Stress testing abandons probability. You do not ask how likely a 20% two-day index drop is; you apply it and read the damage, then check whether the resulting equity still meets margin, still allows you to trade, and still leaves you willing to trade.
A workable retail version takes ten minutes. Apply to every open position: minus 20% on the index with betas applied, correlations forced to 0.9, volatility doubled, and bid-ask spreads tripled on exit. Then ask three questions - what is the loss, does it breach maintenance-margin, and can the positions actually be exited at those spreads? Options books should add a separate volatility shock, since a long-premium book can gain in the same scenario that ruins a short-premium one.
The output is a decision, not a report. If the stressed result is unsurvivable, the fix is smaller size or defined risk today, while the market is calm and the choice is still yours.
Related: scenario-analysis, worst-case-loss, correlation-breakdown, forced-liquidation