3 ms·

Pity, it is a bad book. Im still very much against HFT.

by igl 13y ago

aet 13y ago
There is no point to being "against HFT" -- you need to support an alternative market design and explain why it would be better.
- fsk 13y ago
  The best proposal I've heard: Instead of continuous trading, have an auction every 1-60 seconds. One trade per minute is plenty.
  - yxhuvud 13y ago
    What problem would that solve?
  - kasey_junk 13y ago
    A couple of issues with this: 1) most of the arguments for call auctions instead of continuous trading assume a centralized call auction that in and of itself adds no cost above our current decentralized exchanges. This seems like an extremely troublesome assumption given that we have seen the negative impact of centralized monopoly exchanges. 2) If it isn't a centralized auction there needs to be some price sync. mechanism. The most obvious one seems to be either latency arbitraging HFT (which for some reason people don't like) or a governmental central auction block the type of which is already causing legislative arbitration in the current markets. 3) Call auctions in and of them selves do not solve the major issue with our current prioritization system. If the auction algorithm has several prices that match the same amount of participants, or if the price it determines has an uneven number of buyers & sellers, who gets priority? 4) Price discovery is currently harder in call auction instruments. That could be because they are only done in illiquid markets and liquidity itself will fix that, but it is an issue with the current examples of this style of trading. 5) "Market" order mechanism. Many, many market participants are more worried about making sure they buy/sell than about high fidelity price fairness. If I say, "I want out of Goog no matter what", that complicates any auction methodology and introduces opportunities for gaming. At the end of the day, I can't figure out who will actually be protected in this style of market and to whose disadvantage. That is a lot of added complexity for a very vague benefit.
- ccurtsinger 13y ago
  Why? I see no problem with a position of "this is bad." And since this is a discussion about a book, it's worth pointing out that the book is focused around an alternative proposal for market design.
  - kasey_junk 13y ago
    Actually if you look at the documents of the new exchange in question, IEX, they don't propose to do anything about HFT at all. They are still a price-time priority matching exchange, they still allow colocation (well not directly, but they are themselves in data centers that have colocation), they still offer exotic order types, and their much hyped added latency amounts to 350 micro seconds. They aren't trying to remove HFT from the markets, they are trying to disadvantage one class of HFT in favor of another. I personally don't have any problem with their goal, but they aren't being honest about it and that seems problematic.
- igl 12y ago
  Alternative to banning HFT? I can't think of any. (To clarify by HFT i do NOT mean algorithmic trading‎ in general) The stock market is a service to provide equity for real production companies to grow. HFT distorts the real value of a stock. There is also 0 chance of any regulation authority to ever check the amount of paperwork HFT produces.
seanieb 13y ago
This is far from being a "bad book. Some of the points raised in this review are out of context. While some descriptions of the core HFT processes/issues in the book are naive or simple, they are revisited multiple times and more detail is added as the characters in the book learn more.
- tptacek 13y ago
 You should back those arguments up with evidence. Which points in the review are refuted in the book?
 - camelite 12y ago
 I'm in the middle of the book. Point 10, "Without co-location, the investors will fight for the spaces close to the exchange -- across the street, a block away, 2 blocks away, etc.", seems to my unsophisticated reading to be already directly addressed: 63.36%: "The NYSE had ... built this 4000,000-square-foot fortress in the middle of nowhere ... the moment they announced their plans HFT firms began to buy up land surrounding the fort ... the NYSE somehow persuaded the SEC to let them make a rule for themselves: Any banks or brokers or HFT firms that did not buy space inside the fort would be allowed to connect to the NYSE in or of two places: Newark, New Jersey, or Manhattan. The time required to move a signal ... undermined HFT strategies ... "There was a precedent: They'd let NYSE do it,"..." So to my unsophisticated eye, this sophisticated reviewer is blowing a bit of smoke. Not all smoke, but a bit of it at least.
 - camelite 12y ago
 Still reading. On #6, I never got that impression; Lewis refers repeatedly to the digitalisation of the exchanges bringing prices down. e.g.: 75.73%: "A few thought it was important to remember that technology had lowered their trading costs from what they had been decades earlier - and half-turned a half-blind eye to the stunts Wall Street intermediaries had pulled to prevent technology from lowering those costs even further."
 - camelite 12y ago
 13) "In August 2013, the Goldman automated trading system generated a bunch of crazy and embarrassing trades that lost Goldman hundreds of millions of dollars..." Yes, but this was on the Options market, on Options exchanges, and has nothing to do with the rest of the book. I'm beginning to suspect the author is flat out dishonest. The Goldman cock-up is used as a defense of Goldman. Lewis says that unlike other firms they had a legitimate excuse for not trading with IEX - they didn't trust their systems. So, no the cockup had nothing in particular to do with HFT, but Lewis never implied that it did, and it is impossible to read that paragraph and come to another conclusion honestly. The review author is not shooting straight.

Pity, it is a bad book. Im still very much against HFT.

by igl 13y ago

Pity, it is a bad book. Im still very much against HFT.

aet 13y ago

There is no point to being "against HFT" -- you need to support an alternative market design and explain why it would be better.

fsk 13y ago
The best proposal I've heard: Instead of continuous trading, have an auction every 1-60 seconds. One trade per minute is plenty.
- yxhuvud 13y ago
  What problem would that solve?
- kasey_junk 13y ago
  A couple of issues with this: 1) most of the arguments for call auctions instead of continuous trading assume a centralized call auction that in and of itself adds no cost above our current decentralized exchanges. This seems like an extremely troublesome assumption given that we have seen the negative impact of centralized monopoly exchanges. 2) If it isn't a centralized auction there needs to be some price sync. mechanism. The most obvious one seems to be either latency arbitraging HFT (which for some reason people don't like) or a governmental central auction block the type of which is already causing legislative arbitration in the current markets. 3) Call auctions in and of them selves do not solve the major issue with our current prioritization system. If the auction algorithm has several prices that match the same amount of participants, or if the price it determines has an uneven number of buyers & sellers, who gets priority? 4) Price discovery is currently harder in call auction instruments. That could be because they are only done in illiquid markets and liquidity itself will fix that, but it is an issue with the current examples of this style of trading. 5) "Market" order mechanism. Many, many market participants are more worried about making sure they buy/sell than about high fidelity price fairness. If I say, "I want out of Goog no matter what", that complicates any auction methodology and introduces opportunities for gaming. At the end of the day, I can't figure out who will actually be protected in this style of market and to whose disadvantage. That is a lot of added complexity for a very vague benefit.
ccurtsinger 13y ago
Why? I see no problem with a position of "this is bad." And since this is a discussion about a book, it's worth pointing out that the book is focused around an alternative proposal for market design.
- kasey_junk 13y ago
  Actually if you look at the documents of the new exchange in question, IEX, they don't propose to do anything about HFT at all. They are still a price-time priority matching exchange, they still allow colocation (well not directly, but they are themselves in data centers that have colocation), they still offer exotic order types, and their much hyped added latency amounts to 350 micro seconds. They aren't trying to remove HFT from the markets, they are trying to disadvantage one class of HFT in favor of another. I personally don't have any problem with their goal, but they aren't being honest about it and that seems problematic.
igl 12y ago
Alternative to banning HFT? I can't think of any. (To clarify by HFT i do NOT mean algorithmic trading‎ in general) The stock market is a service to provide equity for real production companies to grow. HFT distorts the real value of a stock. There is also 0 chance of any regulation authority to ever check the amount of paperwork HFT produces.

seanieb 13y ago

This is far from being a "bad book. Some of the points raised in this review are out of context. While some descriptions of the core HFT processes/issues in the book are naive or simple, they are revisited multiple times and more detail is added as the characters in the book learn more.

tptacek 13y ago
You should back those arguments up with evidence. Which points in the review are refuted in the book?
- camelite 12y ago
 I'm in the middle of the book. Point 10, "Without co-location, the investors will fight for the spaces close to the exchange -- across the street, a block away, 2 blocks away, etc.", seems to my unsophisticated reading to be already directly addressed: 63.36%: "The NYSE had ... built this 4000,000-square-foot fortress in the middle of nowhere ... the moment they announced their plans HFT firms began to buy up land surrounding the fort ... the NYSE somehow persuaded the SEC to let them make a rule for themselves: Any banks or brokers or HFT firms that did not buy space inside the fort would be allowed to connect to the NYSE in or of two places: Newark, New Jersey, or Manhattan. The time required to move a signal ... undermined HFT strategies ... "There was a precedent: They'd let NYSE do it,"..." So to my unsophisticated eye, this sophisticated reviewer is blowing a bit of smoke. Not all smoke, but a bit of it at least.
- camelite 12y ago
 Still reading. On #6, I never got that impression; Lewis refers repeatedly to the digitalisation of the exchanges bringing prices down. e.g.: 75.73%: "A few thought it was important to remember that technology had lowered their trading costs from what they had been decades earlier - and half-turned a half-blind eye to the stunts Wall Street intermediaries had pulled to prevent technology from lowering those costs even further."
- camelite 12y ago
 13) "In August 2013, the Goldman automated trading system generated a bunch of crazy and embarrassing trades that lost Goldman hundreds of millions of dollars..." Yes, but this was on the Options market, on Options exchanges, and has nothing to do with the rest of the book. I'm beginning to suspect the author is flat out dishonest. The Goldman cock-up is used as a defense of Goldman. Lewis says that unlike other firms they had a legitimate excuse for not trading with IEX - they didn't trust their systems. So, no the cockup had nothing in particular to do with HFT, but Lewis never implied that it did, and it is impossible to read that paragraph and come to another conclusion honestly. The review author is not shooting straight.