New book on how traders are beating the HFT guys

[Alpha Trader]

This was an interesting interview, but it's a shame that he didn't go into more detail. Was it because he chooses for us not to understand it or does he lack the knowledge further from what he explains.

It would be nice if someone could elaborate or clarify some of the points below. Thanks in advance.

TCR: What made you think that high-frequency trading was behind those strange price movements you were seeing?

GARRETT: For one, the movement didn’t look like anything I’d seen in the past. It didn’t match human action. It was too fast, too consistent.

Anomalies would randomly pop up on my screen. A particular stock would drop 10% in one second, then run right back up a second later. I asked colleagues what these movements were and where they were coming from, but no one had an answer. Even the shortest-term charts, in which every data point represents one second and the data is extremely granular (or so I thought) didn’t yield any answers.

I've seen this happen occasionally; more frequently a few years back rather than now. I'm wondering how these systems can drop price that big and that fast when there are other orders that it has to chew through. Sometimes, this may include block orders. Additionally, I'm sure that there are other algos prepped for this kind of action and to take advantage of it. Would they not enter with their own trades and leave the primary algo high and dry?

TCR: So you’ve made a career of exploiting the dislocations caused by HFTs. What’s your answer to the question on everyone’s mind: does high-frequency trading affect the average retail investor?

GARRETT: Absolutely, although the impact varies based on what type of investor you are. For a shortterm trader, someone who makes many trades per month, the effects are huge.

I think the best way to understand HFT’s impact on you is to understand its advantages over you. There are three major ones.

One, HFTs have better access to the market. They have what we call direct access, which means they don’t have to go through a broker to execute their trades. When you place an order with, say, Scottrade, Scottrade will choose which exchange the order goes to, and they’re going to execute the order where it’s best for them. They’re going to buy it at the best price they can and then sell it to you.

Two, HFTs obviously have a major speed advantage over other investors. They glean this advantage in many ways: by putting their servers right next to the exchanges’ servers, by using very sophisticated equipment, and also simply by virtue of programming a computer to act on pre-set instructions, which it can do much, much faster than a human ever could.

Third, the best HFTs have an impeccable understanding of the market micro-structure: what happens after you submit an order to your broker? Where does your order go, how is it executed, how are orders prioritized? HFTs are experts on this, but very few retail investors even understand the basics.

1) Who day trades through Scottrade?!

2) My orders are much smaller than theirs and for the most part I get filled pretty nicely on limit.

3) Why do I need to understand this if I am satisfied with my fills. If they need a better understanding to fill their large order and take the time to educate themselves in order to get those fills, so be it. It doesn't bother me in the slightest if they make a million, so long as I get my fill.

TCR: What are trending movements?

GARRETT: Trending movements are when an HFT deliberately moves a stock price up or down for its own benefit. For example, a computer can submit an overwhelming number of sell orders, knock the price of a stock down a few percentage points, then buy the stock back cheaply.

I'm assuming most seasoned traders know that this happens and probably take advantage of it. Also, if there are other algos programmed for this, they can get in before the primary algo to take advantage of the anomaly and in the process leaving the primary algo wanting.

GARRETT: It is, which is why investors need to understand how to protect themselves. One of the most important tips I can give you is to never enter stop-losses into the market. There are algos designed to sniff out stop-losses and manipulate them against you.

I’ve seen this many times: prices drop 2-4%, clear out stop-losses, then run up for substantial profits. So the poor retail investor gets his stop-loss tripped and sells on the cheap to an HFT, whereas the HFT buys cheap and profits once the price ramps back up.

Now this is of real interest to me for various reasons. Most times, I do not put in hard stops. However, when I am dealing with more positions that I can handle, I will place hard stops in the ones that are behaving well.

In saying this, sometimes I swear by God that they could see my stop and literally came to pick it up. This would be the only HFT activity that would upset me. It's hard work finding that sweet entry, only to be booted out before the party bus leaves.

TCR: So HFTs try to predict what you’re going to do before you do it. Do the brokers admit to selling this information? Can traders opt out?

GARRETT: This data is standard and available to anyone who wants to buy it, so it’s not that HFTs are purchasing illegal information. But the data set is huge and is only of practical use to players with very fast and powerful computers – meaning HFTs. And yes, most brokers I have encountered will allow you to opt out of having your unique number attached to your information.

To be clear, I’m not saying HFTs track your individual account and literally jump in front of you right before you trade. But they do use this information on the aggregate to model traders’ behavior. So an HFT could have a very good idea of when traders on, say, E*TRADE’s book will enter into a certain transaction.

Individual accounts are too minuscule for HFT firms to play with. Is this really a fact or is this guy making this up? Also, I am pretty sure bigger firms protect their strategies so as to prevent other HFT firms from benefiting from it. So what's the deal here?

Above were the main points of interest, though a lot more can be questioned from the interview. I still remain in the camp that HFT is not something that effects my trading. Either I have learnt to adapt or it really does not interrupt with my stlye of trading.

 

[Grandluxe]

This was an interesting interview, but it's a shame that he didn't go into more detail. Was it because he chooses for us not to understand it or does he lack the knowledge further from what he explains.

It would be nice if someone could elaborate or clarify some of the points below. Thanks in advance.



I've seen this happen occasionally; more frequently a few years back rather than now. I'm wondering how these systems can drop price that big and that fast when there are other orders that it has to chew through. Sometimes, this may include block orders. Additionally, I'm sure that there are other algos prepped for this kind of action and to take advantage of it. Would they not enter with their own trades and leave the primary algo high and dry?



1) Who day trades through Scottrade?!

2) My orders are much smaller than theirs and for the most part I get filled pretty nicely on limit.

3) Why do I need to understand this if I am satisfied with my fills. If they need a better understanding to fill their large order and take the time to educate themselves in order to get those fills, so be it. It doesn't bother me in the slightest if they make a million, so long as I get my fill.



I'm assuming most seasoned traders know that this happens and probably take advantage of it. Also, if there are other algos programmed for this, they can get in before the primary algo to take advantage of the anomaly and in the process leaving the primary algo wanting.



Now this is of real interest to me for various reasons. Most times, I do not put in hard stops. However, when I am dealing with more positions that I can handle, I will place hard stops in the ones that are behaving well.

In saying this, sometimes I swear by God that they could see my stop and literally came to pick it up. This would be the only HFT activity that would upset me. It's hard work finding that sweet entry, only to be booted out before the party bus leaves.



Individual accounts are too minuscule for HFT firms to play with. Is this really a fact or is this guy making this up? Also, I am pretty sure bigger firms protect their strategies so as to prevent other HFT firms from benefiting from it. So what's the deal here?

Above were the main points of interest, though a lot more can be questioned from the interview. I still remain in the camp that HFT is not something that effects my trading. Either I have learnt to adapt or it really does not interrupt with my stlye of trading.

When the algo knocks down the price, their other algo swings in immediately in ms and scoops it up immediately before you can even type up the stock. It just registers on your filter.

 

[drownpruf]

Individual accounts are too minuscule for HFT firms to play with. Is this really a fact or is this guy making this up? Also, I am pretty sure bigger firms protect their strategies so as to prevent other HFT firms from benefiting from it. So what's the deal here?

Ridiculous. Crowd behavior aggregated from a bunch of retail accounts? IMO, one the first institutional HFT algo was from SUSQ. An "interest" algo that stepped in front of ECN limit orders and actually improves NBBO. Most of the stuff is arbitrary. We're not talking AI here. These are relatively basic principles that can operate in the ms-level. This guy makes it appear as "transcendence" and it's not even close.

 

[Alpha Trader]

When the algo knocks down the price, their other algo swings in immediately in ms and scoops it up immediately before you can even type up the stock. It just registers on your filter.

Interesting Grandluxe. However, if it is algo vs algo, is there a chance that another algo could outdo them? If so, they would be in a bad place.

 

[Alpha Trader]

Ridiculous. Crowd behavior aggregated from a bunch of retail accounts? IMO, one the first institutional HFT algo was from SUSQ. An "interest" algo that stepped in front of ECN limit orders and actually improves NBBO. Most of the stuff is arbitrary. We're not talking AI here. These are relatively basic principles that can operate in the ms-level. This guy makes it appear as "transcendence" and it's not even close.

Drownpruf, thanks for the response. It seems like you may have better knowledge about HFT. If time permits, would you be kind enough to go through the questions I had posed (first post on this page). If not, I understand.

 

[drownpruf]

Drownpruf, thanks for the response. It seems like you may have better knowledge about HFT. If time permits, would you be kind enough to go through the questions I had posed (first post on this page). If not, I understand.

My experience with HFT is limited to D1 hedging in single names, but large caps (AAPL, GOOG, ETF). I can't offer much into the mechanics beyond what I experienced some 10Y ago, but from what I have seen (at large(ish) funds) it's no longer trending toward efficiency on the micro-level. The early algos were very beatable. You spoof on the buy, small... SUSQ et al would come in to penny you and you'd hit the bid on your intended short. No moral issue here, your intent is to get filled, not an overt-edge related trans.

Macro-wise there are two diametric forces at play. The HFT smoothes the daily vol (lower var) via sub-penny and the majority being MR strategies, but makes the microstructure chaotic. It can be easier to fill in the vol-markets then it is in shares. It's very difficult to fill non-marketable share orders. I model all D1 fills as marketable. It's not a viable method if it cannot survive lifting offers.

Low vol serves nobody but the HFT. They shut-down when overrun and reduce vol for all participants over the long-term. Not even systematic vol-sellers are served by HFT at the nominal line on vol is too low to short with impunity.

 

[murray t turtle]

What are some of the tactics they are using then?

====================
Grande;
I like that short term trader in the new Jack Schwager[HFT Market Wizzards if I remember my newest trade book title]

He has a short bias, likes to pick tops in a long bull trend/up trend , like the , SLV + silver ''got hunted up several year ago] NOT the mr Bunker Hunt UP trend but the more recent silver trend to $50/+.LOL. I think he bought TYCO some also, but it was not clear to me, if that was a buy to open or or buy to close his position???????????????????????????????????????????????????????????????????????????

NOT that I want to trade counter trend much;
but I like what he is doing, his counter-trend helped trend followers get out at good fills. For example , he is buying to cover his SLV shorts, when some trend followers finally sell to exit longs[main SLV, silver UP[trend. Forgot his name but remembered his trading plan...................................................................................................................

 

[Alpha Trader]

My experience with HFT is limited to D1 hedging in single names, but large caps (AAPL, GOOG, ETF). I can't offer much into the mechanics beyond what I experienced some 10Y ago, but from what I have seen (at large(ish) funds) it's no longer trending toward efficiency on the micro-level. The early algos were very beatable. You spoof on the buy, small... SUSQ et al would come in to penny you and you'd hit the bid on your intended short. No moral issue here, your intent is to get filled, not an overt-edge related trans.

Macro-wise there are two diametric forces at play. The HFT smoothes the daily vol (lower var) via sub-penny and the majority being MR strategies, but makes the microstructure chaotic. It can be easier to fill in the vol-markets then it is in shares. It's very difficult to fill non-marketable share orders. I model all D1 fills as marketable. It's not a viable method if it cannot survive lifting offers.

Low vol serves nobody but the HFT. They shut-down when overrun and reduce vol for all participants over the long-term. Not even systematic vol-sellers are served by HFT at the nominal line on vol is too low to short with impunity.

Drownpruf, appreciate your input. It's a good think we can make a living..........for now.

 

[textrader4]

Just got done reading Traders Of The New Era. It is an excellent read! Being a full time trader I read a lot of trading books. At first mostly technical analysis books and then most recently trading psychology books. This book is a lot closer to psychology because you really get into the heads of real traders like ourselves. I felt like the traders being interviewed were everyday traders that I could totally relate to. When they talked about issues/concerns and what they do to avoid them I found myself saying to myself "yes! I have that issue too". The difference is they have been around long time and are highly successful so they give good advice on what to do.

Unlike other books Traders Of The New Era doesn't try to teach you specific setups or trade management. However, I felt like I gained more knowledge from it than any book I can remember. If you are a trader that is active in the equity markets I highly recommend it.

I haven't read the book yet (just ordered it today) so I can't comment yet on what the traders in the book were referring to. It is very difficult to backtest a setup that a discretionary trader is looking for. For a systematic trader you need to back test but I don't believe you do need it for a discretionary trader. Instead, a discretionary trader might start with trading 100 shares of a setup and as he feels that its working slowly increase share size. If it doesn't work he moves on. The testing is done in the market...nothing can be more realistic than that.

 

[xandman]

Get concrete with tables, charts and statistics. You may just produce a classic must-read for this trading era.