Budgeting & Forecasting
Rolling Forecasts
A rolling forecast is a continuously updated projection that always covers a fixed future horizon (e.g., the next 12 or 18 months). As each period ends, a new period is added and estimates are revised with the latest information, so the forecast never expires at year-end and always looks the same distance ahead.
Real-world example
Each month-end the team drops the completed month and adds a new one, keeping a constant 12-month view.
The Budgeting Process
Types of Budgets
Rolling Forecasts
A rolling budget is a formal, approved plan continuously extended by adding new periods, retaining budgetary-control authority. A rolling forecast is a more flexible, frequently updated estimate of likely outcomes, used for guidance and decisions rather than as a fixed control target. Many organizations keep an annual budget but overlay rolling forecasts.
Real-world example
The firm keeps its annual budget for accountability but updates a rolling forecast monthly to steer decisions.
Types of Budgets
The Budgeting Process
Rolling Forecasts
They keep planning current and forward-looking, adapt quickly to change, reduce the 'stale budget' problem and year-end gaming, improve decision-making with up-to-date estimates, and maintain a consistent planning horizon for capacity and cash planning. They suit volatile environments where an annual budget becomes obsolete fast.
Real-world example
When demand shifts mid-year, the rolling forecast updates immediately rather than waiting for next year's budget.
The Budgeting Process
Cash Budgeting
Rolling Forecasts
They require frequent effort and can consume resources if done in excessive detail; they can cause 'forecasting fatigue'; without a fixed target they may weaken accountability; and they need good systems, driver-based models, and discipline to be efficient. Over-detailed rolling forecasts replicate the burden of budgeting more often.
Real-world example
Monthly full-detail re-forecasting overwhelms the team, so they switch to a lean, driver-based rolling forecast.
The Budgeting Process
Rolling Forecasts
Rolling Forecasts
Update frequency (monthly or quarterly) and detail should match decision needs and the pace of change—frequent, high-level, driver-based updates are usually more valuable than infrequent, highly detailed ones. Forecast the key drivers and material lines, not every account, to stay efficient while capturing what matters for decisions.
Real-world example
The team re-forecasts quarterly at a driver level, drilling into detail only for the current quarter.
The Budgeting Process
Cash Budgeting
Rolling Forecasts
A driver-based forecast models financial outcomes from a small number of key business drivers (e.g., revenue = units x price; costs = headcount x cost per head), rather than extrapolating each line item. Changing a driver automatically updates the forecast, making it faster, more transparent, and easier to run scenarios—ideal for rolling forecasts.
Revenue = active customers x average revenue per customer.
Real-world example
Modeling revenue from customer numbers lets the team re-forecast instantly when the customer trend changes.
The Budgeting Process
Cash Budgeting
Rolling Forecasts
Beyond Budgeting replaces fixed annual budgets with adaptive processes, and rolling forecasts are a central tool—providing continuous, forward-looking guidance without a fixed annual target. Combined with relative performance targets and decentralized decision-making, rolling forecasts let organizations plan continuously and respond to change rather than being anchored to an outdated annual plan.
Real-world example
A Beyond Budgeting company steers by rolling forecasts and relative KPIs instead of a locked annual budget.
The Budgeting Process
Zero-Based Budgeting
Rolling Forecasts
Techniques include time-series methods (moving averages, exponential smoothing, trend and seasonal decomposition), regression/causal models linking drivers to outcomes, and judgmental methods incorporating management input and market intelligence. The choice depends on data availability and stability; often quantitative models are combined with judgment.
Real-world example
The analyst combines a seasonal time-series model with sales-team input to forecast the next quarters.
Cash Budgeting
Rolling Forecasts
Rolling Forecasts
A moving average forecasts using the simple average of the last n periods, weighting them equally. Exponential smoothing gives exponentially decreasing weights to older data, so recent observations count more, and it reacts faster to change via a smoothing constant. Both smooth noise; exponential smoothing is more responsive to recent trends.
Exp. smoothing: Forecast = a x actual + (1-a) x previous forecast.
Real-world example
Exponential smoothing tracks a recent uptick faster than a simple moving average of the last 6 months.
Rolling Forecasts
Cash Budgeting
Rolling Forecasts
Decompose the series into trend, seasonal, and residual components. Estimate the underlying trend (e.g., via regression or moving averages), calculate seasonal indices to adjust each period, and combine them to forecast. Deseasonalizing data before trend analysis, then reapplying seasonal factors, produces forecasts that reflect both the growth path and recurring seasonal patterns.
Forecast = Trend x Seasonal index (multiplicative model).
Real-world example
Applying quarterly seasonal indices to a rising trend line produces a realistic seasonal sales forecast.
Rolling Forecasts
Cash Budgeting
Rolling Forecasts