Convergence Trades and Relative-Value Arbitrage
The core strategy LTCM traded: betting that small, temporary pricing gaps between closely related securities would converge back together.
LTCM's core strategy was relative-value arbitrage, sometimes called a "convergence trade": identifying pairs of securities that were fundamentally very similar (or, in some structures, mathematically linked) but temporarily trading at a small price discrepancy, then betting that discrepancy would converge back toward its normal, historically observed relationship rather than betting on the direction of the broader market itself. A classic example the book walks through is on-the-run versus off-the-run U.S. Treasury bonds — two bonds with nearly identical cash flows and credit risk, where the most recently issued ("on-the-run") bond trades at a small, typically temporary premium over an otherwise nearly identical older ("off-the-run") bond simply due to liquidity preference, a gap LTCM would bet would narrow.
The critical feature of these trades, which made them attractive to LTCM's model-driven approach, was that the individual price gaps being bet on were typically tiny — fractions of a percentage point — meaning a trade needed enormous size, financed through enormous leverage, to turn a small, statistically reliable-seeming edge into a meaningful dollar return, a structural feature of the whole strategy that this course returns to directly in the next chapter on leverage.
| Position | Bet |
|---|---|
| Long the cheaper, off-the-run bond | Its price rises relative to the on-the-run bond |
| Short the more expensive, on-the-run bond | Its price falls relative to the off-the-run bond |
| Net bet | The small price gap between the two nearly-identical bonds narrows, regardless of which direction rates move overall |
Because these trades were structured to profit from a price gap narrowing regardless of which direction the broader market moved, they were internally described as close to riskless relative to a directional market bet — the fund was not betting stocks would rise or bonds would rally, just that two nearly identical securities' prices would converge as they historically almost always had. The critical, largely unstated assumption buried inside this framing, which the book's later chapters unpack in detail, was that these historically observed convergence relationships would remain stable even during a genuinely abnormal market environment — precisely the kind of assumption the turkey problem and Mediocristan/Extremistan framework from this Book Club's The Black Swan course warns against trusting.
- LTCM's core strategy bet that small, temporary price gaps between closely related securities would converge back toward their normal historical relationship.
- Individual price gaps were typically tiny, requiring enormous leverage to turn the underlying edge into a meaningful dollar return.
- These trades were framed internally as close to "riskless" relative to directional market bets, resting on the largely unstated assumption that historical convergence relationships would hold even in an abnormal market.