What Happens When You Actually Understand Your Own Financial Plan

I'm not a financial advisor. I'm not a wealth manager. I'm a Head of Data who thinks too much about data – including his own.

The starting point was simple: my wife and I wanted to understand what different life decisions mean financially. Not abstractly, not in a spreadsheet that nobody understands after three months – but interactively, immediately, with real answers to real questions.

The question that really occupied us wasn't: "When can we stop working?" The question was: "What decisions are we making today that will have what consequences in ten, twenty, thirty years – and what does that mean for our freedom?"

That's a fundamental difference. And it shaped the entire project.

The Problem: No App Can Model My Life Situation

I tried. Really. There are good finance apps, solid pension calculators, decent budgeting tools. But they all fail at the same point: they're built for an average situation.

My situation isn't average – and I suspect yours isn't either. Two separate investment portfolios. A house. My wife's self-employment. Two different pension paths. And specifically: what does that mean for our monthly travel budget – the thing we really want to treat ourselves to?

No app answered these questions. So I built the app.

The Approach: A Simulation Engine That Actually Calculates

The heart of the dashboard is a custom-built simulation engine in TypeScript – around 500 lines of code that calculate the portfolio trajectory over more than fifty years on an annual basis. No averaging, no simplification: each year is simulated individually, with all income, expenses, savings rates, pension entitlements and phase transitions.

The special part: the travel budget is not an input. It's an output. The simulation uses binary search to find the highest monthly budget at which the portfolio still contains a defined buffer at the end. That sounds technical – and it is. But the result is intuitive: you immediately see what you can afford if you make this or that decision today.

All parameters are interactively adjustable: portfolio balances, savings rates, return assumptions, inflation, timing of a potential property sale, retirement age, buffer target. Every change instantly updates all views.

Two Modes: Today's Purchasing Power vs. Actual Numbers

A detail I had underestimated: the choice between nominal and real calculation makes an enormous difference – not in the results, but in understanding.

Mode A calculates in today's purchasing power: 5% real return, no inflation. The numbers are smaller, but they mean something. You see what you can afford in today's euros.

Mode B calculates nominally: 7% return, 2% inflation. The numbers are larger and look more impressive – but they're in future euros that are worth less.

The difference sounds academic. It isn't. When you first see how strongly inflation erodes over thirty years, it changes the way you think about savings rates and return assumptions.

The Death Scenario: The Uncomfortable Question

There's a feature in the dashboard that I put off for a long time. Not because it was technically difficult – but because the question behind it is uncomfortable.

A slider simulates what happens if I die before my wife. That sounds morbid. But it's one of the most important planning questions a couple can ask.

The calculation is complex: depending on which life phase a death occurs in, pension entitlements differ, income losses vary, and the widow's pension under new law is calculated differently. The dashboard shows for each scenario: portfolio at time of death, widow's pension, Julia's total income, monthly shortfall, portfolio requirement until age 100 – and a clear traffic light status.

The result surprised us both. Not because the numbers were bad – but because we had never actually calculated how good or bad the protection actually is. Now we know.

What I Learned While Building

The development process was iterative – and it taught me more than the finished dashboard.

The first bug was symptomatic: the calculated travel budget was suspiciously low. I questioned the numbers, went through the logic – and discovered that my wife's self-employment was completely missing from the calculation. An income source that I naturally factor into everyday planning, I had simply forgotten in the model.

That's the value of your own model: it forces you to be explicit. Every assumption must be translated into code. And when an assumption is missing, it immediately shows in a result that doesn't add up.

Another bug was more subtle: the special value for the accumulation phase – a fixed amount that wasn't supposed to be dynamic – was interpreted as an actual multiplier in two components. The result was a negative travel budget in the accumulation phase. Technically a small error. Conceptually an important reminder: edge cases must be handled explicitly, not implicitly.

The Dashboard as a Decision-Making Tool

After several evenings of development – in 60-minute sessions after work, with Manus AI as a sparring partner – the dashboard is finished. And it has served its purpose.

Not because it told us what to do. But because it helped us understand what different decisions mean. What an earlier or later retirement means. What happens if my wife's self-employment ends sooner or later. What one additional working year means in pension points.

We had these questions before. We didn't have the answers. Now we do.

What the Dashboard Is Not

I want to be honest, because honesty is the brand identity of this blog.

The dashboard is not a financial advisor. The tax calculation is simplified – no capital gains tax on portfolio withdrawals, no private health insurance costs in retirement, no stochastic simulation of sequence-of-returns risk. It calculates with fixed returns, not market volatility.

It's a tool for understanding, not for optimising. For real financial planning, you need a real financial advisor.

But as a tool for understanding – as an instrument that translates abstract life decisions into concrete numbers – it's the best thing I've ever built for our financial planning.

How Manus AI Helped

I'm often asked: what does Manus AI actually do in projects like this?

The honest answer: it's not the developer who builds everything alone. It's the sparring partner who is always available. Who doesn't get tired when I have a question about widow's pension calculation at 10pm. Who finds errors I've overlooked. Who suggests alternatives when my first approach doesn't work.

The simulation engine is written in TypeScript – a language I don't speak fluently. Without Manus AI, this project wouldn't have come together in this form. Not because I didn't have the idea – but because the implementation would have taken too long to be realistic in 60-minute sessions.

That's the real value of Agentic AI for managers: not automation, but acceleration. The idea comes from me. The understanding comes from me. The decisions come from me. But the implementation – that now happens in hours instead of weeks.

Conclusion: Understanding Is the First Step

The dashboard is not public – it's a private tool for us. But the idea behind it is universal: anyone who truly wants to understand their own financial situation needs to model it. And anyone who wants to model it needs to be explicit – about every assumption, every phase transition, every income source.

That's uncomfortable. But it's the only way to turn abstract numbers into concrete decision-making foundations.

What questions do you have that no standard app can answer?

Write to me – I'm curious to hear.