 30-60 minutes.
How often should a cash flow model be updated?
At each annual review at minimum. Also whenever there's a material life change (job change, house purchase, inheritance, divorce). The model earns its value when it's kept current; a cash flow model built once and never touched is a souvenir, not a tool.
What assumptions should I use in a UK cash flow model in 2026?
Common UK defaults: CPI inflation 2.5-3%, nominal growth on a moderate portfolio 4-5%, salary growth in line with CPI plus 0.5-1%, state pension indexed under current triple-lock rules. Stress-test with inflation at 4-5%, a 30% drawdown, and early retirement.
Getting the Input Step Right
Cash flow modelling is the highest-leverage analysis in financial advice, but only if the inputs reflect what the client actually wants and what their finances actually look like. The math is straightforward; the raw material is what usually fails.
That's the part Heavenly was built for. Structured, AI-generated client meeting notes capture the exact language clients use when talking about retirement, spending, goals, and concerns, and preserve the numbers discussed in the meeting so you're not rebuilding the conversation from memory when you open Voyant on Tuesday.
If you want to see it run on one of your own client meetings, book a 20-minute demo, or download Heavenly and try it on your next call.