Methods and accountability
Methods and assumptions
Understand what a result measures before using it. Each tool states its inputs, units, formula, worked example, and exclusions alongside the result.
Updated
Put it to work
Check a result in three steps
- Find the matching example below and reproduce it in the calculator. All example figures are hypothetical.
- Change one input at a time, then read the limitation beside that result.
- If the result still looks wrong, prepare a correction report with those same inputs.
Position sizing
Try this hypothetical example
- Enter
- Account 100,000; risk 2%; long entry 50; stop 45; whole units.
- Expected result
- Risk budget 2,000 ÷ risk per unit 5 = 400 units. Planned notional is 20,000; modeled loss at the stop is 2,000 before costs.
- Check next
- Re-enter a stop of 40: size falls to 200 units. Confirm buying power and execution assumptions separately.
Planned risk budget equals account balance × risk percentage ÷ 100. Divide that budget by the absolute distance between entry and stop, then round down to the chosen tradable increment.
This limits the modeled loss at the entered stop before costs. It does not guarantee a fill at that price or establish buying power, liquidity, or a suitable portfolio allocation.
Risk and reward
Try this hypothetical example
- Enter
- Long entry 100; stop 95; target 115.
- Expected result
- Risk 5; reward 15; reward-to-risk 3:1. Simplified break-even win rate: 5 ÷ 20 = 25% before costs.
- Check next
- Move the target to 110: the ratio becomes 2:1 and the break-even win rate becomes about 33.33%. Neither figure estimates the chance of winning.
For a long trade, risk per unit is entry minus stop and reward is target minus entry; a short trade reverses those directions. Reward divided by risk is the reward-to-risk ratio.
The simplified break-even win rate is risk ÷ (risk + reward) × 100. It assumes repeated full wins and full losses of those sizes before costs; it is not a probability forecast.
Weighted average purchase price
Try this hypothetical example
- Enter
- Buy 10 units at 100 and 20 units at 70, in the same currency.
- Expected result
- Total cost 2,400 ÷ 30 units = 80 per unit before fees.
- Check next
- The simple average of the two prices is 85 and is wrong here because the quantities differ. Confirm that no sales or adjustments need accounting for.
Add each purchase quantity multiplied by its price, then divide by total quantity. All rows must represent the same asset, currency, and quantity unit.
The result describes entered purchases. It does not incorporate sales, fees, dividends, corporate actions, or tax-lot rules, and it does not measure strategy performance.
Constant-return scenarios
Try this hypothetical example
- Enter
- Initial 1,000; annual return 10%; one year; annual compounding; no contributions.
- Expected result
- Hypothetical ending value 1,100, of which 100 is growth.
- Check next
- Re-run with 0% return: the ending value is 1,000. These are assumptions for checking the tool, not expected returns.
The annual return input is effective: using the annual rate as a decimal, the periodic rate equals (1 + annual rate) raised to (1 ÷ periods per year), minus 1. Contributions occur at the selected beginning or end of each period.
A constant return is an assumption. The output separates cash contributed from mathematical growth and excludes inflation, costs, taxes, withdrawals, and variation in market returns.
Single-option expiry outcomes
Try this hypothetical example
- Enter
- One long call; strike 100; premium 5 per unit; multiplier 100; expiry price 120.
- Expected result
- (20 intrinsic value − 5 premium) × 100 = 1,500 profit before costs. Expiry breakeven is 105.
- Check next
- Re-run at expiry price 100: loss is 500, the full premium. This does not price an exit before expiry.
Call intrinsic value is max(expiry price − strike, 0); put intrinsic value is max(strike − expiry price, 0). Long profit/loss equals (intrinsic value − premium) × contracts × multiplier. A short position reverses the sign.
Premium is per underlying unit. The model assumes a nonnegative underlying price and a positive whole-number multiplier. Long-call potential profit and short-call potential loss are unbounded in this model. A negative put breakeven is marked unattainable. Pre-expiry pricing, margin, settlement funding, fees, taxes, and early-assignment consequences require separate analysis.
Company evidence and valuation
Research distinguishes reported facts from calculations, estimates, scenarios, and editorial judgments. Changeable claims carry source references and an as-of date. Operating results, cash flow, and valuation answer different questions.
A valuation requires a dated share price and a defined earnings or cash-flow basis. We do not treat a one-time gain as recurring earning power or replace an unknown figure with zero.
Price context
A historical close is a dated observation. A small number of price observations does not establish a durable trend, tested support, executable quote, or current valuation. Manually maintained snapshots are labeled accordingly.
Conditional scenarios
A scenario states a trigger, outcome, invalidation condition, horizon, and assumptions. Editorial thresholds are not calibrated probabilities or management guidance unless explicitly identified as such.
Bull, base, and bear conditions can be non-exhaustive. Mixed or unclassified signals require reassessment rather than automatic assignment to the nearest label.
Company comparisons
A numerical difference is first company minus second company. It is shown only when units, accounting bases, period dates, and durations match; valuation observations must also share a date. Percentage changes are compared in percentage points.
Different reporting periods stay visible and no aligned-period difference is shown. Matching dates do not make different business models equivalent. There is no weighted investment score or automatic winner.
Precision and limits
Calculations use raw inputs; display formatting happens afterward. Decimal arithmetic is used for supported arithmetic operations; compound growth uses floating-point exponentiation. Extreme values outside the supported range produce an explicit error or unavailable result.
Automated checks cover formulas, boundaries, and publication rules. They do not establish that evidence is current or that an investment is appropriate. Source checks and editorial review are separate responsibilities.