1998: When Global Markets Caught Russia's Contagion

On August 17, 1998, the Russian government defaulted on its domestic debt obligations and devalued the ruble. The signal that reached trading desks in New York, London, and Tokyo was unambiguous: a G-8 nation had broken its debt commitments. What followed over the next three months was not a contained regional event but a rerouting of global liquidity that left equity investors everywhere exposed to drawdowns they had not modeled for.

The S&P 500 entered the episode near 1115.75 on April 14, 1998, and by November 4 had closed at 1118.67 — a net move of just 0.3%. That flat closing figure was beside the point. The same 144-session window that produced the Default · S&P 500 Index (1998) scenario recorded a maximum drawdown of 19.3% and a largest single-day move of 6.8%. Participants replaying those bars faced a market that oscillated between panic and relief, delivering a net-zero result through a maximum intermediate loss of nearly one-fifth of peak portfolio value.

The more acute stress showed in smaller, less liquid markets. The Jakarta Composite Index — the subject of the Default · Jakarta Composite Index (1998) scenario — fell 7.2% on a net basis (from 430.53 to 399.46) over 146 sessions, yet its maximum drawdown reached 47.4% with a largest single-bar move of 10.7%. Investors who survived that drawdown endured repeated rounds of margin calls, forced selling, and a banking sector in simultaneous collapse. The Straits Times Index in Singapore recorded a 39.1% maximum drawdown despite a 5.8% net gain. The S&P/TSX Composite in Canada — another Default · S&P/TSX Composite Index (1998) scenario — fell 14.9% net while suffering a 31.3% maximum drawdown.

The mechanism that amplified the shock was leverage, and the case that crystallized systemic risk was Long-Term Capital Management. LTCM had built positions on the assumption that asset correlations were stable and that volatility would mean-revert. Those assumptions held — until they did not. When Russian default compressed liquidity and drove previously uncorrelated assets into simultaneous decline, LTCM's 25-to-1 leverage ratio turned a mark-to-market loss into an existential crisis within weeks. The Federal Reserve Bank of New York facilitated a $3.6 billion coordinated bailout on September 23, 1998, not to save LTCM but to prevent losses from propagating through its prime brokerage counterparties.

The Crisis · S&P 500 Index (1998) scenario reproduces the same 144-session window as the default case (April 20 to November 10, 1998) but frames it explicitly around the LTCM collapse. FinFable's scenario data for this window shows that the 6.8% largest single-day move coincided precisely with the period of maximum LTCM unwinding. What looked like an S&P 500 event was, in reality, a prime-brokerage and repo-market event wearing an equity-market disguise.

Not every market fell. The OMX Helsinki Uptrend Replay (1998-06-16 to 1998-07-21) captures Finnish equities grinding 2.6% higher (4857.94 to 4982.95) through a summer that was already proving brutal elsewhere. The largest single-bar move of 3.1% was modest compared with double-digit daily swings in Jakarta or Abu Dhabi. Nordic insulation from Asian contagion was real but temporary — the OMX Copenhagen 20 Uptrend Replay (1999-03-01 to 2000-10-01) shows that Finnish and Danish indices were pulled into the subsequent European correction regardless.

The FTSE 250 Downtrend Replay (1998-06-09 to 1998-10-08) illustrates the European dimension of the same shock. British mid-caps fell 13.1% on a net basis (5570.5 to 4843) with a maximum drawdown of 28.8% and a largest single-day move of 3.3% — lower single-day volatility than emerging markets but a sustained, grinding six-month decline that tested position management more severely than acute crashes do. Eleven decision points, four marked critical, forced participants to repeatedly choose between holding through deteriorating conditions and reducing exposure at what proved to be intermediate lows.

For markets, the immediate consequence was a repricing of emerging-market risk premiums that did not fully normalize for eighteen months. For institutions, the LTCM episode exposed the gap between Value-at-Risk models calibrated on normal correlations and the actual behavior of leveraged portfolios during stress. The leverage ratio that looked appropriate under historical precedent — 25-to-1 — proved catastrophic when two sigma events ceased to be independent. For the regulatory record, the episode prompted a re-examination of prime brokerage concentration and the implicit too-big-to-fail subsidy embedded in counterparty credit lines to hedge funds.

For decision trainers today, the 1998 episode is not primarily a story about the Russia default itself — it is a story about what happens when the correlation assumptions inside a leveraged institution stop working. The S&P 500 delivered a net return of 0.3% across those 144 sessions, yet any participant who failed to predefine a drawdown tolerance threshold would have endured — and possibly been forced to exit during — a 19.3% peak-to-trough loss. The Jakarta Composite delivered a net decline of only 7.2%, but the same window included a 47.4% maximum drawdown that would have destroyed most accounts without disciplined position sizing. The lesson is not that crises are survivable in retrospect; it is that the process of surviving them — through pre-defined sizing rules, correlation stress tests, and a clear distinction between intermediate noise and structural break — is what distinguishes disciplined execution from outcome-blind holding.