The whole business model is misunderstood by your average internet commentator.
When you have a market, you need someone to be there to provide liquidity. Imagine if you're a farmer and you show up to the market with your wheat, but all the bakers have gone home that day. Or the baker shows up and there's no farmer. The market maker stands around all day offering to buy and sell so that you don't have to wait for the guy you're really trading with. Of course this middle man wants to get paid for it, but your cost as an average Joe is next to nothing. This is trading in time.
Now imagine you want to cook a meal and your ideal meat is beef, but actually you're ok with pork, so long as the pork is cheap enough to make it worth it. How much cheaper should it be? Well your fellow who knows all the pork and beef people will be able to gauge where the balancing spread is, given the amount of interest. In fact he will from time to time do the trade when the spread is out of line. This is trading in space.
So why all the fancy tech? After all market makers used to stand around in a pit in a colored jacket, and they didn't have degrees. My first boss in the market was one of these guys.
Well, things have gotten very tech heavy because as soon as prices are out of line, there is money to be made. Or rather, lost. As a market maker, you are constantly out there with your prices, offering to buy or sell at a very small spread. If some news happens that affects prices in a big way, you can be sure that you will buy when it's going down and sell when it's going up. In order to both have tight prices and avoid this "adverse selection", you really want to be able to react as quickly as you can when your system decides that something's up.
Markets were possible. But what about the current offering of ETFs and similar products? Despite all the fancy talk, many HFT firms are just market makers for the products that enable average joes to invest well and cheaply.
You don't have to be that old to remember how terrible the experience used to be, and how easily even sophisticated individuals were ripped off at every point of the process.
I explained that. It doesn't matter whether it's a computer doing it or a guy in a jacket, but once one guy gets a computer all the market makers need it.
Retail consumers aren't exactly disadvantaged by the gradual narrowing of spreads, so what are you trying to fix? The latency arms-race is exclusively professionals fighting each other.
Getting back on topic, being a software developer for an HFT shop can be a lot of fun: serious technical challenges and a rapid feedback loop in a role where you may be the revenue stream rather than just another cost center. But posters highlighting NIH syndrome are spot on: you risk pigeonholing yourself.
You could easily solve all of this without HFTs - you might introduce a few seconds of latency but virtually no real investors care about that level of latency and it would save them billions per year.
It absolutely would if you remove the cut HFTs take out of the market. I'm sure you'll make some argument about liquidity and spreads but, in the aggregate, the money from higher spreads goes to market participants so the market as a whole still saves by eliminating HFTs, even if the spreads go up a bit.
This is like saying if we eliminated millimeters from rulers, and only used centimeters, everything would get closer together.
HFTs do not take a cut out of the market, their cut comes 100% from other market makers. Those are the only folks who would benefit from eliminating HFTs, and unless you’re one of them, who cares?
I don’t care who buys when I am liquidating a position. That’s the whole point of a market. Therefore I don’t care about HFTs.
Unfortunately, yes. It never sat well with me. Colocated systems saw latencies no mom-and-pop investor could ever dream to achieve. That is one of the reasons why the second company I joined was a non-profit that built systems to help track human rights violations, enforce labor rights, monitor democratic elections etc. Third job was in the food industry and the last one, in housing / property management. So I'd like to think I've made sufficient amends.
Generally, it's not about front-running normal people. Instead its racing other HFTs. Either a race to capture arbitrage opportunities, or to update your orders on one market based on trades made in another market.
The aim of the game is to collect a small spread on a lot of transactions.
The scary thing for an HFT like that is being 'run over' by a big party selling or buying a lot over a day.
If a pension fund is dumping 10٪ of their holdings of shell, it's hard to collect a spread because you need to find buyers for every stock you buy from them. Meanwhile the price is dropping as they sell off.
Hence you occasionally see 'front running' with price improvement. Instead of buying from the market at 1.02 a HFT gets to sell to you at 1.01, yielding you a better price. This is a good deal for the HFT, because a normal person isn't going to run the HFT over. And it's a good deal for the normal person because they get a better price.
No, it's a zero sum game. HFT firms fight to be the first to pick up a penny, but they can't force anyone to drop more pennies. They don't cause prices to go up more than they would have otherwise, or extract more value from the poor and middle class.
Because that's what a lot of people think online. Maybe it's more nuanced. I'm curious how someone from that carreer sees it?