Past returns are not forecasts. These are starting assumptions; change them and every result updates.

Ups-and-downs and correlations will default to values measured from bundled data in Phase 4, with a "Use measured / Use assumptions" switch.

Formulas, in plain sight

• Scenario return per asset

  r = beta × share move − duration × rate change + ½ × convexity × rate change² + currency link × rupee move

• Cash raisable quickly

  value × share sellable × (1 − crisis discount × severity), severity = fall ÷ 20% capped at 1

• Share of risk

  weight × (covariance × weight) ÷ total portfolio variance

• Goal simulation step

  value × e^((return − vol²/2) × dt + vol × √dt × random), monthly

• Crash path recovery

  trough + (start − trough) × (1 − 0.5^(months ÷ half-life))

Data sources

No market data is bundled yet — every number above is an assumption and labelled as such. Phase 4 adds a bundled 10-year monthly history (with each source and caveat documented) and lets you upload your own.

What this model cannot see

• Results depend on the assumptions below; change one and the answers change.
• Extreme events beyond anything in the model can and do happen.
• Shocks are applied instantly; real markets unfold over months.
• Beta, duration and crisis discounts are rough values, not measurements.
• Taxes, fees and the real sector mix inside funds are not modelled.
• Crisis correlations are partly rule-based until measured data arrives (Phase 4).
• Past returns are not forecasts; nothing here is a promise or advice.

Educational tool, not financial advice.