Crisis · KOSPI Index (1998)
The KOSPI plunged 41.9% from 579.25 to 336.63 between October 1997 and May 1998 amid one of Asia's fiercest crisis periods. Replay each daily bar, navigate 9 critical decision points, and receive an AI coach review of your simulated choices.
150 daily bars · 26 decision points · 1997-10 to 1998-05 · Stocks
Key facts
This window spans 225 days with a peak-to-trough drawdown of 48.4% and a 63.8% run-up.
- Max drawdown
- −48.4%
- Max run-up
- +63.8%
- Duration
- 225 days
- Largest single-day move
- −11.0%
- Bars
- 150 Daily bars
- Decision points
- 26
Every figure above is computed programmatically from this replay’s own candle series — no editorial input, no LLM. Methodology: Methodology
The market advances day by day. You only see candles up to the current day — as if you were really there.
At key decision points choose Buy / Trim / Exit / Hold. Each decision advances the timeline by one day.
After settlement you get a six-dimension score and an AI coach review. Your result enters the scenario leaderboard.
Event Replay
What happened — and what it teaches
Between October 1997 and May 1998, the Korean stock market stood at the center of one of Asia's most punishing financial crises, and this replay lets you walk straight through it one daily bar at a time. The dataset spans 150 trading sessions, a long enough runway for conviction to build, for panic to spread, and for policy headlines to dictate the rhythm of every open and close. The stakes for any investor, trader, or treasurer embedded in this market were nothing less than the preservation of capital under conditions of acute uncertainty.
Over the replay window, the KOSPI index fell from 579.25 to 336.63, a drawdown of 41.9%. The peak-to-trough stress reached deeper still, with a maximum drawdown of 48.4% that captured the worst moment of forced selling and liquidity strain. Volatility was extreme: at one point a single session moved 11.0%, a reminder that in a crisis regime, intraday positioning and overnight exposure carry very different consequences. Twenty-six decision points are scattered through the timeline, nine of them flagged as critical, where the framing of the next day's risk would change the shape of the entire replay.
The training value here is process discipline rather than pattern prediction. Replay the policy-driven shocks, document the trigger for each trade, and notice how often a clear narrative in hindsight was accompanied by real-time ambiguity. The lesson is to size for the largest plausible single-bar move, to predefine exits before conviction turns into commitment, and to separate the analyst's urge to act from the risk officer's duty to preserve optionality when drawdowns can approach half of starting capital.