Hacker News Reader

Comment by codeshaman | Hacker News Reader

"... millions of trades a day".

What is that ? How can that be the reflection of what's going on in the 'real' economy ?

I interviewed at an HFT firm once and they were doing kernel-bypass networking, latency measured in nanoseconds and all kinds of really low level systems programming tasks.

Interesting technical challenges. Except... what does that have to do with the economy ? Why is it that I can make money if my algo runs a microsecond faster than yours ?

Why is that time delta measured in money ? I know there's the very rational technical explanation of how the first to react executes the trade, but just step back and look at the bigger picture here - what the fuck are you guys really doing ?

Are the people writing these algorithms even remotely thinking that the trades can ruin or starve millions ?

Are you even considering that these high frequency "trading" systems have an actual impact on the planet, the air, the water, etc ?

Of course no. They don't give a fuck. In fact, you can only survive in this industry by not giving a single fuck. Otherwise you have to look at yourself and realize what a parasite you really are.

The stock market made sense when securities actually reflected the "real world" performance of companies.

Now it's all reversed - price determines company performance - it's like all companies work for the stock market itself.

I don't like these guys. A bunch of crooks and gamblers who have no idea what they are doing, while everyone thinks they do.

That's why the prognosis for the world economy is pretty bleak - because the insanity is institutionalized already and we have no idea how to stop it.

View on HN

Millions of trades a day each contain teensy, tiny little updates about the state of the economy. Think of them like tweets, not like State of the Union addresses. This married couple wants to send their daughter to college more than they want Google. That hedge fund likes Coke more than Apple. This pension fund doing it's weekly buy of an index to prepare for 2075.

The market is these little packets bouncing off each other and summarized. High numbers of them are not unhealthy. On the contrary, it means more information is moving with less overhead loss to middlemen. (By analogy: What's more expensive, 100k tweets, 10 phone calls, or a FedEx envelope? If you say "clearly it's the tweets because there are so darn many" that would be a weird answer for an engineer to believe.)

If you believe the world economy is going to heck in a hand basket, there exist ways for you to be richly compensated for that expert opinion by sending your tweets into the machine. (Well, if you're proven right.) HFTs will happily minimize the transactional costs you incur while making that statement, as compared to the sweaty, yelling colluding-against-you jocks they have replaced.

All that you said is true.

But that which you control, controls you back with equal force.

While it is assumed that the state of the stock market reflects outside reality, it is in fact both ways - the state of the stock market affects and changes outside reality. And now it's a feedback loop, which involves software on one side and humans and "natural resources" on the other.

So now the married couple's daughter future is dependent on the outcome of this "game of trades" that HFT algorithms play day and night. And the people who write those algorithms do so to satisfy themselves, they don't know or care about the married couple.

Just like Google Maps - when there are few drivers using it, it's a handy little helper, if everyone uses Google Maps, it now defines the traffic pattern.

If markets are about information, they are, then, an information channel.

As Shannon told us, information is the resolution of uncertainty, and information channels can be more or less noisy.

It occurs to me that it's possible that adding more data to the channel is making the channel more noisy not more informative.

But what is the social benefit of the stock markets being synced to the state of the economy at the micro-second / nano-second time scale ?

That's surely taking things a little too far. Nothing in the real economy changes that fast.

There's probably not much social benefit to getting pricing resolution down below 100ms.

However, there is an enormous benefit to getting pricing resolution down from the double-digit seconds level it was at when humans were able to keep up. The market driven by human intermediaries was crooked as a barrel of fishhooks. When these discussions come up, I always urge people to Google [odd eighths scandal] for an example of what things were like prior to electronic trading.

Once you have electronic trading, it makes sense that there either needs to be some "figure of merit" to allow different electronic traders to compete, or else some sort of permanent monopoly for the most successful initial electronic trading firm. Better, I think, that firms compete pointlessly to price instruments at subsecond granularity than that Goldman Sachs simply buys the (thankfully nonexistent) market maker version of Google.

>But what is the social benefit of the stock markets being synced to the state of the economy at the micro-second / nano-second time scale ?

It's cheaper than people doing it, and the lower the latency, the faster a trader can move his position to match changes in the market.

What is the social benefit of Twitter loading in 100ms instead of 150ms? Why is latency important for peoples' stupid mobile apps, but not important for financial transactions?

>Nothing in the real economy changes that fast.

Sure, but you don't want to have to wait for an earnings report to sell your shares when you want to buy a house or something. You want people to always be trading your stock so it's easy to buy and sell. Similarly, when a big pension fund buys or sells a lot of your stock at once you want "people" trading to reduce volatility.

Dude its not 100 ms we're talking 100 microseconds.

> But what is the social benefit of the stock markets being synced to the state of the economy at the micro-second / nano-second time scale ?

This sort of HFT smooths out prices over this time frame. The thing is, as it smooths out the peaks and valleys, the value in doing so decreases as those peaks and valleys are what made it profitable in the first place.

>That's surely taking things a little too far. Nothing in the real economy changes that fast.

Why? Computers have changed the game. We went from mailing our brokers or visiting them in person to calling them to conducting trades over the Internet. The 'real' economy changes quite quickly.

Market makers can have actual value by reducing bid/buy spread.

However, I don't think reality supports the idea that HFT are adding real signals to the market. HFT software has little understanding of the 'world' so they add noise which reduces signal. Taking the markets further from reality.

See: flash crashes. They don't actually have anything to do with the wider economy just feedback loops devoid from reality.

PS: Some software is designed to automatically read and interpreted news, but they often get it wrong moving the market in the wrong direction.

Flash Crashes also have no effect on the wider market. When you look at end of day prices for various stocks and indices, there is only a single symbol where you can actually see the effect of any flash crash:

https://www.chrisstucchio.com/blog/2012/flash_crash_flash_in...

Also, HFT lowers the price for people who do have information to trade with it. In this way, while HFT adds little information of it's own, it does enable better information transmission.

In other words in the short term can and will distort the market. So, the question becomes do they discover a economically meaningful price in the short term or is the processes essentially random.

As to lowering the price, arbitrarily increasing or decreasing the price is not an economically useful activity. The value of markets is price discovery, at best HFT reduce transaction costs though this is somewhat debatable as they are extracting money from somewhere.

You should read the article to see which symbol actually displays the effects of HFT shenanigans by end of day. (Hint: it starts with K, and it deserved everything it got.)

To see how cheaper liquidity directly enables speculators, read the section "Why Speculators Need Liquidity" in this post: https://www.chrisstucchio.com/blog/2012/hft_apology2.html An example is worked out in detail illustrating the point.

Also, HFTs are extracting far less money than the humans they replace. So if extracting money is what you want to prevent, congrats.

This paper studies the effect of a high-frequency market maker entering the Dutch stock market: http://www.amf-france.org/technique/multimedia?docId=workspa...

The trader they study reduced the bid/offer spread and trades 80% passively (he is primarily providing liquidity in the order book for others to trade with). His limit orders are more informative about "hard" news such as changes in the index future price than other traders, meaning he tries to buy stocks that are cheap and sell those that are expensive relative to the overall market, steering them toward their correct prices. This is just a simplistic HFT from almost a decade ago, modern prop traders have teams of researchers building prediction models with many more inputs.

And for all that, he earns a realized spread of 0.72 basis points, before paying for computers and employees. 1 basis point is 1/100 of 1%. That's in 2008, a time when HFT was relatively new and competition was less fierce, margins are considerably lower these days, not like making 0.72 basis points is exactly robbing widows and orphans to begin with. No human would make a market for such low margins, no human could react nearly instantaneously to changes in the index/stock prices to set his quotes, and no human could trade hundreds of stocks at once.

How do you think HFTs cause flash crashes? If you look at the middleman's net position on page 3, he never accumulates more than 10000 shares in one direction, and his position mean reverts every few minutes. How could a trader who primarily trades passively (others choose whether to elect his orders sitting on the book), never accumulates a significant directional position, quotes both buy and sell prices on the order book, and who aims to flatten his trades relatively quickly (whatever selling pressure he could create is quickly turned into buying pressure after he's given a short position) have such a large impact on prices?

I don't think trading off news feeds is a typical HFT trade. Some market makers may use these feeds to get out of the way during announcements. Maybe it's an automated trade for hedge fund types, but it's a high-risk, high-reward infrequent type thing. If a trader is incorrectly predicting prices long enough, they'll lose money and go out of business.

The crucial aspect is the "crunch per bit" of HFT. We know that an electronic system is mostly an improvement over having pit traders yell at each other all day.

What we do not know and cannot prove without experimenting is whether our electronic markets benefit from following smarter, more complex rules of play instead of the same ones faster.

> Millions of trades a day each contain teensy, tiny little updates about the state of the economy. Think of them like tweets, not like State of the Union addresses. This married couple wants to send their daughter to college more than they want Google. That hedge fund likes Coke more than Apple. This pension fund doing it's weekly buy of an index to prepare for 2075.

Can you explain this further in context to the parent comment? You lost me after "married couple wants to send their daughter to college more than they want google".

I think the gist of what's being argued here is that the increased volume of trades (facilitated by HFT's drive to minimize transaction time) allows more 'real time' information about the relative value of instruments to be communicated.

That said, the first example quoted don't really make a lot of sense, in my opinion. A couple selling off some stock to pay for their child's tuition probably won't ever have their transaction directly represented in the market (it will likely be lumped in with or divvied up from other orders handled by the same brokerage), so the information conveyed is ultimately somewhat diluted.

It's no more an update on the state of the economy than a gambler making a bet on a betting exchange is an update on who is going to win a football match.

It's simply another mechanism for gambling.

HFT also rarely involves value investing (how much research can you do in a millisecond?). It usually uses low-to-zero-profit market-making strategies as cover for higher profit less than ethical trades.

HFTs front run the trades.

No, they don't.

You and patio11 are smart enough to understand that all the HFT strategies are equivalent to front running. Each and every HFT strategy exploits some aspect of front running. In this world the market is by definition a misnomer because it does not exist on a single exchange and is fragmented.

Sure you cannot front run a single exchange when every buy and sell is on that single exchange but that's not how the actual market and front running works.

HFT requires that the market is distributed. You can say this is market maker arbitrage but it is actually equivalent to front running the market the same way a retail to broker initiated trade can be front run.

Market makers are highly unlikely to "ruin or starve millions". Consider some graphs of the biggest HFT screwups ever:

https://www.chrisstucchio.com/blog/2012/flash_crash_flash_in...

I have no idea why you think an HFT system is somehow more polluting than an ordinary server displaying cat pictures. Could you explain?

Insofar as the stock market doesn't reflect "real world" performance that's primarily due to tax induced distortions. That situation is only getting worse as businesses go global, and the crazy US system of taxing foreign income turns double taxation into triple taxation.

> A bunch of crooks and gamblers who have no idea what they are doing, while everyone thinks they do.

You've just done a huge rant on how they have a smooth-running machine destroying the economy, making billions, being ruthless, and now you say they "have not idea what they're doing". Sorry mate, but these people know exactly what they're doing. In fact, most of them are really fucking smart, and that's why they're able to get away with legally extracting tens of billions of dollars from the markets.

10s of billions? That's a bit optimistic. The best HFTs, of which there are a handful, have revenues of roughly $1 billion. That's revenue. Their net profits are usually in the $10 to $100 million range. (See for example, Virtu's recent IPO filing.) These are successful global businesses, but they're hardly outliers in that crowd. Hell, Twitter had more revenue last year.

It's amazing to me that people get so worked up about HFT when the same arguments apply to 90% of "tech" companies.

> Are the people writing these algorithms even remotely thinking that the trades can ruin or starve millions ?

No they can't, by definition, high frequency trades have very small price deltas and each one matters very very little to the broader market.

An abstract way to think of HFT is that it condenses money out of information. This is exactly what ad-driven businesses like Google and Facebook do too, but HFT doesn't require massive databases of information about all of us, plus it is more energy efficient because it doesn't have to send the same copy of a cat GIF over the air to 350 million people just to make half a cent in ad sales.

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

> 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.

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.

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.

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.

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.

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?

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.

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

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.

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.

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.

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.

Depending on the exchange there can be many feeds that are price differentiated by feature set (including speed requirements). There are no non-paid feeds, if you are getting market data it is being paid for by someone.

In particular though, most exchanges follow a pattern where market data is broadcast (usually udp) and order management is bidirectional unicast (usually tcp/ip). On the order management side, no one sees your order until after the exchange does. The exchange then executes the order (filling it if there is match on the other side, or adding it to the order book). It then propagates down the market feed side the outcome of that order. Only then do other participants see it, regardless of how fast they are. There is no opportunity for the fast operator to get in front of the order as the GP describes.

You can verify this easily for any exchange you like (at least SEC regulated ones in the US) by reading the technical specifications of the electronic trading platform.

I don't know what google search or RBC story you are talking about, but in Flash Boys for instance, they make pains to imply that HFT are front running orders as the GP describes but they never say it outright. Because they know it can't happen. A good rebuttal to Flash Boys is Flash Boys, Not So Fast.

A very interesting rebuttal to "Flash Boys, Not So Fast" is the many factual errors as pointed out in the top 1-star Amazon review of that book: http://www.amazon.com/Flash-Boys-Insiders-Perspective-High-F...

I'll credit you for indeed finding something "interesting", but there's a telling comparison to be made between the criticisms in this Amazon review and the criticisms made in Kovac's book.

The Amazon review is based on the premise that Kovac doesn't know what he's talking about at all (that's a tough sell, since Kovac runs a trading firm), and builds that argument through fuzzy error claims like "Kovac got the number of trading venues at a particular date wrong" (the whole value of Kovac's prop firm was that it traded in zillions of markets) or "Computerization was less impactful in bringing down spreads than decimalization".

Kovac's book, on the other hand, is hundreds of pages of page-by-page technical criticism of the Lewis book.

This reviewer was supplied by Kovac with a mountain of rebuttable claims. What they came up with was "Academic and industry researchers have certainly found compelling evidence of high frequency trading firms front-running demand". But "front-running demand" isn't a technical term; it's an advocacy label coined by people who oppose algorithmic trading!

When your mic-drop closing criticism, the one you expand upon in your "updated" critique, is a tautological restatement of the position you're advocating for, you're doing something other than technical criticism. Which is fine, but let's be clear about it.

It's front run between exchanges or from dark pools.

Also the RBC story is the central story of Flash Boys.

The different feeds are irrelevant in this case, because no exchange sends (to any feed) information about: a) incoming immediately-executable orders, or b) unexecuted portions of an order which will be routed to another destination. The first anyone will hear about an incoming order is when it executes, and the first they'll hear about the routed portion is when it executes somewhere else.

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.

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

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.

It can be solved by an exchange that makes trades every second or on some time interval that is minuscule to a person but enormous to a computer.

People seem to defend HFT for some reason (the front running kind, not the necessary arbitrage kind) but they never seem to answer the question of why someone who is not an HFT firm would want to trade on an exchange that caters to high frequency traders (as basically every exchange does).

> People seem to defend HFT for some reason (the front running kind, not the necessary arbitrage kind)

I'm not defending the front running kind. The argument is that the front running kind (which is in no way actual front running and is instead latency arbitrage) is the same as, and fundamentally a requirement of the arbitrage kind.

Further, most peoples model of what latency arbitrage actually is, is so flawed as to create strong biases when they shouldn't exist.

"latency arbitrage" ?

The euphemism HF traders use to help themselves sleep at night.

Thats one way to describe it. Another way to describe it is "hey prices at 2 different venues get out of whack some times, when that happens there is an opportunity to make money by bringing them back into alignment".

We don't assume other kinds of middle-men require sleep aids when they do this kind of behavior (and in fact we usually encourage it), so I don't know why HFT folks would need them either.

And confidence tricksters perform trust arbitrage.

This is all just ways of saying HFT is exploiting a loop-hole in a system to make money for themselves.

They could all be eliminated overnight and the free-market system would not notice their absence at all.

They don't provide a service to anyone besides themselves, they don't provide liquidity, they are nothing except parasites.

This is a little like saying the CAP theorem is a euphemism that distributed systems designers use to help themselves sleep at night.

... they never seem to answer the question of why someone who is not an HFT firm would want to trade on an exchange that caters to high frequency traders (as basically every exchange does).

There are a number of incorrect assumptions built into your comment (most notably that prop HFT are the only ones using algos to trade), but the simple answer to this question is that those exchanges have larger displayed size at better prices.

> most notably that prop HFT are the only ones using algos to trade

Can you point out where I made that assumption?

>It can be solved by an exchange that makes trades every second or on some time interval that is minuscule to a person but enormous to a computer.

I do not see how that will solve the problem. Suppose firm A and firm B both want to buy stock X at prize Z. There is only enough X at prize Z for one of those firms. Both firms send their trade orders withing milliseconds of each other. In your time interval scheme, which firm gets to buy the stock?

CyberDildonics is basically proposing frequent batch auctions as described in the Budish paper. It won't work, largely for reason you state. The appropriate solution is for exchanges to add a _random_ delay to each order (emphasis on _random_, a fixed delay a la IEX does nothing). ParFX is doing just this.

Random delays already exist. The infrastructure of every exchange introduces them. Its a standard part of any HFT model to think about what happens when you hit one. It certainly doesn't remove the speed game.

I think a better way to change exchanges is to dramatically decimalize the price levels. Right now the difference between 2 price levels a) adds a floor to the minimum spread and b) prevents strategies from truly competing on price requiring them to compete on time.

I think a better way to change exchanges is to dramatically decimalize the price levels.

The idea that reduced tick sizes would force traders to compete on price rather than time is a popular HN market microstructure solution, but it's wrong or at least partly wrong.

1) Liquidity taking strategies will still "compete on time" in order to trade at the most favorable prices. Market making strategies will still "compete on time" in order to avoid getting picked off.

2) Inverted fee venues and midpoint order types offer sophisticated traders the opportunity to trade at prices within a 1 penny spread. If you watch the tape for a stock like ZNGA you will see this type of trading.

3) There are a number of high priced / high volume stocks that don't trade at a penny spread. GOOG is a good example. Do you find GOOG trading to be somehow cheaper, more orderly or efficient than e.g. MSFT? The GOOG spread is almost never a penny and is regularly over 20 cents. If the benefits are there, they ought to be obvious.

First, let me make it clear, I don't view market participants competing on time as a problem and physics means it will always be a component of the price of the risk associated with trading. But if we are going to be making changes to markets, I'd much rather change the pricing requirements than adding esoteric random time differences, or assuming batch auctions will make things better.

The reason I like reducing tick sizes (and I mean dramatically like 1/1000th or more) is less about the bid ask spread at the middle, but rather the competition at the +1 levels. A lot of the latency advantage right now is in being able to cancel a few levels near the midpoint while leaving tons of other orders stacked.

Reducing tick sizes would make these deep stacking strategies less viable, which seems (though I have no proof) like it would make the positions of the market makers less risky in a systematic way.

As for GOOG vs MSFT, I think that is obviously explained by the much higher price of GOOG right? Nothing is going to make holding a stock that costs more less risky, even speed.

I thought you were making the argument that reducing tick sizes would reduce the amount of effort put into reducing latency (or reducing the overall amount of prop HFT volume), but I guess I was wrong about that. The idea that reducing tick sizes would ultimately reduce market maker risk by making certain quoting strategies less viable is interesting.

I think ultimately I'm just not being clear about something I'm not certain of in any case, but a better restatement of my position might be:

"Having some participants whose primary competitive edge is speed is not in and of itself bad, but speed advantages tend to consolidate around a few natural monopolies, which is. So if we could add something else into the mix that allows for other participants to compete outside of those monopolies, that is good. Dramatically reduced tick sizes would do that, would at times make the spreads smaller, and might even make market making less risky systematically and doesn't seem to have the downsides of other options proposed".

Do you think there are monopolies in electronic market making? I think there are tons of firms out there trying to be good at it in different asset classes, and a few that are really good at it in lots of asset classes. I'm not sure it looks like "a few natural monopolies."

I don't know if we are there yet, but it certainly seems to be trending towards consolidation. I don't have any data to back that up, just my impression from around town.

Of course that could have been due to macro trends around volatility that have reversed recently as well, I suppose.

Let me clarify: my random delay shuffles incoming orders, while random exchange delays (outside of CME iLink games) do not. Removing determinism hugely hurts speed.

The value of further decimalization seems questionable. It would in all cases reduce displayed size. Would transaction costs still decline due to tighter spreads? For the liquid lower-priced stocks, probably. For stocks whose natural spread is > 1 cent (most of the expensive stuff), certainly not. For everything else, hard to say.

Maybe a Japan/Europe-style tick increment schedule is a good compromise.

If there is more demand than supply, wouldn't you expect the price to go up?

You're not allowing it to, because your proposal quantized the market! That's what it means to batch trades and execute them at an interval: the price can't change inside that interval.

So you cancel the order, raise the price and let it go another cycle. Why wouldn't the seller want that?

Yes, but now you haven't accomplished what your proposal sets out to do, because the fastest people to the quantized order book get the best prices.

How? In every quantized unit of time everyone would be treated equally. If there is more demand than supply the prices goes up until there is not more demand than supply. This is also not even taking into account orders having ranges acceptable prices, which would add much more flexibility and granularity without needing more temporal resolution.

I don't think you've thought this through very carefully. Most orders in the market aren't "market orders"; they specify a price. Meanwhile, assume you resolve pricing in an automated auction: there's a spectrum of prices. Who gets the better prices?

If prices are given as a floor for selling and ceiling for buying they can be fit in a fair and methodical way through many different methods with any remainder being left over for the next tick. I'm surprised it is even such a controversial idea that a fair and fluid exchange can be made without resorting to a first come first serve structure.

This suggestion supposes that there is a reasonable "floor" and "ceiling" for prices in batch auctions. But in fact that spread of prices is exactly what market makers figure out organically.

If you instead mean that individual traders could input orders with floor and ceiling prices, that's something that already exists, but now you're back to the original problem of resolving whose orders execute when (inevitably) most of the orders don't match.

It's not back to the same problem, there are many ways to do it including but not limited to matching the sells with the lowest floors to the buys with the highest ceilings, working inward from the extremes of the spread. Why are you so convinced that a fair exchange is impossible?

> they never seem to answer the question of why someone who is not an HFT firm would want to trade on an exchange that caters to high frequency traders.

There are network effects in trading: people go to transact where everyone else is transacting. Because HFT is allowed on the most popular venues, the economic thing to do is hold your nose and trade there too. This is true even if HFT somehow makes the venues "worse" (which it doesn't).

If HFT doesn't make those venues worse, then how are they making money?

Makers collect the spread from participants who want to transact right now (the price of immediacy). "Real" investors are better off because they paid someone to take a position of their hands that they didn't want. They made their trading problem someone else's.

Takers exploit the option value of resting orders by trading when "fair value" moves but those orders don't. "Real" investors are better off because they cheaply acquired a security they may hold for years (the fact that the price will shortly move against them by a penny or two is irrelevant). Market makers are worse off--but they're HFT guys, and it's not really your problem.

Do "real" traders want microsecond immediacy? I would think a person would rather have a better price, even if it means waiting a whole second.

There is a simple flag you can set (it's called "Add Liquidity Only" or "Post Only", depending on the venue) which will result in HFT's never trading with you. Strangely, most "real traders" never set this flag.

https://www.chrisstucchio.com/blog/2014/how_to_not_get_rippe...

The reason is that it's not a matter of waiting a whole second and being guaranteed a fill. It's placing an order and accepting execution risk; maybe you get a fill, maybe you are stuck holding a tanking stock.

If you define HFTs as passive spread/rebate-capturing single instrument market-makers, then yes, but many prop HFTs operate a "hybrid" model where they sometimes cross the spread to flatten their book, or cross if an arbitrage presents itself. There are even HFT strategies that are primarily liquidity taking.

Obviously not. Humans just want something that feels like "now." That this has changed from seconds to milliseconds to microseconds is an inevitable consequence of the rules of the game--and is completely irrelevant to you. Your horizon is much, much longer than that of the professional trader to whom these details matter.

So a professional trader is not a person and they do care about microsecond as opposed to seconds? That sounds like a bot, not a person.

Another place people seem to get jumbled up in these discussions is in comparing institutional traders to individual retail traders.

A person buying or selling individual stocks for their own portfolio isn't impacted by HFT at all. Their orders probably never see a real exchange. Instead, internalizers pay extra money to offer brokerage customers the best possible prices for what they're looking to trade.

That's because HFTs aren't the apex predator in the markets. Informed institutional traders like Goldman are. Market making operations are built around mitigating the risk of getting run over by giant block trades from institutions. When those happen, they take out whole swathes of the order book and leave adverse price changes in their wake.

An electronic market maker will pay a premium to make safe profits off the spread of retail orders, because to a first approximation none of those orders are going to take out the whole book.

Meanwhile, among institutional traders, there are two kinds of firms to look at.

The first kind, the Royal Banks of Canada of the world, are getting paid a tidy sum of money to make large block trades for the cheapest possible price. The RBC trader has knowledge that the rest of the market doesn't: many tens of thousands of shares of something are about to trade, and that trade is going to move the market. They make money by trying to disguise their intent, so that instead of the market price reflecting that intent, they can capture the premium and leave everyone else in the market to hold the bag.

Those kinds of traders hate HFT.

The other kind, the Vanguards of the world, buy or sell based on long-term strategies. They're buying CSCO or MMM to fill out an index. They want the best possible price, of course, but they mostly care about determinism and low cost of trades.

Those kinds of traders like HFT. (You can see the Chief Investment Officer of Vanguard, the world's most trustworthy mutual fund company, saying that repeatedly and emphatically). HFTs reduce spreads and make trading cheaper, and Vanguard doesn't earn profits by taking it out of the hides of their immediate counterparties.

As a public policy matter, I'm not sure why I'm supposed to support regulations that increase Vanguard's cost of trades so that an equities desk at RBC can charge higher premiums to fleece the markets on behalf of their hedge fund clients.

All of this is misdirection to the point that high frequency traders definitely make money, but it extremely difficult to justify what they add to the market. People making decisions don't want to trade somewhere with high frequency traders and they don't need to for any technical reason.

If they didn't, they aren't required to pay for it. If they want to wait they can. Better yet, if they want the exchange to wait for them until the price gets to be what they want, they can do that as well, without paying for any immediacy at all.

when there was no HFT and market makers were humans, you had to wait a long time for your trades AND spreads were huge (aka, prices were worse).

You are making an assumption that computers mean high frequency trading. We can use computers to trade without having high frequency traders.

The same way market makers have made money since prices were literally printed on a stream of tape: by outcompeting other market makers to capture spreads.

They get paid to bridge demand over time. By efficiently doing this they can make the venues better in aggregate and still make money.

Does the demand between millionths of a second need to be bridged?

No, there is probably nothing especially productive about pricing things at that level of granularity. But speed is a reasonable figure of merit that allows different algorithmic trading firms to compete for the business of making markets, rather than having all of market-making owned by one of the big investment banks.

No, but by being able to react fast they can price what the service of bridging more efficiently, and thus more cheaply to those taking part in the service.

How is that not circular logic?

Market Makers sell the service of taking on the risk of bridging prices for a good over time. The price for that service is directly correlated with the amount of risk they take on.

HFT reduce the risk of the inventory they hold by being able to adjust the prices of that inventory fast.

HFT Market Makers can therefore price the service cheaper as their risk is less, due to the speed.

There's an unobvious upside: HFTs generate liquidity. Liquidity is good for everyone in the end.

No, they don't. In any panic, HFTs disappear like anyone else. They are middlemen, not producers. They insert themselves into the middle of trades to make profits for themselves.

If I put in a market order when the price is $100, and a HFT firm sees my order, beats me to market, buys up all the $100 offers and then resells the stock to me at $105, my liquidity hasn't been improved in the slightest. All that has happened is that someone has front-run my order and cost me $5/share, extracting profits from my pocket without providing any useful service to anyone.

If you want this comment in more technical terms, here it is:

http://www.zerohedge.com/news/2015-08-25/cutting-through-hft...

They don't "insert" themselves, they were there before you came along. Your market order trades with them because their price is the best one currently displayed. As for an HFT firm "seeing" your order and front-running--you're just making stuff up.

You are totally wrong, to put it simply. He is not making stuff up.

Flash Boys is an entertaining write-up of how the firms are inserting themselves, and how they are "seeing" orders at one exchange before the order arrives at another exchange.

There you go again with "insertion." HFT market-makers don't thrust themselves between two people who are about to transact. Their orders were already in the market, and you came to them.

The "Flash Boys" case is a funny one. The HFTs did see the orders--but so did EVERYONE else, and in basically the same instant, because all transactions are publicly disseminated. Your point is right in a very narrow sense and very wrong on every axis that actually matters.

I'm mystified as to why Flash Boys didn't valorize high-frequency trading. It's a Moneyball story where the scrappy nerds use computers and math to out-compete the dinosaurs. Oh well.

And if you put in a market order at $100 at Exchange A and the HFT beats the order to the best-priced Exchange B, buys it all up, and the instrument declines in price to $99.98, they lose money.

A lot of HFT argumentation only makes sense when you assume prices only go up.

It's not a reflection of the economy.

Capital markets are a place where firms go to acquire capital, people with excess capital buy securities, and firms (eventually) return that capital. The markets are only a reflection of the pricing of those securities, which sometimes but doesn't always correlate to what's going on in the economy.

The feeling is not shared by everyone in the IT field, some people only care about the technology part and the challenges. I agree with you and also consider this industry unethical and I'm not planning to work in it any time soon but I'm sure lots of people don't mind.

hft is about as unethical as that guy buying that apple you liked before you. speed isn't evil; it's been the deciding force in market based transactions for eons.

Except, it's not some "guy", but some stupid algo which has been given the power to ruin markets.

how on earth do algos ruin markets?

As with many other industries, humans can be come obsolete for the day to day operation, in this case the second by second changes. People still write the rules by which they operate and those rules have made the day of floor traders more and more obsolete.

I really see no difference here than the days cars replaced horses, milk delivery came to an end, newspapers struggling with the digital age. Come a generation if not less people will look back at our ways and go "how quaint". The solution is to adapt to what technology brings so that we can better ourselves. so while it may be unfair now the technology will spread to where everyone operates that way

Why do you think the stock market exist to be a reflection of the real economy? It exists to provide liquidity for selling and purchasing stocks. Nothing more, nothing less.

Being faster actually helps with that purpose.

It's really a national failure that this entire industry isn't shut down with ruthless force. Utter financial parasitism by people who don't give a damn about who is affected downstream.

Reply on news.ycombinator.com