A Reliable Cash Forecast for a Developer-Builder
Without clear visibility on its cash position, a developer-builder was drawing its credit line well beyond what it needed. We put in place a reliable forecast and simple drawdown rules, so that it pays only for the credit it actually uses.
The company
A developer-builder with 90 employees, running several residential projects in parallel, from land acquisition to handover.
The problem and its cost
Cash was tracked across several files kept by project managers and by accounting. Sales receipts, construction loan drawdowns and payments to contractors were never brought together in a single view. With no reliable forecast over the coming weeks, management drew on the credit line as a precaution and kept a buffer well above its real needs.
The cost shows up on two lines. On average, CHF 1,200,000 was drawn unnecessarily at 4.5%, which is CHF 54,000 of interest per year. On top of that, 20 hours per month went into rebuilding the figures by hand, at a loaded hourly cost of CHF 95, or CHF 22,800. The calculation is: CHF 1,200,000 × 4.5% + 20 h × 12 × CHF 95, or about CHF 77,000 per year.
The strategy
We first made the data reliable before modelling anything. A forecast is only useful if every input has an owner, a source and an update date. We therefore started with a review of cash flows by project, relying on the tools already in place and adding none.
We then built the forecast, set the drawdown rules, and finally established the steering rhythm. The order is deliberate: drawdown rules only make sense once the forecast is credible, and the steering rhythm only lasts if the teams accept the first two and the bank understands them.
The execution
Team: We brought together a dedicated contact, our finance and data team, and an expert in real estate project financing from our network.
- →A consolidated cash view by project, fed by the existing accounting and site-tracking tools.
- →A rolling 13-week forecast, completed by a 12-month outlook with three scenarios (sales pace, construction delays, payment slippage).
- →Written drawdown rules: trigger threshold, minimum buffer and conditions for early repayment of the line.
- →A monthly steering meeting built around a one-page dashboard that also supports the dialogue with the bank.
We used AI to detect recurring gaps between forecast and actuals, and to estimate real payment delays from invoice history. The finance team reviews every suggestion before it enters the forecast.
The rest is a matter of organization: one owner per data source, a fixed update calendar and a clear definition of an acceptable buffer. These simple rules matter more than the sophistication of the model.
The result
Expected annual saving: 55% of CHF 77,000, about CHF 42,000. For an investment of about CHF 30,000, payback comes in 9 months.
The expected gain is 55% rather than 100% because a safety buffer remains necessary against sales and construction uncertainty, and the analysis does not disappear: it is lighter, but still needs review every month.
- Management decides on shared figures rather than out of caution.
- A clearer dialogue with the bank, backed by documented forecasts.
- Time given back to the finance team for analysis rather than compilation.
The assumptions
- Credit line drawn unnecessarily (average): CHF 1,200,000.
- Interest rate: 4.5%.
- Analysis hours per month: 20.
- Loaded hourly cost: CHF 95.
Scenario type, built from real missions and anonymized.
In the same cocoon
