Holy Grails
Here is an original promotional document from Long Term Capital Management (10 MB). This LTCM document is only available online here at TurtleTrader.com.
Download the LTCM promotional document (PDF).
LTCM still, a few years later, is a fantastic case study and a monster cautionary tale. You can read about those who won LTCM’s losses in the book Trend Following, understand the background in the book When Genius Failed: The Rise and Fall of Long-Term Capital Management by Roger Lowenstein and read a great white paper from the CATO institute on LTCM.
Read the CATO Institute white paper on LTCM.
The Original LTCM Document
The promotional document available here is the fund’s own pitch to prospective investors. Reading it with the knowledge of what happened subsequently is one of the more instructive exercises available to any serious student of markets and risk. The document presents the intellectual framework, the mathematical models, the team credentials, and the risk controls that made LTCM appear to be the most sophisticated and best-protected trading operation in the world. Two Nobel Prize winners. A former Vice Chairman of the Federal Reserve. The most rigorous quantitative models in the industry. Risk management systems that had been tested across every historical scenario the team could model.
None of it was sufficient. The scenario that destroyed the fund was one the models had not adequately assigned probability to: a simultaneous global flight to quality following a Russian debt default, in which every convergent position the fund held moved against it at the same time with amplified force because every other major institution was simultaneously attempting to reduce exposure to the same positions. The model said these correlations were unlikely. The market made them certain.
The LTCM case is the Efficient Market Hypothesis failure documented in complete operational detail. The fund’s partners believed markets were efficient enough that prices would converge to fair value over time, and that their models correctly measured the probability and magnitude of deviations from that convergence. The model was wrong in the way all models are wrong about tail risk: it measured how often extreme events had occurred in the historical sample, not how often they would occur in the future. The future arrived in August 1998.
Who Won LTCM’s Losses?
Trading is zero-sum. LTCM’s losses went somewhere. The book Trend Following documents the traders on the other side of those losses. Systematic trend following operations, positioned across global markets in the direction of the price movements that LTCM’s forced liquidation accelerated, captured significant returns during the period that destroyed LTCM. The correlation data showing that multiple independent trend following managers made money in the same months of 1998 is the fingerprint of a common systematic response to the same market opportunity: large, sustained directional moves across currencies, bonds, and equities produced by the largest convergent strategy failure in financial history.
The holy grail framing in the original page is precise. LTCM was the market’s most credentialed attempt to build a holy grail: a system that used superior intelligence, Nobel Prize-winning mathematics, and exhaustive historical analysis to produce consistent risk-free returns. It produced consistent returns for four years and then catastrophic failure in one. The search for the holy grail always ends the same way. The alternative, a systematic approach that acknowledges uncertainty and responds to price rather than predicting it, produces a less impressive pitch document but a more durable track record.
Frequently Asked Questions
What is the original LTCM promotional document?
It is the fund’s own pitch to prospective investors, presenting the intellectual framework, team credentials, mathematical models, and risk controls behind Long-Term Capital Management. Reading it with knowledge of the subsequent collapse reveals how even the most rigorously documented and intellectually credentialed approach can contain the seeds of catastrophic failure. It is available exclusively at TurtleTrader.com.
What does the LTCM collapse teach about model risk?
That models measure the probability of historical events, not future ones. LTCM’s models correctly assessed the probability of extreme events based on historical data. The 1998 collapse was an event that the historical data did not assign adequate probability to. The models were not poorly built. They were built on an assumption that the future would resemble the past in its distribution of extreme events. That assumption failed at the worst possible moment given the fund’s leverage levels.
How does the LTCM story relate to trend following?
LTCM ran a convergent strategy, betting that prices would revert to historical relationships. Trend following runs a divergent strategy, following price wherever it goes. The event that destroyed LTCM, a simultaneous large move across multiple markets in the same direction, is exactly the environment where trend following produces its largest gains. The money LTCM lost went to participants who were positioned in the direction of the market movements LTCM’s forced liquidation created.
Trend Following Systems
Want to learn more and start trading trend following systems? Start here.
