What's your opinion about HFT - High Frequency Trading?

Just read Flashboys and am really confused.

If the book's arguments are correct then IEX is definetely a great idea, but what if it's biased? It still odd that the author promotes an exchange and represent Brad Katsuyama as the hero of the financial markets.

How good is HFT for the markets? Does it provides liquidity or does it have more negative effects?

 

Does provide liquidity, tighter bid/ask spreads. IEX is a joke and a PR show for the same firms that exploited HFT in the first place. Isn't it surprising how the first guys routing their orders through that "revolutionnary" exchange came from the firms that got bad press for HFT. Lewis depicts them as borderline heroes for investing in a shite exchange. I am a retail trader and HFT does not affect me at all, or if it does(unknowingly, most likely) the impact is negligeable. HFT has existed for many years already, but now that media has developed an addiction to it, HFT is now a problem. There are bigger problems with the markets, such as stock market all time highs as the economy is just starting to kick off again.

 

While I think talking down HFT as a whole is a bit much as it encompasses many different strategies, how anyone can support SPECIFICALLY the front running operations is beyond me. Sitting on numerous exchanges to sense orders, and then rushing to the other exchanges through faster connections to beat the rest of the order is absurd. It accomplishes absolutely nothing except giving the people who do it an obvious talking point to defend their actions. "We add liquidity, just look at how many more trades occurred with our presence". Yes.... more trades occurred because you out ran an ACTUAL order and then resold it thus doubling the necessary transactions. This didn't add liquidity as the trade would have occurred either way only at a better execution price (albeit a minuscule difference- thus why most retail traders dgaf).

Again, this isn't all HFT- so to hate the WHOLE thing is a bit much. But ppl defending the front running I don't understand*. If you're making money off it, then it's a different story as you're just tryna protect the golden goose. But 99% of ppl in finance have zero connection to it so their defense amounts to little more than "the Michael Lewis talked down on the finance!?!? I work in the finance!! Screw him!!!". Meanwhile, it has absolutely nothing to do with them. *I read it a while ago but as far as I recall he too was only criticizing the front running, not all HFT strategies.

 

Market stability in times of high duress or volatility are the biggest pitfalls to HFT. They only add liquidity up to a point. If things get too crazy, they can flip the switch on their systems and take a step back; no obligation to sit around and soak up orders. SEC is still too dumb and lazy to realize all of what's going on and so there should always be potential to make money in HFT. It's just going to get harder to squeeze fractions of pennies out of trades.

HFT doesn't affect the small at-home people, only the funds that are looking to trade huge $$$ blocks. As I mentioned, though, the biggest risk is for an extreme liquidity shortage causing overloaded circuit breakers. You still see mini-crashes all the time and a majority fix themselves right away. The big kahuna is the day when prices don't snap right back up and people start shting their pants and sell.

 

Read Flashboys as well and I think he was criticizing front running more than HFT. Additionally, I don't think he made IEX seem like a savior by any means, he just intended to talk about people who were trying to curb the front running. It just levels the playing field, HFT are in fact allowed to be associated with IEX, they will survive if they play a legit function in the markets as we know(supposedly) them.

Also reading Dark Pools by Patterson just cuz I am intrigued.

 

I want to address all of the points which you make for their existence. However, let me preface this by saying that I have mixed opinions on HFT.

1) You state the problem yourself, "exchanges are forced to cater to HFT." Being as close to the exchange as possible and having the best possible connection increases the speed of the data transfer. If the average joe approached the exchange with a pci-e connection and his computer they would never let him link up. This essentially allows those funds with the best access to 'see' information sooner albeit milliseconds.

2) My understanding is that they create a tighter spread only in a rising and sideways market. In instances, of major market collapse they have not been shown to increase liquidity and in fact can exacerbate the problem by pulling the plug.

3) It is true that IPOs would not be affected, but companies initiate buy-backs and make secondary offerings. Additionally, shareholder value is affected by the activities of the secondary market, and other forms of lending are certainly affected by equity. The secondary market is (or should be) about efficiently allocating capital. You buy a stock because you expect that the company will make money over time and the value of your equity in the company will increase as a result.

However, this is the future barring any major legal judgement.

 
slotmouth:
I want to address all of the points which you make for their existence. However, let me preface this by saying that I have mixed opinions on HFT.

1) You state the problem yourself, "exchanges are forced to cater to HFT." Being as close to the exchange as possible and having the best possible connection increases the speed of the data transfer. If the average joe approached the exchange with a pci-e connection and his computer they would never let him link up. This essentially allows those funds with the best access to 'see' information sooner albeit milliseconds.

2) My understanding is that they create a tighter spread only in a rising and sideways market. In instances, of major market collapse they have not been shown to increase liquidity and in fact can exacerbate the problem by pulling the plug.

3) It is true that IPOs would not be affected, but companies initiate buy-backs and make secondary offerings. Additionally, shareholder value is affected by the activities of the secondary market, and other forms of lending are certainly affected by equity. The secondary market is (or should be) about efficiently allocating capital. You buy a stock because you expect that the company will make money over time and the value of your equity in the company will increase as a result.

First, I’d like to say, good comments. Now let’s discuss your points:

A) Market-making HFT don’t participate directly in price discovery, instead they make-markets which reduces transaction costs and provide liquidity. Their co-loction allows them to make markets in between broker-dealer price quotes. They add price depth to the book and also some exchanges, like ARCA, offer payment to adding liquidity to the market. So if a bid/ask NASDAQ spread is $0.015 @ 500, an HFT can ping this and intervene momentarily to offer $0.005 @ 100, $0.010 @ 300, and $0.015 @100 and someone bids 500 at these prices the HFT gets a payment and the buyer has lowered their transaction costs (greater for larger volumes), thus side-stepping the specialists/broker system.

B) They’re may be some validity to your point here although I don’t see why market-making couldn’t still be used because even as prices decline, shorts will eventually have to be covered and there will be a need for liquidity in the market. What I can say though, is that in reading Andrei Schleifer’s, Inefficient Markets: Chapter 4, essentially what he is saying is that arbitragers face this exact problem of liquidity in a price deflationary equity market. Precisely, when they need the most capital is when investors are taking capital out and as a result the positions further go against them instead of being able to realize them. While there are arbitrage HFT, I don’t specifically know how many, but I would err that that it’s not endemic to HFT.

C) This is along similar thinking to my other question that maybe you can answer, if HFT competes on noise (spreads) than does this effect the signal (valuation)? I tend to think, so far, that any effects at this point are instantaneous at best but if you're a student of Chaos Theory, this could have some type of significant unforeseen impact upon a single security/s market. Research by Carran, Hunsader, Johnson, Meng, Ravinder, Tivnan, and Zhao, "...found 18,250 ultrafast black swan events that were uncovered in stock price movements between 2006 and 2011."


Just:
I do not understand Chris and his HFT support. "HFT is not breaking the markets" but it is not about the markets, it is about the people, businesses, investors. Are you talking about the market as something virtual, electrical or mechanical, non-existent in human world? The markets do not exist without investors and businesses. America is a society of business owners. Asset ownership is evident through independent businesses, direct market participation or over 401K holdings. Why should investors who own stakes in US businesses tolerate some short term price takers who do not have vested interest in such businesses but are the ultimate source of risk not associated with true riskiness of owned asset? Why do businesses with their assets and human capital should tolerate such culture if they do not operate at the speed of a high frequency trading? Why don't all these high frequency traders take their beloved models and algorythms and find ways to please their wives or educate their kids? This is an absolute nonsense to assume that all the benefits of HFT culture for the market are good for society.

My article is simply addressing a book that argues against HFT as an execution system that's supplanting an older one, not replacing speculators or hedgers which algorithmic trading has been doing for some time now. In my opinion, in think its shear economics as I tried to convey in my first question, is this simply creative destruction of the human market-making system? (see above response) I’m merely trying to show that Arnuk and Saluzzi made some weak claims and that it’s a status quo argument.


Here is part two of Chris Stucchio’s article on HFT, How HFT's Make Their Money.

Who Am I? | See what GMngmt is all about at About.Me
 

I do not understand Chris and his HFT support. "HFT is not breaking the markets" but it is not about the markets, it is about the people, businesses, investors. Are you talking about the market as something virtual, electrical or mechanical, non-existent in human world? The markets do not exist without investors and businesses. America is a society of business owners. Asset ownership is evident through independent businesses, direct market participation or over 401K holdings. Why should investors who own stakes in US businesses tolerate some short term price takers who do not have vested interest in such businesses but are the ultimate source of risk not associated with true riskiness of owned asset? Why do businesses with their assets and human capital should tolerate such culture if they do not operate at the speed of a high frequency trading? Why don't all these high frequency traders take their beloved models and algorythms and find ways to please their wives or educate their kids? This is an absolute nonsense to assume that all the benefits of HFT culture for the market are good for society.

 

HFT is a mixed blessing at best. Just wait until the market drops and all the HFT/Algo firms pull their liquidity from the market within moments of each other, it will shock you how extremely this accentuates a plunge.

But Rhaegar fought valiantly, Rhaegar fought nobly, Rhaegar fought bravely. And Rhaegar died.
 

Thanks' for comments. I should admit that my points may not match your intentions but I see that the many occasions of broad-based discussion regarding HFT in media and internet are not addressing the fundamental economical and social issues. Trading cost, liquidity which by the way is not quality liquidity in the sense it should have dedicated flow into businesses rather than stock price short term fluctuations, might be beneficial for velocity and efficiency of the market but it undermines the fundamental principles of asset ownership. Massive complaints from investors regarding HFT is the best evidence that something is wrong with it's application in the marketplace. I am mostly concerned about the rise of automated computer trading of which HFT is a particular case but empirically the most evident one in the markets. I believe that HFT should be allowed for market making business and as execution platform for customer large order execution but as an independent and automated trading solution should be banned. SEC should be monitoring institutions that could utilize HFT for trading their own accounts and ban any quantitative institutional trading with HFT platform that is not in market making business. The problem of HFT discussion that instaed of discussing HFT benefits and drawbacks people should realize what kind of market they want to see, on what economical principles it should operate, and for what fundamental reasons we should have such markets.
What is SEC's fundamental assessment of US equity market? How does SEC want the US equity market operate? The problem of our regulators that they seem to agree with HFT supporters and some academic advocates that further understanding of HFT is necessary before any action undertaken. It seems that such supporters and academia are mostly focused on price related quantitative research regarding HFT and it's impact on capital markets. While the arguments about low cost and liquidity could have some legitimate grounds, additional research and understanding are not needed because this is about philosophy of how you want to see the market operate, this is about market system design which has little to do with market pricing. This is about the internal market culture. If you want to have democratic market where investors buy and sell assets based on their human assessment of businesses then HFT traders should be banned or at least their activities should not have such profound effect on businesses and investors. However, if you look at the market as an electronic gaming industry with high volumes from trading algorithms and low cost for fast trading but no longer representative of broad public participation, then you sabotage numerous complains from investment public who carry no vested interested in very questionable and socially harmful trading technology but at the same broadly exposed to the market through 401K accounts. This is not about HFT market benefits, this is about what social and economic value they provide for businesses and investors, economies and people.

 
Best Response
mCobb:

well because people are so butt hurt about there not being an equal playing field in terms to execution of trades

People are butt hurt about everything these days. The only way to please everyone would be to nuke the planet back to the Stone Age and start as amoeba again.

Finance was/is never about an equal playing field. I find it funny that people vilify HFT, but forget all of the other ways high finance sticks a rod up the retail client's ass. You really think you're getting good fills as a retail guy on any sort of fixed income order that isn't a Treasury? That when Uber IPOs, GS/MS won't try their hardest to skim even higher commissions? What about all of those investment reports from the ER department - like there isn't an incentive to recommend certain equities (especially when said companies are trying to raise)?

HFT is a function of the markets which provides pros and cons to the participants involved. Market makers aren't a charity, they're not here to provide liquidity for a "thank you", and they're also not the Salvation Army. If there's money to be had, they're taking it. People love to throw out "front running" and latency arb like it's some collusion while forgetting that these same firms are the ones providing the two-sided markets and instant execution when things hit the fan. Also, front running is only bad for retail traders trying to scalp the markets as well (but lack the tech/infrastructure to do so). If they actually planned to buy and hold the securities as they claim, a few cents on the fill will be insignificant in the overall run. They're just upset they're not getting to make the easy cash.

What cracks me the most is that people think life in general is supposed to be an "equal playing field". Except for select relationships with your family/close friends, the two parties are constantly trying to get an edge on the other. When you get a new car, the dealership is trying to edge you out. Same when you go buy a new phone (waaaah Apple frontrunned me cause they knew I wanted the new iPhone and they raised the price). Getting a damn burger at Shake Shack is them charging you a premium for the service they provide. HFT is no different. If anyone is so upset about it, they can extend the duration of their investment horizon to the point where any latency superiority from the algo firms is nullified. The finance people (the public's opinion of electronic trading is beyond irrelevant) crying foul about HFT aren't upset that there's low hanging fruit ripe for the taking in the markets. They're upset that for the first time in a while, they aren't the ones playing others, they're the ones getting played.

 

"Not being an equal playing field" ?

Well you're a free man. If you wanna pay for recruiting quants; pay an exchange collocation, DMA connection and market data fees, spent months to develop market microstructure models... you can. If you think all these costs would increase your trading costs more than your profit, and you rather choose to pay fees to an algo broker to execute your orders (and letting him develop all the stuff)... you can. Then, if you think the algo broker fees are too expensive and would be less profitable than just going directly through a standard exchange... you can. In fact, you're free to choose the technology you want, the execution strategy you want, and the providers you want.

But what you cannot do is sitting on the ground, looking people estimating their costs and adopting execution strategies and telling to the Lord: "what an unfair world ! "

:)

 

I'm not against HFT but all those small percentages of a penny can add up in the long run. I think that's were people have the biggest problem with it but if you're willing to skim only a percentage of a penny multiple times through algos then go ahead, power to you.

 

Yeah, have you not seen Office Space.

Peter Gibbons: Ah no, you don't understand. It's very complicated. It's uh it's aggregate, so I'm talking about fractions of a penny here. And over time they add up to a lot. Peter Gibbons: Um... the 7-11. You take a penny from the tray, right? Joanna: From the cripple children? Peter Gibbons: No that's the jar. I'm talking about the tray. You know the pennies that are for everybody? Joanna: Oh for everybody. Okay. Peter Gibbons: Well those are whole pennies, right? I'm just talking about fractions of a penny here. But we do it from a much bigger tray and we do it a couple a million times.

 

Skimming or front running is just one of the strategies (and probably the only debatably nefarious one) an established HFT can deploy. Other strategies include market making, taking liquidity rebates from exchanges, various forms of arbitrage (intermarket, statistical, latency, etc.), and regular directional trading based on announcements/news/momentum.

If you notice, these strategies are all strategies used by retail and institutional investors. The only thing that sets HFT apart is the speed and frequency at which it is done (and therefore allows them to capture many "easy" opportunities). And to everyone crying about front running, that is basically what trading is. Buying something you believe another counterparty will want in the future, so that you can sell it to them at a higher price. If Warren Buffett sees that MSFT is undervalued, and that based on their financials are worth a higher price, he will buy it, and wait for the price to rise before selling it. He knows eventually someone will want to pay more for MSFT because it has strong fundamentals. Anyone who debates this, please explain to me how this is different from Jump Trading knowing a whale wants to buy MSFT, and buying it first before selling it at a higher price? Show me what changes apart from the timescale.

Average investors are barely, if at all, affected by HFT. If anything, it allows them to get faster and better execution. As I said in my post above, any fear of front running can be nullified by lengthening the horizon of your investment, and/or using a limit order. If you think a HFT will front run you, and drive the price up, set a maximum order on the price in which you're willing to pay, and you won't have your fill. If you're using market orders or getting filled at an adverse price that is still below your limit, that is your fault and you (should) know what you're getting into.

HFT will only hurt you if you try to scalp spreads/front run other people (which is a terrible idea anyways). And even if you managed to continually get your bid hit/offer lifted, the pennies you'd lose to the algos would be a fractional percentage of the commissions you'd be paying TD Ameritrade or whatever broker you're using. Go complain to them.

 

It increases market efficiency, so I don't see a problem. However, if you're approaching this from an angle of "fairness", nothing is really fair and nothing is promised to be fair in the markets. Institutional investors will almost always beat individual investors when it comes to information asymmetry and data analysis. So why are HFTs singled out, when banks are just as "unfair"?

Solid question tho.

 

I think a part of the general investment community's ire with HFT stems from the notion that it's more a mechanical exercise than an intellectual one - and before hft coders and advocates get butthurt and flame me for saying hft isn't "intellectual," hear me out - a lot of people in the money management industry attach a romantic sentiment to the idea of battling intellects, that "I'm going to do the very best research and come up with the most elegant arguments and prove that I'm the smartest by building a track-record of my investment theses being correctly proven out." Why then are money managers happy to tolerate traditional bank broker dealers when their function may or may not be just as "mechanical?" Maybe because traditional dealers been around since the dawn of time and "that's just the way it is." HFT is a new entrant, playing a different game, and many people aren't going to like that. Whatever your moral argument is, the current response is natural given human nature, and I think to see it play out any other way would really be an exception.

 
Lizard Brain:

I think a part of the general investment community's ire with HFT stems from the notion that it's more a mechanical exercise than an intellectual one - and before hft coders and advocates get butthurt and flame me for saying hft isn't "intellectual," hear me out - a lot of people in the money management industry attach a romantic sentiment to the idea of battling intellects, that "I'm going to do the very best research and come up with the most elegant arguments and prove that I'm the smartest by building a track-record of my investment theses being correctly proven out." Why then are money managers happy to tolerate traditional bank broker dealers when their function may or may not be just as "mechanical?" Maybe because traditional dealers been around since the dawn of time and "that's just the way it is." HFT is a new entrant, playing a different game, and many people aren't going to like that. Whatever your moral argument is, the current response is natural given human nature, and I think to see it play out any other way would really be an exception.

This isn't an argument; their feelings are irrelevant.

 

Technological advancements have tradeoffs. I spent a good amount of time on the desk constantly battling the machines... A quick rant of my frustrations:

Let's say you need to buy 5,000 options of XYZ stock, and you see 5,000 offered on the screens @ 2.82. If you send an order direct to market to buy 5k with a 2.82 top, it first hits the BATS exchange (which is owned by HFT) where let's say there are 650 offered there. Your order then continues through the fiber onto NYSE/AMEX/PCoast/Boston/CBOE/ISE/ect which are farther away through the wire. Because HFT firms are co-located, they can pull their offers to 2.84, faster than you can buy what's offered on the other exchanges. In the beginning, you were constantly left scratching your head why you only got done on 1k when there was ample liquidity on the screens when you swept 'em. You cancel the balance and start playing with the machines. Depending on how you displayed your order size, the algos will try to figure you out.

For example, if you re-send the order for the balance "iceberg", which only displays a tiny portion of your order size, the machine will bang a small odd-lot off you and see if the bid is still there. Then it knows you are iceberging. Similarly, the machine can walk its offer down. If they are immediately swept away, it knows you are "prowling" (btw these are GSEC's algo names, each banks' algorythmic trading division has their own branded/white-labeled names that perform the same/similar functions).

Long story short institutional traders end up paying up a little or receiving a little less...

From the exchange's point of view, they love HFT. It has led to explosive volume and the maker/taker model where they pay you rebates if you are a liquidity "provider", meaning you were the NBBO and someone hit/lifted you and you got a little rebate. The amounts seem small but in aggregate can become somewhat substantial. Regardless it created some degree of behavior change on both sides as the exchange world became more competitive.

With the rise of HFT, spreads became much tighter, which was beneficial to retail investors (what was once a 2.75/2.85 market 10,000 up or wider is now 2.82/2.83 250 up. This is the argument that HFT firms love to lean on, but in reality I don't think it does much for the retail guy who can't tell his ass from his elbow and makes the institutional investor pay more.

Over time, the demise of the "spread so wide you can drive a truck through it" to these penny-wide spreads that also inhibited the crossing of larger trades on the floor (your floor broker has to sweep the screens to put up (cross) a large order so that it complies with NBBO regs) that it simply took a lot more effort to trade larger blocks. With the locals (floor brokers) and mkt making firms winding down, the machines became a much larger part of the liquidity out there. This matters because HFT firms don't really care what they are trading so long as markets are "risk on" - meaning things are calm. A lot of work in recent years has been on the efficiency side of HFT with respect to turning OFF the machines when mkts become turbulent (news releases, headlines, etc). This "short vol" bias does the EXACT OPPOSITE of what HFT firms claim they do when mkts become volatile, which is provide liquidity. Case in point, the "flash crash", where the machines shut down.

These are some of my frustrations with this world. If you would like to learn more about how they manipulate the facts or even pick people off in some markets, have a look at Virtu's S-1 filing:

http://www.sec.gov/Archives/edgar/data/1592386/000104746915001003/a2219…

TLDR; HFT has pros and cons. You can't say HFT provides more "liquidity" without specifically defining what that means, because in times of crisis HFT often don't. The complexity of "market microstructure" has led exchanges to partner with HFT firms on maker/taker models and co-location services that may not serve institutional investors' best interests.

"Bulls take the stairs, bears take the elevator" "Sell a teenie, lose your weenie"
 

Other than reading flash boys I really don't know anything. So I am not really qualified to comment intelligibly. From a broad perspective I would say it's fine. However, if the order types described are quasi market manipulation and or front running. There is a fine line in paying higher fees to receive better service and having an unfair advantage.

 

Actually there are a number of legitimate reasons to dislike HFT. One of them is spoofing. A number of HFT firms have been caught spoofing and have been fined; there's hardly any defense against it.

A second case are flash crashes, for example the 2010 one. Now, for the guy who above mentioned that HFT are ''tremendous benefit for markets'' please enlighten me how exactly flash crashes do. Otherwise you are just repeating someone's propaganda.

Third and more generic case are various manipulations. Once again, various HFT firms have been caught and fined for that reason but it can be easily argued that the fines with tiny compared to the profits made via manipulation.

For example: http://www.bloomberg.com/news/articles/2014-10-16/athena-to-pay-1-milli… https://www.fbi.gov/contact-us/field-offices/chicago/press-releases/hig… http://www.reuters.com/article/us-optiver-settlement-idUSBRE83J01220120…

Of course that doesn't make high frequency trading all bad per se, but it sure paints a dark image on the industry.

Never discuss with idiots, first they drag you at their level, then they beat you with experience.
 

I was one of the people defending HFT above, but I also agree that there are valid negatives to it. No system only brings pros to the table, every one of them has some cons.

All of the points you listed are valid. Here are some of the things I'd like to bring up though:

1) Spoofing: it's incredible difficult to define what spoofing is. The SEC definition is an order which is entered with the intention of being cancelled. There is no concrete way to determine whether or not an order is a spoof order. Unless you can document the trader or algorithm's intention when they entered it, the only thing we are basing this off of is time. So what is the duration that would make an order a legitimate one versus one that is a manipulation order? Without a proper method of determining this, it's easy to just use this reason to target HFTs. Of course their cancels would come in faster, they have the firepower to do so. If I manually place a click trade in a highly illiquid market and wait 1 minute before cancelling it, then sell after the uptick, is that a spoof? People forget that even when floor trading was a thing, there were all sorts of games being played between participants. Guys were duping each other long before algos took over.

2) The Flash Crash is an isolated event. To use something that has happened HFT as the scapegoat for market manipulation. It's foolish to think that it only happens from the algo firms. I'd venture a guess as to say that there are just as many dubious things that go on in more illiquid fixed income markets where a small number of dealers are shrouded in secrecy.

Many of the things that people slam HFTs for aren't particularly unique to the industry. A lot of it is due to the fact that computerized trading is so quiet and secretive that people jump to all sorts of conclusions. I highly doubt people would care half as much about the effects of HFT if they were given completely transparent access to what goes on in institutional sell-side trading desks.

 

I think it's important to realize the role HFT and algorithmic trading plays in the market making context. Many of the private firms listed make markets in some areas, and every BB uses algos to make markets, it would probably be impossible to fairly price equities today without it.

 

My understanding of HFT:

Bus Company A has access to high-tech GPS system. Bus Company B does not. The GPS system is highly sophisticated. It provides the fastest possible routes in real time. As a result of the GPS system, Company A has shorter rides and therefore gains a competitive advantage. Company B can no longer compete.

Company B, in response, rather than upgrading to the GPS system, complains to the government that Company A has an unfair advantage. Company B pundits write books about how Company A is destroying the role of the classical bus driver. Company B argues that Company A is anti-competitive.

Government bans the use of the GPS system. All bus rides are longer. Company B is once again able to compete.

“Elections are a futures market for stolen property”
 

Saw an interesting article on Reuters the other day about Flash Boys. Two UC-Berkeley professors did a big study on HFT and are pretty much saying that the theory of HFT "front-running" retail investors isn't quite true. For anyone interested, it's titled "Berkeley Study Finds Scarce Evidence of Market 'Front-Running.'"

 

As I stated in the other thread, it's easy to blame one's failures in trading to "big brother" HFT or the Fed.

From what I learned through the years, technicals are great for longer term trading. For short term trading, they are not that good. Most of the successful daytraders do not rely on charts too eavily. Instead, they use filters that, according to their preference, will show you which stocks are popping up more than a specific % and then they jump on them for a few minutes or less. They call it "momentum trading" or sometimes "event driven" trading if the move is caused by some news.

Can HFT be a problem when trading like that ? I don't think as bad as some conspire. Some people make it sound like you are in direct competition with the HFT machines. Not really, all you are doing is trying to get a chunk of the move. If the machine gets a bigger chunk because of its speed, well good for the machine, but as long as you got your small chunk, you should be ok.

Now if we are talking about employment competition, then yes, machines will wipe out a lot of jobs. But for personal trading, I don't see a big problem.

 
MrFuture:
As I stated in the other thread, it's easy to blame one's failures in trading to "big brother" HFT or the Fed.

From what I learned through the years, technicals are great for longer term trading. For short term trading, they are not that good. Most of the successful daytraders do not rely on charts too eavily. Instead, they use filters that, according to their preference, will show you which stocks are popping up more than a specific % and then they jump on them for a few minutes or less. They call it "momentum trading" or sometimes "event driven" trading if the move is caused by some news.

Really? I tend to believe the opposite (technicals win on the short, while fundamentals win on the long). And charts are pretty much a necessity for any daytrader. I trade index and commodity futures and I use technicals, a number of different time frames, and tape action to make my trades.

 
trade4size:
Seems like every other decent thread out there now ends up getting into a discussion about how HFT are gaming the markets so I wanted to get a dedicated thread to everything about them.

I will start....

Seeing a lot of threads about how HFT has taken over and there simply is no point in trying to beat the markets anymore that they are making both fundamental and technical analysis useless. I think thats basically a pile of shit.

Hypothetical Example of HFT's at work. Jobs report comes out way above estimates, prior months report gets revised up, unemployment rate drops, and we see a large uptick in financial and technology payrolls. A HFT is programmed how to react to this and has already put the trades on before a human can even load the page. S&P futures shoot up and begin to trade in a tight range.

WHAT THIS MEANS FOR TRADERS: tough to get in right away as the machine processes the information before human can react. However PMs and traders that run money will take the time to read the statement and make their adjustments. The quick trade off the news becomes impossible and the only way to really do it is be quick on the keys and use stops in the event you get whipped out.

HFTs have taken away the speed trading edge that short term traders used to have. That does not mean there will not be subsequent plays. Different players will react differently to information.

This should be enough to get us started.

You are SO far off the mark on what HFT is about, it's embarassing. It's not about "pressing the button fastest on NFP #s" - a lot of HFT is about gaming market microstructure and making "slow money" pay for liquidity/transparency. Most of the early pairs trading was about synthetically charging mutual funds and institutionals for liquidity in portfolio repositions. The fact that it's called "High Frequency Trading" and the fact that there are the soundbites on "thousands of trades a second" lures idiots like you into thinking it's about pressing the button fastest on NFP #s. There is some stuff going on with variance ratios and lead/lag indicators to find hidden relationships, there is some algorithmic/voice recognition stuff to detect microstructure patterns but no serious HFT shop is working on ramping up S&P futures to a tight range, one Friday a month. The reason serious market guys are concerned about HFT is because it will suck a lot of easy money out of the system. Since you have such a boner for Market Wizards, a good example would be a HFT shop implement Richard Dennis' trend-following system - except dynamically optimizing on time scale and doing it on 50 asset classes at a time.

 

ideating I used a simple example to get discussions started using a simple example that everyone could understand. Can you please drop the drama that is literally 3 years old now?

"Oh the ladies ever tell you that you look like a fucking optical illusion" - Frank Slaughtery 25th Hour.
 

This is off base. Very few HFT shops, if any, use the strategy described in the initial post. The vast majority of HFT are flow based. They look at the activity in dark pools and ECNs to assess the likelihood of a millisecond later shift in the NBBO. Also, they look at flow patterns to determine when huge blocks of stock are being executed and then jump in front.

While technical/fundamental traders can certainly use high frequency approaches to secure alpha, the discussion in the media and by the regulators is all about flow and microstructure.

 

Kevin my initial post is about something that would show a change in flow. In that situation there would be a rush to buy but before the orders were even able to adjust the HFT algos would have drove the price up. I am very aware of what has been happening to the microstructure and its all about front running institutional flow.

Im hoping that HFT gets oversubscribed and the algos start to kill each other but i think a more likely result will be some sort of government intervention/change of rules.

"Oh the ladies ever tell you that you look like a fucking optical illusion" - Frank Slaughtery 25th Hour.
 

ideating is right, the major fear around HFT is that it does take the easy money away, it takes scalpers and banks with creazy margins away. It hurts the guys who just move around shares at big companies and live off their client relationships.

The next level and where HFT's now are is understanding market psychology, so while you think the algorithm is trying to setup ways to process data as soon it comes out and put on trades. The computer is actually sitting and saying "ok this report is out, now trade4size will react in 4 ways"...With each one of those reactions the computer already knows your next 2 moves, they are writing formulas around figuring that out. A good HFT of the future won't be putting the trades off market noise, they will be putting off how you trade in certain situations.

 

Yes marcellus I do agree with that very much. It has turned into a chess match but I think thats because the rest of the industry has not came up with a way to fight HFT tactics because there has been so little information available as to how they operate. They have no adapted their strategies to counter the things HFT is doing. A year and a half ago I heard about HFT for the first time and there was no information on it. A few posts on zerohedge about it but that was it. I actually am a pretty good example of someone that got gamed by HFT. I got burned so many times on these momentum setups that turned into false breakouts I didnt take note of the big picture (HFT) and thats mostly because I didnt even know it was going on. So I am front and center someone that got gamed by it and am aware of it.

I think I am just now getting a better feel for how it works but the HFT's are as you put it adapting the key is keeping up with them as they adapt instead of being 2 steps behind. Now different things react differently.

Ideating since you bring up Richard Dennis I think its pretty amazing the way people are reacting about HFT now are acting the same way they did in regards to the trend following systems of the 70s and 80s. Back then the key to winning was to understanding the systems being used by the big players... today the key to winning is understanding how HFT algos work. In 20 years HFTs will probably be obsolete superceeded by another form of analysis.

Also HFT would not be "dynamically optimizing on time scale and doing it on 50 asset classes at a time." this is simply system trading and has been possible for atleast the past 10 years. Perhaps doing it on the micro timeframe but plenty of people of still designed systems to trade on shorter timeframes.

In reality its overhyped and just another buzzword people are using as an excuse for not making money on the market or claiming that everything outside of it is dead.

"Oh the ladies ever tell you that you look like a fucking optical illusion" - Frank Slaughtery 25th Hour.
 

So- you don't know what HFT is but you think "in reality it's just a buzzword that people use as an excuse to not make money."

Quality thread.

P.S. "I actually am a pretty good example of someone that got gamed by HFT. I got burned so many times on these momentum setups that turned into false breakouts I didnt take note of the big picture (HFT) and thats mostly because I didnt even know it was going on. So I am front and center someone that got gamed by it and am aware of it."

You realize this contradicts your conclusion, right?

Granted, I know tons of CS guys and went to school with engineers who took classes with professors who work with MIT/Stanford folks.

But there is tons of articles and books out there in the academic world about HFT, about writing the algorithms to understand how trading flow is created and what can be done with it.

Also I am pretty sure in 2005/2006, WIlmott forums started discussing HFT. WSO has just got to it, again shows how HFT was mainly an academic world for while.

 

Ideating for people daytrading it can be a hindrance and for me it definitely was. For people taking positions holding for days/weeks/months then I it probably just hurt their total returns by raising their average price on the buys and lowering it on the sells. For one set of people it can spell doom but have a little impact on others.

Stop micro-analyzing my posts like you are some trial lawyer trying to catch me contradicting myself. I am not perfect, I dont know that much I just wanted a friendly conversation on HFTs. Like it or not I got shitty grades from a non target and networked my way into exactly the job I wanted it just took me 2 years.

"Oh the ladies ever tell you that you look like a fucking optical illusion" - Frank Slaughtery 25th Hour.
 

i am far from an expert in HFT, but i think about it in this way: For years floor traders had a liscence to steal and used to rob people blind in stocks and futures. Now they have been cut out of the loop but in the rush to do so a bunch of quant geeks have stepped into a flawed system and taken their place. Eventually the marketplace will be leveled but it will take time and in the meantime some people will be able to get "between the wall and the wallpaper" like the floor traders used to. I still think there is plenty of room to make money and in fact I think the market is probably much "fairer" to the buyside trader then it was back when the floor traders ruled.

 

All the HFT market makers pulled their plug prior to the announcement. A 50k share order moved the SPYs almost 20c. If the HFT guys were in there, it would have been filled in 1c. The volatility after the announcement is more indicative of what the market is like when the HFTs are not at work.

 

Ok good point but why are they pulling the plug when they have the ability to make wider markets during times of uncertainty when they have the computing power to parse through the statements and react to the order book movements.

"Oh the ladies ever tell you that you look like a fucking optical illusion" - Frank Slaughtery 25th Hour.
 

They have no obligation to make a market.

If they think there might be an increase in volatility or a big directional move that could run them over they can simply pull their quotes. Also, I don't think we can assume that ALL automated market makers are operating at essentially zero latency. If the slower guys pull, I'm sure the faster guys will thin and widen their markets due to the relative illiquidity accompanying the departure of slower players.

 

So I understand the idea of HFT and how people make money off of it - they eat the spread. Fine, whatever.

But here's my question.

Outside of printing money, why does the government allow this sort of thing to exist? Adding liquidity to the market at the millisecond doesn't do anything to make the markets more efficient - in fact the opposite argument could easily be made (see the flash crash, due, probably, to automated trading). I just don't get it. If I was god almighty and lord of the universe I would make it a rule on exchanges that they only update once a minute or even five minutes and NO FLASHING, everyone gets the ticker data at the same time. No tasting either. Your firm wants to flood the market with 5 million buy orders it gets banned from the exchange for a month.

Anyone smarter than me want to tell me why I'm wrong? Honest question here.

 

I want to clarify my above comment. I see nothing wrong with making money on the loophole once it exists, sure. In fact, I am a huge proponent in exploiting loopholes to their maximum potential. I just don't know why exchanges or the general public allow people the loophole as it's more or less a license to print money.

I hope that makes sense.

 

Not sure if you are guys are getting this. Maybe ideating needs to drop the hammer again.

HFTs do not work on loopholes or fundamentals/market noise. HFT analyze trends and work in milliseconds.

You know why because finding correlations in milliseconds and when the market is tame is the most advantageous to someone playing the psychology game. HFTs are patient do the same thing day in and day out and make a slow growth of cash. They do not look to print millions in a day.

So as mentioned it makes no sense for a good HFT to do anything, because today there is Fed Announcement, today there is morons out there looking to make or lose millions. HFTs do not care to join those morons, they stay to their rules and are patient. This is why HFTs pulled their orders today, and when the flash crash happened.

Also making you wait for a minute for the market to update, is just going to make it easier for people to manipulate one way or other.

 

im only comparing bubbles and what happened afterwards. there are many quants, 5% or so of them will make it past the next 2 years.

many dotcomers(owners) adapted because they were enterpreneurial, quants are quants, HFT trading on your CV will spell death in the near future, mark my words. Thats why I think its a terrible area of finance for new graduates to pursue

 

Another HFT basher, just with a higher profile name.

Jim Simons gives a solid response to this, once in an CNBC interview, and more completely in response to a question at a presentation at MIT (in a video linked on this site somewhere).

Cuban's premise is about what happens when computers makes mistakes plays on this "terminator" ideology that machines can destroy the planet (so can humans). What I think is a meaningful comparison is the risk of a computer making a mistake compared to the impact of a human making a mistake. Given that it took the markets months to recover from human caused crashes, and 10 minutes to recover from an algo "caused" crash, I actually think he's making a solid case for HFT.

The people that would have lost money on this are not the institutional investors, the so called "good guys" in the market. There are lots of points he conveniently omits, mainly the markets were already down 3% or so on that day, so people were nervous. JS covers it better than I ever could.

 

Edit: Sorry, I went on a total bender from too much coffee this morning when I first responded. I don't think the two cancel each other out because the problems are different. I doubt that simply banning or curbing hft's is going to make liquidity dry up in a meaningful enough way that speculators can just blast prices all around. And vice versa. Then again, I don't particularly worry about HFT so I'm not very impartial which was your original point.

Here's a fun thought. Why not say that if you are deemed to be 'speculating' you can't lever the trade. Shit. That's why I would do it anyway? Who could resist levering up 10-1 and making speculative bets. Besides, 90% of it isn't even your money! Hell, probably 100% of it isn't even your money. I'm of the opinion that speculation is not an inherently bad thing and is a part of any market. I mean christ, even a farmer who is hedging his crops is 'speculating' that the prices he might have to sell at in the future are less than it is now. How is his need any different than a meteorologist who studies weather thinks that crops are actually going to be less bountiful and takes the other view? I mean let's be honest, what are we actually trying to do here? Protect the retail investors? Or are we panicking because there are about 10 million other factors that are more meaningful towards the inflated prices of commodities that have nothing to do with speculators.

In regards to HFT and Algos, I think the question is how are you banning HFT? What precisely are you going to ban? The amount of time between trades? Banning the practice of quote stuffing? Or ban the use of algorithmic trading entirely, which is actually a somewhat separate issue. I'm not trying to argue an issue of liquidity providing being the ultimate high minded goals of hft guys. There are plenty of algorithms combined with super fast connections speeds that just crush the markets.

To my mind, if I have to HFT guys competing to get there the fastest to buy my stock i'm selling and then flip it to the next guy I could give a shit. So let them trade as fast as they want. If I can get there in .01 seconds and the other guy gets there at .015 seconds it doesn't really matter. They buy it and then flip it making the spread. If they start quote stuffing and putting up phantom bids at thousands per second to drive the prices upward or 'scare' me away that is a different issue. But frankly, I've still yet to figure out what retail investors are being hurt by any of this? Maybe none of them are around anymore to care but I don't think that speed is necessarily as big an issue as an algorithm that relentlessly pounds a stock to jump on the momentum ride up or down; or hell even create that momentum itself. Besides, an algorithm could still achieve this even if you slow the speed down quite a bit. Thus I'd argue that it is more dangerous to have a bunch of guys sitting around dreaming up algorithms that go in and do all kinds of crazy things for any number of reasons.

 

I'm somewhat surprised you read Ars Technica.

Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.

Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.
 

it's mostly nerdy stuff. that's why I read it..

and i don't think it's popular with bankers, but maybe i'm wrong.

Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.

Only two things are infinite, the universe and human stupidity, and I'm not sure about the former.
 

Doesn't impact long term investors really. It is actually helpful as far as liquidity goes in the marketplace, but that is juxtaposed against an increase in volatility in the marketplace. I don't think it negatively impacts the markets necessarily, however, I think in instances yes it can be a problem and push the market one way or another. The last thing needed right now are more regulations in the marketplace. A financial tax is messy and impractical. If nothing else it is more political theater than any real, practical solution to a problem I'm not even sure exists.

 
IlliniProgrammer:
For small scale individual investors, HFT means nothing.

The problem is that if you invest in an actively managed fund that buys not in single lot increments but trades thousands of lots every time it wants to buy or sell, that's where HFT hurts you.

I manage my own stock portfolio and I don't have stop orders so I don't worry about HFT.

No stop orders? I'm curious, are you constantly monitoring your portfolio? Or do you have a relatively conservative investments?

 
IlliniProgrammer:
I am a long-term investor. My portfolio is essentially 95% bricks and mortar dividend stocks that pay at least 3%.

If you are a long-term investor that manages his own portfolio, HFT has no advantage over you.

Indeed HFT probably helps individual small-time investors fractionally; spreads are narrower (better execution on market orders), and there is far more liquidity in small lots.

 

^Don't mind them, their just being a couple of dicks.

When a plumber from Hoboken tells you he has a good feeling about a reverse iron condor spread on the Japanese Yen, you really have no choice. If you don’t do it to him, somebody else surely will. -Eddie B.
 
hedgefundfreak:
I personally think that this is just a phase the market is going through and that people will soon realize that it is totally unreliable and can produce devastating consequences. Do you think that the SEC should put a ban on high frequency trading activities?

What are you basing this opinion on? Calling a broad strategy "totally unreliable" leads me to believe you know very little about high frequency trading. Well, that and all your previous posts.

 

Also I wonder how dedicated will Eric Schneiderman be when it comes to putting HFT to the sword? we shall see.

I am surprised to see that Virtu is indefinitely postponing their IPO.

i'm not smart enough to do everything, but dumb enough to try anything
 

The only way HFT really affects 'traditional' investors is by creating a more volatile market which ends up trading on a bunch of technicals rather than worrying about fundamental strength of companies. That really is what lies at the heart of the market nowadays is a shift from fundamental qualities of companies to technical strength of markets and how you can use technicals to trade. What you run into is that the market becomes more and more unstable because if there is a large move it is exacerbated by the prevalence of machines in the market. You see it normally in large downward moves because machines, HFT guys specifically, have no repsonsibility to provide liquidity into the marketplace and if things are getting ugly they can easily just flip off the switch. Then again, so can humans.

I think the practice of quote stuffing and things of that nature are the most worrisome to me. If I can fire off thousands of bids and asks you end up with the ability to drive the price whichever way you want and move the stock around in fractions of a second. It hurts price discovery at the same time the increased liquidity in the marketplace helps it. It really is a double edged sword. Exchanges love it because the fees they are getting from HFT's are driving their revenues and earnings even when trading volumes are falling off a cliff. At some point, we might already be there, exchanges are no more than tech companies providing faster and faster access to their clients.

 

Iure voluptas soluta voluptatibus quae voluptatum doloremque. Excepturi voluptatem architecto impedit vero officiis dolores dolorum. Voluptates maxime et temporibus tempora aliquam quis.

Voluptatem est consectetur voluptatem id voluptates et. Quis omnis odit quis non eius. Error iure beatae dolorem molestiae dolore quia vero. Odio asperiores quo aut ab aut consequatur saepe quis. Neque esse earum eum unde.

Molestiae dolores cupiditate id qui qui eaque et nesciunt. Saepe sed consequatur velit repellat harum voluptatem. Hic numquam modi totam explicabo sed. Autem voluptas corporis qui.

Non consequuntur et harum aut dolorem qui culpa. Qui velit ducimus sed ut. Nihil quia consequatur quibusdam architecto et.

 

Tempora tempore porro corporis soluta. Commodi iste et omnis vel enim omnis. Dolores blanditiis porro nostrum ratione molestias fugit hic.

Aperiam qui eum fugit. Animi non odio reiciendis deleniti atque aut. Excepturi nostrum libero aliquid est consequuntur qui eligendi.

Vel inventore modi consequatur perferendis. Tempora sint in sed reiciendis similique itaque. Dolor velit officia vero nemo omnis ex. At at velit provident perferendis magnam voluptates.

Qui et et earum nihil quis. Deserunt nostrum perspiciatis accusantium at voluptatem aut. Sint sed nihil soluta quaerat a placeat. Saepe saepe nostrum ea eos qui tempora. Corrupti minima placeat animi qui quaerat nobis. Omnis assumenda reiciendis aliquid et omnis.

 

Quidem a dolorem commodi reprehenderit. Consequatur quod illo ut enim omnis perspiciatis occaecati. Mollitia impedit ut vitae sit fugit ut. Ipsa nostrum nesciunt dolor alias. Totam optio doloremque dicta quis qui.

Aliquid omnis rem sed voluptate eligendi eum. Aut officia aliquam pariatur facere nisi. Doloremque et assumenda rerum minima veritatis quis.

Nesciunt beatae unde nihil sed dolorem quia iusto. Iure blanditiis occaecati cum voluptas. Exercitationem voluptatem similique expedita. Aliquid aut explicabo commodi et exercitationem. Nulla voluptatem deleniti hic aspernatur aperiam veniam eum. Veritatis rem non consequatur quo doloribus molestias. Sint quia non sit quisquam.

 

Nesciunt ut at et ex quae ex. Eaque illum officia exercitationem dolores. Voluptas totam soluta rerum hic. Veniam officia exercitationem commodi non dolor repellat. Quo eveniet nisi consequuntur delectus doloribus. Quaerat eos tenetur dolorem nisi ex consequuntur illum.

Nam consequuntur quasi iusto. Veritatis voluptas at ut necessitatibus sint.

Eos qui consequuntur quasi sit vel aut. Error explicabo expedita voluptas qui at sed tempora ut. Et magnam illum at reiciendis voluptas dolorem rerum.

Sapiente commodi voluptate similique est est velit. Nulla aspernatur fugit omnis deleniti. Quibusdam minus impedit architecto id at qui dolorem ut.

 

Voluptatum dolores eos voluptates voluptas magni. Aut natus qui at eaque. Consequuntur nostrum rem non tempora accusantium non et occaecati. Eos eos molestiae ullam iure. Maxime quis quasi totam omnis.

Rerum ab qui consequuntur et similique quia excepturi hic. Vero et rem dolorem earum. Illo in iusto vero aut eos culpa. Saepe illum consectetur sint. Nesciunt sunt pariatur libero officiis. Eius deserunt et sunt.

 

Optio reiciendis architecto praesentium. Sint similique est voluptatem aliquid dignissimos laudantium. Ipsa eius ea incidunt ipsam voluptas. Illo atque et perferendis temporibus accusantium velit occaecati.

Ullam aliquid asperiores magnam est ea. Eos qui quae dolorum qui qui dolores qui. Inventore eum nulla aspernatur delectus. Voluptatum fuga dicta similique.

Perferendis qui et eveniet enim praesentium qui expedita. Sint et expedita quibusdam architecto dolorem neque. Sit quidem omnis enim delectus adipisci occaecati.

Career Advancement Opportunities

March 2024 Investment Banking

  • Jefferies & Company 02 99.4%
  • Goldman Sachs 19 98.8%
  • Harris Williams & Co. (++) 98.3%
  • Lazard Freres 02 97.7%
  • JPMorgan Chase 03 97.1%

Overall Employee Satisfaction

March 2024 Investment Banking

  • Harris Williams & Co. 18 99.4%
  • JPMorgan Chase 10 98.8%
  • Lazard Freres 05 98.3%
  • Morgan Stanley 07 97.7%
  • William Blair 03 97.1%

Professional Growth Opportunities

March 2024 Investment Banking

  • Lazard Freres 01 99.4%
  • Jefferies & Company 02 98.8%
  • Goldman Sachs 17 98.3%
  • Moelis & Company 07 97.7%
  • JPMorgan Chase 05 97.1%

Total Avg Compensation

March 2024 Investment Banking

  • Director/MD (5) $648
  • Vice President (19) $385
  • Associates (85) $262
  • 3rd+ Year Analyst (13) $181
  • Intern/Summer Associate (33) $170
  • 2nd Year Analyst (65) $168
  • 1st Year Analyst (198) $159
  • Intern/Summer Analyst (143) $101
notes
16 IB Interviews Notes

“... there’s no excuse to not take advantage of the resources out there available to you. Best value for your $ are the...”

Leaderboard

1
redever's picture
redever
99.2
2
BankonBanking's picture
BankonBanking
99.0
3
Betsy Massar's picture
Betsy Massar
99.0
4
GameTheory's picture
GameTheory
98.9
5
kanon's picture
kanon
98.9
6
Secyh62's picture
Secyh62
98.9
7
CompBanker's picture
CompBanker
98.9
8
dosk17's picture
dosk17
98.9
9
numi's picture
numi
98.8
10
Jamoldo's picture
Jamoldo
98.8
success
From 10 rejections to 1 dream investment banking internship

“... I believe it was the single biggest reason why I ended up with an offer...”