As far as my understanding goes, your example would be more accurate if you had alluded to showing up at a dealership looking for a Toyota and some other guy was following you around, waited until you settled on a car, and then jumped in front of you to buy it. He then turned around and sold you the car for a slight markup. The same with the Ebay example. Ebay’s part is played by the exchange. The HFT is the sniper who buys the product right at the buzzer and then turns around and offers to sell it to you for one extra penny.

Maybe it’s just my lack of financial sophistication but I still don’t understand how that is helpful.

They are not providing shares to buy. They don’t hold anything. If I can buy it from them, then I could have bought it from someone else, the same person they did, within seconds of when I got them.

Why should I pay a guy even pennis for holding onto a stock for a second? The brokerage fees may have been higher in the 70s, but at least those companies charging fees actually were making markets in the shares. They held shares and took risks in holding them. HFTs only hold them for mere moments.

Looking at markets other than stocks, does this seem realistic? Real estate professionals take huge cuts out of real estate transaction volume, even when they do very little work. Everyone is mad when they have to pay the fee, but no better system has been invented. Ebay takes a huge cut on auctions, even though they are just a listing service where the user has to do all the work. People complain, but no one has created an effective and free or cheaper competitor. Car salesman make a nice living for providing the service of “Oh, you would like to buy a Prius? Well, fortunately we are a Toyota dealership so we can get you one in a month, as long as you like yellow. Pay me!” Diamond rings have a huge retail mark-up built in, but a large fraction of grooms pay this mark-up rather than finding the owner of an existing diamond who wants to sell. I could go on.

In this framework, HFT looks pretty good. If you want to buy 100 shares of IBM and it’s trading around 190 all week, you can go into the market and pick your limit price in the 190 neighborhood and get your shares quickly, paying an implicit price of a pennies per share to your intermediary. You’re not paying 6%, as you do in many other markets and as stock traders used to (look up what brokerage fees were in the 70s). It’s not like Knight Capital and its peers were making gangbusters profits before its misstep – the remaining market-making-type firms are survivors in an industry where many older firms have gone out of business due to profits drying up due to more efficient markets.

Really, TFF? Curmudg – IMO it’s not quite fair to compare this piece to that Amtrak burger-fetish bit, which actually had something (thin) to it. Seems more appropriate to put it in a class with, and second-only to, that (thankfully passing) infatuation someone had about finding a ‘free lunch’ through creative Target2 default accounting. (Ugh!)

Maybe someone will attempt a more systemic cost comparison soon. . .

Great piece.

“But those investors always get filled at NBBO — the best possible price in the market — and they do so immediately.”

Wrong, B/Ds have significant leeway to fill outside the NBBO and the whole point of many HFT algos is to manipulate the NBBO anyway.

“We’re seeing the same thing with the fiasco at Knight Capital, where a highly-sophisticated high-frequency stock-trading shop lost an enormous amount of money in a very small amount of time, and small investors lost absolutely nothing.”

You’re not happy if your IRA had a stop-loss on in one of the names NITE temporarily cratered.

But yes, the LIBOR “scandal” is a joke. It more or less exactly tracks Fed funds anyway – the favorite tool of centrally planned ZIRP which will royally screw millions and millions of people in the end.

+By ensuring that the spread goes to professional traders rather than to other naturals. This was one of the primary justifications for the deregulation of the specialist market. Intermediation rates used to be in the high teens or low 20s. Today, they’re 50% or more. Pragma Securities has a research note on just how much this costs institutions, and, by extension, the public. See http://www.pragmatrading.com/sites/defau lt/files/pragma_commentary_hft_and_cost_ of_deep_liquidity.pdf;

+By using the enhanced data feeds exchanges supply to HFT firms so HFT firms can trade off of order information. To be sure, other firms buy these feeds, but they were originally designed in hand with HFT firms, and these firms well knew they could use them to predict very short-term price movements and profit from them. This style of order information trader is just an old-fashioned scalper, often takes more liquidity than it supplies, and is a tax on all investors. For more details on just one strategy using these feeds, see http://papers.ssrn.com/sol3/papers.cfm?a bstract_id=2004231. See also https://www2.bc.edu/~taillard/Seminar_sp ring_2012_files/Hirschey.pdf;

+By generating excessive and largely meaningless quote traffic, socializing many costs of HFT strategies while of course keeping HFT profits private. This is different from deliberately manipulative strategies like “quote stuffing” and is simply the ever-increasing technological cost of handling HFT chaff. See http://blogs.reuters.com/felix-salmon/20 12/08/06/chart-of-the-day-hft-edition/;

+Through manipulative strategies like quote stuffing, which, when inadvertent, at minimum cause collateral damage throughout the market – and when deliberate surely are used to pick pockets. See http://papers.ssrn.com/sol3/papers.cfm?a bstract_id=2066839;

+By introducing another element – time at the millisecond level – as a complexity and game variable used to extract rents from longer horizon investors. See http://papers.ssrn.com/sol3/papers.cfm?a bstract_id=1924991.

+By introducing higher levels of volatility. See http://papers.ssrn.com/sol3/papers.cfm?a bstract_id=2022034.

So there are many costs passed on to longer-term investors, large and small, either directly to anyone who manages his or her own assets, or indirectly to anyone who invests through an institution.

When we are talking about the complex financial instruments of the sort that involve LIBOR and HFT, they are zero sum gains. To the degree that banksters win, the rest of us lose, even if just by chasing capital from constructive uses to the big casino.

In the case of HFT, it is even clearer: HFT is front running on a microsecond scale. They execute short term trades and collect a toll on the rest of us.

In both cases, there are simple dumb regulations that can ameliorate the situation.

In the case of HFT, requiring trades to be executed, and requiring that the trade duration is longer, and in the case of LIBOR, requiring that loans be cleared on a government operated exchange with immediate public disclosure.

Good regulation is easy: You simply have to make it dumb. The banksters will cheat on the smart stuff, because they CAN, and because their bonuses depend on it.

Assuming that was the case. Sometimes it was low, sometimes not, depending on the goal the traders had for the day. The strategies shifted over time. But obviously, _someone_ was getting cheated, even if it wasn’t always Joe Average Homeowner.

Not quite sure where to look for trading volumes, but I get the impression that it is on the order of 1 trillion shares per year. Seems the “tax” of HFT is less than a penny per share. Less than I typically pay in commission on a trade.

I’m not excusing or defending HFT, just suggesting it isn’t that big a deal. Unlike the estate tax.