CBA Certified Financial Modelling & Analysis with AI

Model Design and Spreadsheet Craft18%

• Select an appropriate model structure, layout and level of granularity for a stated business question and audience
• Apply consistent conventions for inputs, calculated cells, units, signs and time periods across a workbook
• Construct formulae using lookup, aggregation, date and logical functions that remain correct when rows and periods are added
• Distinguish robust formula patterns from fragile ones, including hardcoded values inside formulae, inconsistent ranges and hidden dependencies
• Interpret error values, broken links and check-line failures to locate the underlying fault in a model
• Evaluate a model against integrity criteria such as balance checks, flag totals, one-formula-per-row consistency and documented assumptions

Three-Statement Model Construction22%

• Construct linked income statement, balance sheet and cash flow logic in which every line item has a defined driver or schedule
• Calculate working capital balances and their cash flow effects from days-based or ratio-based drivers
• Build supporting schedules for fixed assets, depreciation, debt and equity, and link them correctly into all three statements
• Distinguish causes of a balance sheet imbalance and select the correct diagnostic sequence to resolve it
• Interpret circularity arising from interest on average debt or cash and select an appropriate treatment such as switch-controlled iteration or an opening-balance convention
• Evaluate whether a completed model's cash, retained earnings and net debt movements reconcile with the underlying schedules

Forecasting and Scenario Analysis18%

• Select appropriate forecast drivers for revenue, costs and balance sheet items given the business model and available data
• Calculate forecast values from growth rates, ratios, run-rates and simple regression or trend logic
• Distinguish between base, upside and downside scenario design and sensitivity analysis, and select the right technique for a stated question
• Construct scenario switches and sensitivity structures, including data tables, that change assumptions without rewriting formulae
• Interpret sensitivity outputs to identify which assumptions genuinely drive the result
• Evaluate a forecast's credibility against history, capacity constraints, market size and disclosed external benchmarks

Valuation and Investment Appraisal20%

• Calculate free cash flow to the firm from model outputs, adjusting correctly for depreciation, capex, working capital and tax
• Calculate and interpret NPV, IRR and payback for a project or investment, including their conflicts and limitations
• Construct a discounted cash flow valuation, including a weighted average cost of capital and a terminal value by perpetuity growth or exit multiple
• Distinguish enterprise value from equity value and execute the bridge between them
• Select and normalise comparable company multiples such as EV/EBITDA and P/E, and interpret differences across peers
• Evaluate the sensitivity of a valuation to discount rate, terminal assumptions and forecast drivers, and identify which conclusions survive

AI-Assisted Modelling, Audit and Assurance22%

• Construct effective task specifications for an AI assistant to draft formulae, schedules or model documentation, including context, constraints and expected output format
• Select appropriate uses of AI across the modelling workflow, distinguishing tasks where AI accelerates work from tasks where human judgement must lead
• Interpret AI-generated formulae, code and explanations to detect logical, accounting and referencing errors before adoption
• Distinguish AI failure modes, including fabricated figures, plausible but incorrect accounting treatments, stale knowledge and lost context in long tasks
• Evaluate the confidentiality, data protection and governance implications of sending financial information to external AI services
• Construct a verification workflow, combining check cells, reconciliation tests, independent recomputation and documented review, for any AI-assisted output