Eric Rosenfeld brings us back to ten years ago what went wrong at LTCM. Refreshing if you link the collapse of LTCM to the current financial crisis.

A few take-away points:

1. The benefits from diversification based on historical correlations is great only if correlations don’t change during sudden move;

2. The collapse of LTCM is not a problem of flight-to-quality, rather, it’s the problem of everybody trying to get out of the same trades LTCM had.

3. We need to seriously deal with counterparty risk, and work out a solution. Otherwise, the too-big-to-fail problem will haunt us, always.