I know several people who make a lot of money in trading, and I hear the liquidity argument constantly as the justification for their behavior. They describe the millions that they make as payment for all the 'value' that they've given to everyone; But, as an ignorant, I can't see how those millions could have come from anywhere than other (less informed) peoples' pockets.
Trading is a legitimately socially useful business, but it's winner-take-all. Yes, they provide liquidity and, in doing so, capture proportionately small amounts of money that other principals don't care about. If you need to move $25 million, are you going to notice a difference of a few hundred dollars that an arbitrageur collects (by taking the other side of a bid he judged to be 0.37 cents high? No. You want your trade to go off. Principals would lose money to the bid-ask spread no matter who's in the market; arbitrageurs narrow it by competing against each other.
So why do traders make so much money? Because they're better or more useful than software engineers? No. Because they steal it? No, not that either. Software engineers are seen by the business as cost centers, even in 90+ percent of startups and even at Google (closed allocation).
For traders, it's a different story. If Bob is a little better than Mark, Bob will get 100% of the business and Mark will get nothing. At this point, to do arbitrage you need to be thinking about microseconds. If Bob can execute in 75 mcs and Mark takes 100 mcs, then Bob is going to get all the trades. Trading shops must be meritocracies because they have no other option. If they can't hire good traders, then there's no reason to keep working.
Because trading is winner-take-all, trading houses put a lot of money back into compensation: 40 to 50 percent profit sharing (in a way that, outside of direct P&L roles, is subject to politically fucked-up performance just like everything else) is the norm. That'd be like a typical software company paying $250k-500k bonuses.
If we, as software engineers, want to make trading money (not the 5-10m outliers, but 250-1M, then we need to think about profit sharing-- http://michaelochurch.wordpress.com/2013/03/26/gervais-macle... -- instead of this startup equity that pays off in the distance future, and is subject to horrible terms). I believe that we, as a group, could be making what we're actually worth, but we'd have to convince businesses that we're as essential to their operations as traders are to trading houses and, thus far, we haven't done so.
If Bob is a little better than Mark, Bob will get 100% of the business and Mark will get nothing. At this point, to do arbitrage you need to be thinking about microseconds. If Bob can execute in 75 mcs and Mark takes 100 mcs, then Bob is going to get all the trades.
If Bob were to be kidnapped by aliens, would society be poorer for it?
Nope, but someone else would get the money instead of his client. Think about you engage a negotiator for buying a house and he only gets 5% bargain when another one could have gotten 10% - the seller gets the money you'd have otherwise.
Not at all, but that's how commodity work works. The provision of the commodity is important, but there's a limited market and superficial or unimportant differences (in traditional marketing, branding; in finance, 75 vs. 100 mcs) determine who gets what share.
What traders do adds a lot of value to society. The difference between 75 and 100 mcs is irrelevant. Ultimately, trading is converging on a circle-jerk of machines throwing numbers at each other, but the world is better off with that circle-jerk, and really doesn't care whether it's Bob or Mark who wins.
Trading is the last commodity job.
However, traders don't make more money than computer programmers or professors because they're more important to society (that's clearly not true) but because of the employer/management filter. For traders, the organization is so sensitive to small differences in individual performance as to justify extreme compensation. Software engineers are worth just as much to the world, but employers still see them as cost centers because, while engineers actually have their employers just as much by the balls, it's not as visceral as it is with traders.
If you think of economic input/output relationships as S-shaped curves (I've dealt with this a lot in exploring convexity and concavity of labor) then trading is an area where the precision/scale parameter has gone to infinity and it looks almost like a step function.
Trading is a legitimately socially useful business, but it's winner-take-all. Yes, they provide liquidity and, in doing so, capture proportionately small amounts of money that other principals don't care about. If you need to move $25 million, are you going to notice a difference of a few hundred dollars that an arbitrageur collects (by taking the other side of a bid he judged to be 0.37 cents high? No. You want your trade to go off. Principals would lose money to the bid-ask spread no matter who's in the market; arbitrageurs narrow it by competing against each other.
So why do traders make so much money? Because they're better or more useful than software engineers? No. Because they steal it? No, not that either. Software engineers are seen by the business as cost centers, even in 90+ percent of startups and even at Google (closed allocation).
For traders, it's a different story. If Bob is a little better than Mark, Bob will get 100% of the business and Mark will get nothing. At this point, to do arbitrage you need to be thinking about microseconds. If Bob can execute in 75 mcs and Mark takes 100 mcs, then Bob is going to get all the trades. Trading shops must be meritocracies because they have no other option. If they can't hire good traders, then there's no reason to keep working.
Because trading is winner-take-all, trading houses put a lot of money back into compensation: 40 to 50 percent profit sharing (in a way that, outside of direct P&L roles, is subject to politically fucked-up performance just like everything else) is the norm. That'd be like a typical software company paying $250k-500k bonuses.
If we, as software engineers, want to make trading money (not the 5-10m outliers, but 250-1M, then we need to think about profit sharing-- http://michaelochurch.wordpress.com/2013/03/26/gervais-macle... -- instead of this startup equity that pays off in the distance future, and is subject to horrible terms). I believe that we, as a group, could be making what we're actually worth, but we'd have to convince businesses that we're as essential to their operations as traders are to trading houses and, thus far, we haven't done so.