I sort of know what Buffet means. It's like in the late 90's with regards to Long Term Capital Management hedge fund:
"At the beginning of 1998, the firm had equity of $4.72 billion and had borrowed over $124.5 billion with assets of around $129 billion. It had off-balance sheet derivative positions with a notional value of approximately $1.25 trillion."
Eventually most of this was lost within a few months of volatility caused by events not taken into account by its computer models, and the fund was ultimately liquidated, losing its investors much of their $1 Billion investment.
The fund used models determined by some of the smartest around including economics Nobel prize winners. This team background then enabled them to raise a lot capital, which was subsequently extremely levered.
From how I understand it, another problem of LTCM (besides for the flaw in the model not taking into account an extremely rare occurence) was due to the fund venturing into different types of trades that weren't fully thought-out once all the arbitrage opportunities ran out in the program's designated purview.
Another aspect of Buffet's hesitancy with "black box" trading can be found in this hedge fund manager's interview: http://news.ycombinator.com/item?id=310284, that is, the programmers doing the modelling generally have the same type of educational backgrounds and thus the trading patterns and performance of their 'boxes' are all similar, thereby giving none a competitive advantage over the other.
At the end of the day, simplifying things, is it better to trust a brain and its computer extension or just a brain? In Buffet's case, being the equivalent in investing of grandmaster Garry Kasparov in chess, he sticks with his brain. Anyone else might want to run with the computer if they can't get shares in Berkshire Hathaway. I think LTCM were just very unlucky and a bit reckless, but could have done very well, if only they had more conservative exit points in winding up the fund.
But, the main gist of my previous comment was that it didn't add much to the thread in terms of discussion, but rather just brought a point in the article to people's attention. And just from that, I couldn't see all the mod points justified. I guess this goes back to my style of modding which is: "I'll mod this up or down, but only if it's under or over valued." ;)
"At the beginning of 1998, the firm had equity of $4.72 billion and had borrowed over $124.5 billion with assets of around $129 billion. It had off-balance sheet derivative positions with a notional value of approximately $1.25 trillion."
http://en.wikipedia.org/wiki/Long-Term_Capital_Management
Eventually most of this was lost within a few months of volatility caused by events not taken into account by its computer models, and the fund was ultimately liquidated, losing its investors much of their $1 Billion investment.
The fund used models determined by some of the smartest around including economics Nobel prize winners. This team background then enabled them to raise a lot capital, which was subsequently extremely levered.
From how I understand it, another problem of LTCM (besides for the flaw in the model not taking into account an extremely rare occurence) was due to the fund venturing into different types of trades that weren't fully thought-out once all the arbitrage opportunities ran out in the program's designated purview.
Another aspect of Buffet's hesitancy with "black box" trading can be found in this hedge fund manager's interview: http://news.ycombinator.com/item?id=310284, that is, the programmers doing the modelling generally have the same type of educational backgrounds and thus the trading patterns and performance of their 'boxes' are all similar, thereby giving none a competitive advantage over the other.
At the end of the day, simplifying things, is it better to trust a brain and its computer extension or just a brain? In Buffet's case, being the equivalent in investing of grandmaster Garry Kasparov in chess, he sticks with his brain. Anyone else might want to run with the computer if they can't get shares in Berkshire Hathaway. I think LTCM were just very unlucky and a bit reckless, but could have done very well, if only they had more conservative exit points in winding up the fund.
But, the main gist of my previous comment was that it didn't add much to the thread in terms of discussion, but rather just brought a point in the article to people's attention. And just from that, I couldn't see all the mod points justified. I guess this goes back to my style of modding which is: "I'll mod this up or down, but only if it's under or over valued." ;)