[PanMan]

I just read Flash Boys[1], which is a fun read, and explains how being faster than others really helps in making money (on the expense of all other traders). Basically: If you buy stocks, the trade is executed on multiple exchanges. When the order arrives at the first exchange, it shows your intent to buy these stocks. If a HFT firm buy them at other exchanges quicker than your order arrives there, they know they can sell the stocks to you, and make (a tiny tiny) profit, at your expense.

1: https://en.wikipedia.org/wiki/Flash_Boys

[kasey_junk]

> If a HFT firm buy them at other exchanges quicker than your order arrives there,

The crux of my complaint about Flash Boys is that it leaves the reader with this impression, which is usually an incorrect representation of what is happening.

A better simplification of what is happening is that the HFT firm is racing to update the price on shares they are already offering (or on buy orders they've already placed).

That is, they are more like a shop keeper changing their prices when they notice demand than a ticket scalper racing you to the ticket window.

[Lawtonfogle]

So like if the ticket scalper was the one running the ticket booth?

Is there really a difference between:

I see you want X, so I go buy X and sell it to you at a slightly higher price to make some money.

I see you want X, so I take my X out of stock and sell it to you at a slightly higher price to make some extra money.

I'm not sure they are really different.

[fisherjeff]

Maybe I'm misunderstanding but my impression is it's more like the scalper's standing off in the distance, waiting. Once they see someone walking up to the booth, they race in to buy up the last ticket at the current price and sell it to the person who was about to buy that same ticket, but now for a little more money. Then rinse and repeat all day long.

Again, I may be misunderstanding, but if this is truly analogous, I completely fail to see the utility.

[tptacek]

HFTs don't corner the market for a particular instrument. So what's really happening is more like the market for limited- release sneakers. They rush to the store, they buy up a bunch of sneakers. Now they're holding inventory. They only make money if they can sell that inventory off for more than they paid for it.

But there are thousands of other places where the same shoes are being sold. If the demand for the particular sneakers they bought climbs, they will make money. But it's just as likely that the demand will fall (in fact, it's much more likely that demand will fall in the instruments HFTs trade --- that's what those instruments do! They take random walks!), in which case that inventory they bought is a liability.

[kasey_junk]

What about this. You just bought all of my stock of X. I have no idea if thats all you want or why you want it, but there is now less X available.

If you think the price shouldn't be impacted by that, it means something very fundamental about markets.

[Lawtonfogle]

So if X are tickets, and my plan (which you don't care about) is to now sell the tickets at 125% of the price I paid... are you saying scalping should happen?

[tptacek]

I think scalping is a really bad example of what HFTs are actually doing (see https://news.ycombinator.com/edit?id=10981812).

But scalping is an interesting example for another reason. Ticket scalping is a response to an extremely inefficient market. Scalping can only be profitable if people are willing to pay more than face value for a ticket.

It might make sense to restrict scalping for tickets, because artists might want to make tickets available to people across a wide spectrum of means. But that is never the case for tradable instruments. Restrictions on scalping deliberately seek to impede price discovery. The whole point of liquid trading markets is that the price of the instruments being traded is supposed to reflect their actual value.

[sjg007]

No, the HFT buys the cheaper shares and then sells them to you at a higher price.

[kasey_junk]

We could probably keep going round and round saying "no that's not how it works" and appeal to our technical expertise in the area in question, but instead lets argue it from first principals.

On the one hand we have an ultra fast HFT that is just sitting there trying to do only latency arbitrage between 2 venues. On the other we have an ultra fast HFT that is making markets on multiple venues. For the first HFT the upside to their strategy is that they can hold very little inventory. Of course they are not going to be able to buy orders that are already at the correct price. That is, orders that could have been put in place seconds, minutes, days or weeks earlier are going to have time priority regardless of how fast the pure latency arb player is. They are also going to be wrong some percentage of the time. Meaning they are going to have inventory they need to unload and all that entails.

Meanwhile the market making HFT has more inventory risk, but they also get some of the latency arb for free, as they are already at those positions. They get some of the latency arb the same way the pure arb player does (ie they are super fast as well) and when they are wrong or lose the race they also have sophisticated inventory management processes in place that they amortize across all of their strategies and not just the latency arb ones.

It turns out that the second model is more profitable, and that is very very important when you are investing in super low latent bespoke networks as part of your operational model.

[sjg007]

Broker sees n shares of x for market price y and wants to buy them. HFT intercepts order and buys at y and now resells to original buyer at price z>y.

[kasey_junk]

Or I guess we could do it the other way. No you are wrong. That is not possible on any exchange that I know of.

Luckily there is an easy way for you to prove your contention. The market data feed and order management specifications for all of the exchanges in the US are available as public information. Find a single one that allows that order of operations.

Until then maybe don't spread lies.

[shiven]

Wait what? I thought there were two feeds. One for the paid subscribers in near real-time and a purposefully delayed feed for the rest of us.

In order to prove what you are asking for would need full access to the realtime feed AND corresponding time-resolved data from multiple brokers where a buy call gets intercepted.

Where is the lie in what the GP wrote? A quick Google search supports what he wrote, eg. the RBC story.

[infecto]

Makes no sense. I only know of one exchange where it was possible to jump the line (NYSE) and that strategy is long gone.

Problems I notice. 1) How are you jumping ahead to buy at Y? 2) How do you know someone is willing to pay at Z.

I guess if in your scenario the spread is larger and there are outstanding orders at Z but the way you explain this problem is incorrect.

[ctlby]

It turns out that the only way to be profitable in trading is to buy low and sell high.

[ctlby]

Speed wins regardless of the number of exchanges. Even with a single venue, the first participant to execute a trade captures the embedded value and removes that opportunity for all other players. As a corollary, in any given market, only one or two firms can successfully run such latency-arb strategies. Most HFT strategies do not do this.