Comment by vlehto | Hacker News Reader

Comment by vlehto | Hacker News Reader

High frequency trading gets shat on frequently. It's stupid. This is only a tool that evens out price differences across geological distance. If someone buys significant amount of AAPL stock in New York, the price goes up in NY. Then someone from London can profit by buying small amount of AAPL stock in London and selling it in NY.

High frequency trading is this, but with very high speeds and lots of competition. This means they have to exert value out of almost infinitesimal price changes.

>As HFT strategies become more widely used, it can be more difficult to deploy them profitably. According to an estimate from Frederi Viens of Purdue University, profits from HFT in the U.S. has been declining from an estimated peak of $5bn in 2009, to about $1.25bn in 2012.

That's 0,007% of U.S. GDP. And probably going down.

>"DEFINITION OF HFT: 1. a strategy which trades for investment horizons of less than one day 2. a strategy which seeks to unwind all positions before the end of each trading day thus HFTs can't deploy large amounts of capital"

http://www.sec.gov/comments/s7-02-10/s70210-129.pdf

Oh, and another round of "we HF-Traders are just misunderstood benefactors of the general public and helping markets being more efficient".

There are heaps of evidence HFT ruins markets, economies and nations, so I call BS.

http://www.zerohedge.com/news/2015-03-20/how-hft-destroys-ma...

I'm not really knowledgeable to have an opinion on HFT, but there's something I'm still not clear about.

HFT firms are not that big; the ones I've seen pointed as the largest have net incomes of less than $100M. Why wouldn't the other market participants, particularly the largest Wall Street firms, allow HFTs to ruin their game? The link you posted mentions "corrupt SEC regulators", but why would they be corrupt in favour of HFTs?

Large Wall Street firms view these guys the way a rhinoceros views the birds snatching bugs off its back. It's not a competitive relationship, they coexist.

To push the analogy, sometimes the rhino gets annoyed and pecked too hard, but overall it's happy to live without as many bugs everywhere.

>there are heaps of evidence

>links to zerohedge

credibility = destroyed

Whenever I see a discussion like this a HFT-supporter, like yourself, will usually come along and say that any links which are made to suggest HFT is bad are terrible (often the book flash boys is mentioned), but I never see any evidence of this terribleness, or alternative information sources. Do you have any?

Actually yes, Chris Stucchio's posts on the subject are pretty in depth starting with https://www.chrisstucchio.com/blog/2012/hft_apology.html

You could also look at the book, Flash Boy's Not So Fast, which takes apart, page by page, all the arguments in the book, Flash Boys

http://www.amazon.in/Flash-Boys-Insiders-Perspective-High-Fr...

Sorry to piggyback on your point and rip off Matt Levine a little bit, but most of the cage rattling seems to ultimately come from people who are mad that it's harder to slip in their huge orders without having a price impact like it should.

Sorry, I'm on my phone, but you could start with "Trading and Exchanges" which is available on Amazon (really, any of the highly rated books on market microstructure are pretty good IMO). Also IMO, It would be helpful to use traditional market makers as a baseline when making any kind of comparison to HFT, and asking if it's really worse. For example, total HFT profits have declined from like $5 billion to a little over $1 billion per year, whereas vanguards fees alone allow savers to keep roughly $40 billion per year. There are also a multitude of interviews with large, institutional investors who are extremely sophisticated and who are arguably the main people affected by HFT activities, for example this interview with cliff asness of AQR, which manages $100 billion ++ in quant strategies. https://soundcloud.com/bloombergview/masters-in-business-aqr... I think he starts talking about it around the 30min mark.

Also, my own experience working on a trade floor as a source.

I skimmed some of the evidence there. Nothing seems to condemn the HCF industry as a whole. Some linked articles about it in Wall street journal and Financial Times are behind paywall. And even if HFC would be bad, it's not significant. Your source also says HFC can't hold much equity.

You could have explained me why HFC is bad. Just slamming around some link with lots of "evidence" is not very convincing. Given how easy it is to find evidence about anything these days. It's bit unfair to assume I should gather your argument for you.

Themis trading wrote a report saying HFT is evil? This is about as persuasive as the taxi industry releasing a document saying that Uber is worse than Voldemort or the hotel industry saying AirBnB is the devil.

Furthermore, many of the issues described there are fairly innocuous in reality.

For example, flickering (aka "quote stuffing") is a software bug which costs HFTs money: http://zacharydavid.com/2014/04/on-hft-part-ii-bugs-features... https://www.chrisstucchio.com/blog/2014/quote_stuffing_is_a_...

The issue of HFT being a zero sum game in the race for latency is predominantly caused by the subpenny rule: https://www.chrisstucchio.com/blog/2012/hft_whats_broken.htm...

Bollocks. Given fund managers are obligated to obtain "best execution" under MiFID (and regulations in general) if HFT were impacting that then they would raise merry hell. The fund management industry is orders of magnitude larger and more influential than all HFTs.

>There are heaps of evidence HFT ruins markets, economies and nations, so I call BS.

HFT itself isn't "bad", period.

Does it have bad actors, stupidity, corruption, inefficiencies and is sometimes a waste? Yes. What industry is immune to this?

Immunity isn't what you should be concerned about. Incentives are the issue here.

Yes the value of notional sizes of derivative trades have little to do with either their actual values (which net to zero for all the parties involved), their risk or the amount of capital actually involved.

Latency arbitrageurs do not provide the value you claim they provide.

They consume liquidity by identifying that one individual is willing to buy at $10.02 at point A and one who is willing to sell at $10.00 at point B. These individuals would discover each other naturally in (literally) the blink of an eye; but HFTs pay to colo everywhere, they pay for 40Gb connects and ASICs. They pay for the fastest possible connections between those colos.

This means that instead of those individuals who truly create liquidity and who truly drive the market finding each other; the HFTs are able to wedge themselves into the middle of the transactions. This works because the NBBO system is slower than their systems. It works because of complex order types that benefit only HFTs (no actual investor requires a hide not slide... that's just a tool for HFTs to cut the line).

HFT does not create liquidity. It does not even out price differences. It is an entirely predatory tool to extract value that was generated by other, better men.

HFT is a place where self-righteous nerds pick everyone's pockets.

HFT is useless trash.

The only place I agree with you is that HFT-based theft, like car radio theft, is a relatively minor crime and one that is declining over time.

True market making is often algorithmic and automated, and it operates quickly. But it is completely different from the nonsense that is HFT.

> High frequency trading gets shat on frequently. It's stupid. This is only a tool that evens out price differences across geological distance.

1. That's a valuable evening out prices may be a valuable function, but one reason HFT gets shat on is because of the enormous effort put into attempts to arbirage over shorter and shorter time intervals. How does society actually benefit if the prices are equalized over a 1ms interval vs a 1s interval?

2. There are other, more parasitic HFT strategies than arbitrage. One I read about consists of the HFTs putting small tripwire sell orders on some exchanges to detect buy activity, then front running big orders by buying up all the stock on other exchanges and offering it at a higher price. The HFT performs no function besides inserting themselves as a middleman. It's like being on Amazon, clicking buy on a $10 item from seller A, then either having your order fail or be filled by seller B (who themselves just bought from seller A) at a higher price.

https://en.wikipedia.org/wiki/Front_running

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

front running is illegal, and flash boys is a book packed full of ignorance. there is no outrage over one consumer buying a house or an apple ahead of another. there's no demand for "fair" auctioning of toyota corollas. speed isn't illegal, and it's been a determining advantage in markets for eternity.

If Flash Boys is full of ignorance, perhaps you could recommend a better book?

It also sounds like the term front running refers to many different practices, this is from the Wikipedia article:

> One common practice of high-frequency traders (HFT) is a form of front running, where they peer into various exchanges and try to detect orders as they propagate from a broker's order router.

> HFT traders place many small orders that indicate buying/selling pressure. Those with the shortest lag in reaching other exchanges then place orders on those exchanges to catch the rest of the order, at a more advantageous price.[6] According to Harvard Political Review writer Austin Tymins, HFT hedge fund Citadel LLC made billions of dollars front-running the trades of large institutional investors, many of which are investing on behalf of middle-class clients.[7]

> If Flash Boys is full of ignorance, perhaps you could recommend a better book?

http://www.amazon.com/Flash-Boys-Insiders-Perspective-High-F...

In contrast to other Michael Lewis fare, Flash Boys sounds poorly researched and based of the anecdotal experience of about 3 people.

You'd have to do a _lot_ of buying on other other exchanges to move exhaust the order book to the point that you can make a profit, and there is no way of knowing that the "buy activity" that your "tripwire" detects is, in fact, a harbinger of larger orders on the way to other exchanges. It could just be a single buy order at your tripwire exchange. You have no way of knowing.

Read "Flash Boys: Not so Fast" for a highy detailed rebuttal of Flash Boys. It's written by an HFT trader, so it's not intending to be unbiased, but from a logical perspective it does seem to demolish Lewis's points.

http://www.amazon.com/Flash-Boys-Insiders-Perspective-High-F...

Oh give it a break. If you had to rest orders for a second, then HFT would evaporate.

That would push spreads apart a bit (increasing the costs of trading for everyone) but it's still clear that someone would be need to making markets. And it would still be cheaper to have computers do this instead of humans.

It would not change anything significant about HFT besides removing the H. There would still be a technology race to execute trades as close to exactly 1 sec as possible.

HFT firms make money by being the first to grab trades for tiny price deltas. If the law said trades could only execute on one-second intervals, the deadline to be first to grab a trade would become a tempo defined by human law, instead of a steady lag due to physical law. But there would still be a deadline, so companies would still race.

Most of the firms I worked for rested orders for days or weeks. Having to rest orders for a second would not materially change their trade other than make them slightly less likely to get into the top of the book, thus increasing the spread.

Reply on news.ycombinator.com