A retirement projection built on a single average return is a straight line, and a straight line is a promise nobody can keep. Real markets deliver their returns in a sequence, and the sequence matters as much as the average — a poor first decade in drawdown is very hard to recover from.
Retire models that directly. Every scenario is run a thousand times or more against sampled market behaviour, and the output is not a number but a distribution: in how many of those futures does the money hold, and in the ones where it fails, when does it fail?
What the engine actually does
Each run walks the portfolio forward year by year, drawing the client's income, growing what is left according to sampled returns, and applying inflation and tax as it goes. The success rate is simply the share of runs where the portfolio outlives the client.
- State Pension captured straight from a gov.uk State Pension forecast PDF
- A minimum of a thousand paths behind every reported result
- Multi-segment historical bootstrap, including overseas exposure held in sterling
- Sequence-of-returns risk modelled explicitly, not averaged away
- Confidence bands, median path, and best and worst cases
Tax is in the loop, not bolted on afterwards
The withdrawal the client asks for is the amount they want in hand. The engine grosses that up through the HMRC bands inside the simulation loop, so the drawdown taken from the portfolio is the real one. Estimating tax after the fact understates the drawdown and flatters the projection.
- HMRC income tax bands and the personal allowance applied per year
- Net-to-gross conversion inside the simulation loop
- Per-year tax shown alongside the income schedule
Portfolios as they really are
A retiree rarely holds one product. Scenarios carry multiple assets — living annuities, discretionary unit trusts, fixed deposits, offshore holdings and cash — each with its own asset mix, drawdown rules and ordering, so the model draws down in the sequence the adviser intends.
- Drawdown pension, ISA and GIA, fixed deposit, offshore and cash asset types
- Per-asset equity, bond and cash weightings
- Drawdown ordering and locked assets
- Fund-level allocation by SEDOL, classified against IA sectors
Something to hand the client
The output leaves the system as a report an adviser can put in front of a client: the projection, the assumptions behind it, the income by year, and the trailing performance and risk of the funds involved.
- PDF and CSV export of any completed run
- Income-by-year schedule
- Trailing performance and risk statistics per fund
- GDPR-aligned handling — client identifiers encrypted at rest, subject access and consent withdrawal built in
About this demo
You will be signed into the real product against a sandbox advisory practice — six client plans, each with a Monte Carlo projection that was produced by the actual engine, not written by hand. It is read-only: every screen, projection and report is open, but starting a new simulation or changing a plan is held back with a note explaining what the live system would have done.
Important
The clients, portfolios and projections in this demonstration are entirely fictional and are shown to illustrate how the software works. Nothing here is financial advice, a recommendation, or a forecast of any actual investment outcome.
Where to start
Start as the Adviser — open the client book and look at the completed projections, particularly the two that are in trouble. Then switch to the Practice Administrator for the whole-practice view, and to the Paraplanner to see the read-only permission boundary the product enforces.
- Every client in the firm, across all advisers
- The fund universe and performance data
- Firm users and their permissions
- Billing, plan and simulation quota
- My own client book
- Scenario building — assets, drawdown and inflation assumptions
- Monte Carlo results with confidence bands
- Client-ready PDF and CSV reports
- Client records and completed scenarios
- Simulation results and reports
- No editing — the permission boundary made visible