[yummyfajitas]

That's where they make all their money, right?

This varies year by year. In recent years, prop desks have either gained (Goldman) or lost (Lehman) huge amounts of money.

But this isn't always true. Sometimes services are the big moneymakers. Prime brokerage (allowing hedge funds to outsource their back office) was big up until the crisis nuked many hedge funds, for example. In a good year, classical IBanking (IPOs, M&A, etc) can be big. Big banks run many desks at either a small loss or small profit just for the purpose of keeping the lights on. When the market changes and that desk becomes important, they rake it in.

Is the economy, say, twice as well off from a financial allocation standpoint compared to 25 years ago? If not, how are the trading desks pulling in twice as much money without being extractors?

They could capture a larger portion of the new value being created. Suppose they created 100 units of value in the past, and captured 25% of it. Now suppose they create an extra 50 units of value, but capture 50% of it. Before this change, the banks captured 25 units of value, the world 75 units. After, the banks capture 50 units of value, the world captures 100 units.

As for payday loans, they always existed to some extent (loansharks were always present, as were pawnshops), but they only got into full swing in the 90's. The internet made tracking defaulters easier, competition in electronic banking made the transfers cheaper, and the Clinton-era wave of bank deregulation eliminated many state level interest rate caps (California's was lifted in 1996, for example).

[jbooth]

Ok, I hear your argument regarding capturing a larger part of the new value.. my sense is that they've gotten a lot better at capturing value and are providing very little additional value. I mean I don't live in the finance world, but in my world, the price of anything does not matter within less than a second ever. And it probably doesn't matter within a minute or an hour either. Commodities probably don't even matter within a week or so. No corporation purchasing a large amount of commodities can turn around a decision in under a week, so who cares if they have subsecond pricing accuracy?

Then I talk to friends who work in finance and I see that insane amounts of resources are being thrown into this stuff. Like, truly insane, you're in this industry, you've probably seen it.

Apple makes an iPad, they get money, consumers get iPads.

Goldman makes a subsecond trading system, they get money, consumers get... ???? If consumers and non-financial businesses don't care about the price of AAPL stock within a second or a minute, then how come Goldman makes so much money creating that price stability? What is broken here? It seems like a tail wagging the dog scenario to me.

[yummyfajitas]

Goldman makes a subsecond trading system, they get money, consumers get... ???

Smaller bid/ask spreads, greater liquidity.

Of course, HFT is just a high tech sideshow to the rest of the market. It's not anywhere near as big as you think. Do some simple math - multiply daily share volumes (here is NASDAQ http://www.nasdaqtrader.com/Trader.aspx?id=DailyMarketSummar... ) by the fraction of trades done by HFT (estimates range from 25%-75%) and multiply that by a typical profit of a tenth of a cent. You don't get a huge number.

[jbooth]

Well, we've been through this before.. but I still don't understand, who cares about smaller bid/ask spreads for their own sake? Say HFT was unilaterally banned (which would be ridiculous and excessive), and now bid/ask spreads will naturally spread apart for maybe MINUTES at a time before someone notices and arbs the difference.. how are we worse off?

My major beef here isn't people making money for doing stuff I think is useless.. I think Us Weekly and People Magazine are useless but I don't begrudge the editors their paychecks. My beef is how much talent is being sucked up into a game that doesn't seem to provide any outside benefit. If all of those people were building actual products and services that people paid for (as opposed to basically hacking the finance system for profit), I feel like the nation and world would be a lot better off for it. In short, I think it's a market failure that they can make so much money without actually creating anything.

[yummyfajitas]

Assuming that by "HFT", you actually mean all sorts of algorithmic trading, we are worse off because we have humans doing the work of a computer. If all computerized trading were slowed down by a constant factor of 1000, the world would never notice and HFTs wouldn't care.

I agree with you on wasting talent winning a race rather than creating new value: http://news.ycombinator.com/item?id=2093334

But I don't agree with you that the majority of the finance industry falls into that category. Most trading is not a race. HFT is not representative of the financial sector. The entire HFT sector is only about $20B, which is a little more than double Goldman's profits last year.