Solow covers a lot of the ground I've covered here on rationality, perfect competition, (Pareto) efficiency and market failure, but it's interesting to me that he does it entirely without once mentioning the concept of ethics.
"I have read that a firm such as Goldman Sachs has made very large profits from having devised ways to spot and carry out favorable transactions minutes or even seconds before the next most clever competitor can make a move. Deep pockets in a large market can make a lot of money out of tiny advantages. (Of course, if you have any such advantage the temptation is irresistible to borrow a lot of money to enlarge your bets and your profits. Leverage is good for you, until it isn’t. It is not so good for the system.) A lot of high-class intellectual effort naturally goes into trying to invent ways to find those tiny advantages a few seconds before anyone else.
Now ask yourself: can it make any serious difference to the real economy whether one of those profitable anomalies is discovered now or a half-minute from now? It can be enormously profitable to the financial services industry, but that may represent just a transfer of wealth from one person or group to another. It remains hard to believe that it all adds anything much to the efficiency with which the real economy generates and improves our standard of living.
If that suspicion is valid--I emphasize that the necessary calculations have not been made and will be hard to make--the conclusion would be that our poorly regulated financial system is not only dangerously unstable, but also too big and too complex, absorbing talent and resources that could be better used doing something else. What is inadmissible is the assumption that, if the market creates a large and convoluted financial system, the market must be right."
Via Ezra Klein