4 ms·

For the purposes of this discussion, it's not a difference at all. The question is not what should be legal or not, but what customers would prefer, and what m

For the purposes of this discussion, it's not a difference at all. The question is not what should be legal or not, but what customers would prefer, and what makes a better market. If you ask a customer "would you like an exchange that gives people enough information to front run your order" or "would you like a broker who front runs your order" the answer in both cases would be a resounding no. The claim in the article is that you can create a better exchange by providing an equal footing for all participants.

mrchicity 11y ago
Everybody wants to trade with no price impact. If you asked me if anyone else should be allowed to colocate their servers, I'd say no, just give me a monopoly please. One participant's desires don't make a good market. Fair and non-discriminatory access to exchanges (not free, but anybody can pay to compete if they wish), fast trading engines and deep order book information are what let traders put up such tight spreads for customers in the first place. Vigorous competition between all market participants keeps prices fair.
- JesperRavn 11y ago
  EDIT: I think you are right, although I would express things a bit differently. I'm writing a top level post with my thoughts. ORIGINAL POST: When you make a trade, there are some information rents associated with that trade. If someone knew you were about to trade, they could capture all of those rents. When there is a monopoly on exploiting these information rents (as in frontrunning), they will be captured by the frontrunner. When there is perfect competition (as when a number of HFTs can trade in another exchange before your trade reaches it), these rents will be all wasted on technology to maximize speed, i.e. no one gets them, they become deadweight loss. I understand that there are other issues, such as knowing whether a trade comes from a single seller or multiple smaller sellers. But this is a separate issue from simply trading ahead of other people. I don't understand why it's so controversial that an exchange would advertise the fact that traders can rely on their orders reaching the exchange before anyone else knows about it as a good thing.
  - mrchicity 11y ago
    The original trader still has the ability to capture his information rents in a single trade on one market. Nobody can see it and step away or react before the execution goes through. That's all anyone is guaranteed. HFTs face the same problem. If they're trying to cross the spread to put on a trade or hedge, they also have market impact costs and slippage. I think it's controversial because a lot of their marketing implies all HFTs do this as their primary business and thus add very little value. It creates this narrative of a parasitic activity that intermediates trades which would have already occurred where no risk is taken. I don't think that's the case in reality. All HFT strategies are looking at order books in the products they trade, and they definitely make money (or avoid losses) when someone executes inefficiently across exchanges, but there's more to it than that. HFTs are very successful in futures and non-US markets where products trade on a single exchange. How do they do that if they aren't good at predicting prices and taking risk?
    - JesperRavn 11y ago
      >The original trader still has the ability to capture his information rents in a single trade on one market... That's all anyone is guaranteed Yes this is the essence of it. Based on other posts, I would guess that if you really wanted to make simultaneous trades in multiple markets, then you could. But even if HFT really did render this impossible, it's not a bad thing, since it trading in two exchanges simultaneously actually gives more rents to the large trader than the theoretical ideal of a single continuous double auction. And if you want to trade in A then trade in B based on the results of A, then you certainly don't deserve to be the first to make it to B.

mrchicity 11y ago

Everybody wants to trade with no price impact. If you asked me if anyone else should be allowed to colocate their servers, I'd say no, just give me a monopoly please. One participant's desires don't make a good market. Fair and non-discriminatory access to exchanges (not free, but anybody can pay to compete if they wish), fast trading engines and deep order book information are what let traders put up such tight spreads for customers in the first place. Vigorous competition between all market participants keeps prices fair.

JesperRavn 11y ago
EDIT: I think you are right, although I would express things a bit differently. I'm writing a top level post with my thoughts. ORIGINAL POST: When you make a trade, there are some information rents associated with that trade. If someone knew you were about to trade, they could capture all of those rents. When there is a monopoly on exploiting these information rents (as in frontrunning), they will be captured by the frontrunner. When there is perfect competition (as when a number of HFTs can trade in another exchange before your trade reaches it), these rents will be all wasted on technology to maximize speed, i.e. no one gets them, they become deadweight loss. I understand that there are other issues, such as knowing whether a trade comes from a single seller or multiple smaller sellers. But this is a separate issue from simply trading ahead of other people. I don't understand why it's so controversial that an exchange would advertise the fact that traders can rely on their orders reaching the exchange before anyone else knows about it as a good thing.
- mrchicity 11y ago
  The original trader still has the ability to capture his information rents in a single trade on one market. Nobody can see it and step away or react before the execution goes through. That's all anyone is guaranteed. HFTs face the same problem. If they're trying to cross the spread to put on a trade or hedge, they also have market impact costs and slippage. I think it's controversial because a lot of their marketing implies all HFTs do this as their primary business and thus add very little value. It creates this narrative of a parasitic activity that intermediates trades which would have already occurred where no risk is taken. I don't think that's the case in reality. All HFT strategies are looking at order books in the products they trade, and they definitely make money (or avoid losses) when someone executes inefficiently across exchanges, but there's more to it than that. HFTs are very successful in futures and non-US markets where products trade on a single exchange. How do they do that if they aren't good at predicting prices and taking risk?
  - JesperRavn 11y ago
    >The original trader still has the ability to capture his information rents in a single trade on one market... That's all anyone is guaranteed Yes this is the essence of it. Based on other posts, I would guess that if you really wanted to make simultaneous trades in multiple markets, then you could. But even if HFT really did render this impossible, it's not a bad thing, since it trading in two exchanges simultaneously actually gives more rents to the large trader than the theoretical ideal of a single continuous double auction. And if you want to trade in A then trade in B based on the results of A, then you certainly don't deserve to be the first to make it to B.