Long Term Capital Management

Jan 9th, 2010 | By | Category: Finance Theory

I've been reading number of articles recently about Long Term Capital Management (LTCM).  LTCM was a hedge fund run by John Meriwether out of Greenwich, CT.  Meriwether assembled some of the best minds in the field of finance theory: Scholes, Merton, etc.  It was both a stunning success (1995, 1996, and 1997) as well as an epic failure (1998).  Many writers have been using LTCM as another example of finance academics and wall street investors running out-of-control.  I do not think that this is the case.  In fact, I think an important accomplishment has been tarnished–the creation and subsequent implementation of a revolutionary risk management tool.

Fisher Black, Myron Scholes, and Robert Merton created the perfect formula.  It is perfect, not because it is infallible but because it pushed the frontiers in finance and ushered in a new era of risk management.   The frontier of science is not a clean and orderly place.  Failure proceeds success and success is defined, if at all, in shades of grey.  History often takes time to write the winners and the true success of an accomplishment may be much different in nature than originally envisioned. 

For all of the disorder and ugly preceding failure, the human race needs to push the frontiers in order to continue to sustain and grow our standard of living.  Black, Scholes, and Merton—as well as Meriwether and the other Long Term Capital Management (LTCM) partners and investors—pushed finance to new places and we as a society are better for it. 

Unfortunately, the world is not often appreciative of it pioneers. 

The epic implosion of LTCM was surely disruptive.  At the very least, it cost several investors a substantial amount of money and put the global financial system under pressure.  The press and politicians saw the opportunity to attack.  (It is interesting just how few politicians and press actually understood what they so fearlessly and courageously lambasted—after the fact.)  Congressional committees were hastily convened.  Several books were published.  The average American could be forgiven for condemning the ‘failed wizards of wall street’. 

But the truth of the matter was that their discovery (the Black-Scholes model for pricing options) provided a vast new set of tools for investors—and if used prudently, could enable investors to effectively manage the systemic risks of their positions.  Building on the work of Louis Bachelier at the turn of the 20th century, LTCM provided a real-world laboratory so that the ideas could be tested under real market conditions.  The first three years yielded enormous profits and demonstrated the potential under general market conditions.  After raising three billion in three months the firm returned 20% net (after management fees) in year 1, 43% in year 2, and 41% in year 3.  But in 1998 there were a series of events that were unanticipated: the Asian crises and the Russian sovereign debt default.  These events were outside the data built into distributions used by LTCM. 

In the end, the models were vindicated (the spreads on their opposing positions eventually converged), but it was too late.  The counter-parties to LTCM required higher-and-higher margins, positions were called, and leverage squeezed to astronomical heights.  LTCM was effectively bailed out by a consortium of banks at the height of the liquidity crises.  (The banks all made their money back—in effect proving the strategy.)

The failure of LTCM proved an effective market signal.  It was capitalism at work, providing a real check on the upper bounds of the financial theory (at least for now) and set the rules for traders.  Capitalism has two sides: success and creative destruction.  We cannot take only the upside benefits—we must accept that the cost of our economy’s success is that periodic failures will serve as guideposts.  In this case, the lesson was that although modern financial theory predicated on probability analysis and statistics can help to manage risks it does not completely remove the element of human judgment.  Financial markets continue to be subject to both rational investment valuation and human subjectivity.  Black, Scholes, and Merton made important discoveries about the former.  LTCM did not manage to remove the latter. 

For me, the point is that models work the vast majority of the time.  To expect perfection is unrealistic.  According to Peter Fisher, Federal Reserve Bank of New York, “models don’t drive financial markets”.  Since there is still an element of human subjectivity and judgment, we can expect the occasional model mishap since models rely on inputs and a rational, functioning market.  Financial modeling is no different than other technologies.  We accept the occasional failure in other technologies because we understand occasional failure is a by-product of operating in the real world.  Product defects are real and we are typically happy if the defect rate can be effectively minimized—but not eliminated.  Perhaps because of the mathematical formulas, the public expects too much of our financial engineers.  With time and public education, we can come to see the benefits of the models while appreciating their limits.

We have been given the opportunity to benefit from the discoveries of Black, Scholes, and Merton—among other economists.  By applying these important risk management tools in a prudent manner we can significantly reduce our risks.  By appreciating the inherent limitations of financial modeling we can apply judgment and intuition.  There is room for both science and intuition in our financial markets and we would do well to appreciate the limitations and shortcomings of both.

Tags: , , , , , , , , , , , ,

6 Comments to “Long Term Capital Management”

  1. Cialis says:

    xaiRaO Excellent article, I will take note. Many thanks for the story!

  2. Great post!
    For me, one of the great lessons of the LCTM collapse was the assymetry of history. The Black Scholes formula assumes that the future will resemble the past. To paraphrase Nassim Taleb, George Bush has lived for 64 years. Black Scholes would call him immortal. It has no ability to predict the unprecendented. Yet unprecendented events shape financial history.
    I would be loath to trust in risk management predicated on an assumption that the future will equal the past. So how can formulas such as the above be supplemented? Or will we be perennially gazing into the rear view mirror?

  3. Stephen,

    I’m a big fan of Taleb. My personal belief is that although models are reliant on inputs and assumptions they are useful tools for helping to make informed decisions about future events. When the model, inputs, or assumptions fail to account for previously unpredictable events, or black swans, the result can be disasterous. However, these unpredictable events happen far less frequently than less, more boring, events occur. So, if I were betting on whether George Bush is alive tomorrow, the models predict yes–and I’d most likely be right. I will someday be wrong; but I’d be right many times more than the one time I was wrong.

    But I do agree that 100% reliance is dangerous. Models need constant updating to account for new black swans, outliers, and previously unthought of outcomes. The models improve and so do the results. I hope an important outcome of the recent CDS mess is that the models are now more robust and financial institutions are better able to price risk.

    So I guess my two cents on the matter is that yes, we’ll probably be using models based on historical data for a long time to come–but that the models will get better and better over time.

  4. Steve says:

    xaiRaO Excellent article, I will take note. Many thanks for the story!

  5. Wonderful web site. Plenty of helpful info here. I am sending it to some friends ans also sharing in delicious. And certainly, thanks on your sweat! How to add color to your black dresses http://www.575a.com/plus/view.php?aid=42431

Leave a Comment