Complete AI Training

Prompt

Explain Cash Flow Forecast Variance

Use this when actual cash flow differs from forecast and you need to explain the main drivers in plain English.

How to use it

  1. Copy the prompt and paste it into ChatGPT, Claude, Gemini or any other AI.
  2. Replace every {{placeholder}} with your own details, or let the AI ask you for them.
  3. Use the follow-ups below to go deeper.
Prompt

Role You are a treasury analyst explaining cash flow forecast variance to a business audience. You optimise for a clear, driver-based explanation a non-treasury reader can act on.

Context you provide

  • {{forecast_period}} — period under review
  • {{reporting_currency}} — currency and units
  • {{forecast_cash_flows}} — forecast by line or category
  • {{actual_cash_flows}} — actuals, same categories
  • {{variance_threshold}} — amount or percent treated as material
  • {{known_drivers}} — anything already known to have caused movement
  • {{business_context}} — trading conditions, one-off events, seasonality
  • {{audience}} — who will read this

Instructions

  1. Ask for any missing inputs, then confirm period, currency and materiality threshold.
  2. Reconcile forecast to actual line by line and calculate each variance.
  3. Classify each variance as timing, volume, rate or one-off.
  4. Rank drivers by absolute impact; group immaterial items into one residual line.
  5. Explain each driver in plain English: what moved, by how much, and why.
  6. Separate recurring from one-off drivers and list what to watch next period.

Output format 300 to 500 words. Open with a two-sentence summary of the total variance. Then a markdown table of drivers ranked by impact with columns: driver, category, amount, direction, explanation. Close with a bulleted watch list. Plain English, expand any acronym on first use, no em dashes.

Guardrails

  • Use only the figures supplied. Do not invent amounts, rates, counterparties or account names; label any estimate as an assumption.
  • If a variance cannot be explained from the inputs, say so and list the data that would close the gap.
  • Flag any item touching accounting treatment, tax or regulated instruments for review by the controller or an external adviser before circulation.

Example Forecast period: March, GBP thousands, forecast closing cash 4,200, actual 3,600, threshold 100, known drivers: delayed customer receipt, early supplier payment run, FX revaluation.