Matt Levine sheds more light on this story[1], backed by evidence whereas the NYTimes is just hearsay. Why would Barclay's screw over institutional investors who account for a large majority of their $4 billion in revenue for HFT who bring in only $3 million? Because without HFT (which is a bad, bad word to the ATG's ears), nobody would be trading on it, and nobody wants to admit that. It just doesn't make sense.
I don't understand how a darkpool could exist without rogue HF traders bumping up the revenue from stock trades. If the current stock market won't allow for it, what makes them think a private market will? At the end of the day whoever operates the pool has to foot the bill if they're trading outside of the official exchange.. Unless they turn it into a Ponzi-type scenario or outright lie to their investors.
The basic theory of a dark pool is that by restricting who can access the other participants in the market, you can provide for your dark pool's clients better execution costs.
That is, by only allowing similar market participants (think other hedge funds, pension funds, etc) and excluding "predatory" speculative market participants (HFT, day traders, pit traders, etc.) you can match "natural" trades to each other, without paying the middle man.
In reality, this never happens. Speculative "predatory" traders are a necessary component of the market and without them there isn't sufficient liquidity for the market to operate.
In this particular case, an ibank stands accused of lying about this fundamental fact to their clients. It has nothing to do with the underlying validity of the market structure.
HFT add liquidity on a second by second basis, but nothing on even a short term basis. Market makers speed up transactions slightly, but the cost for doing so is vary high.
IEX is trying something slightly different to get around allowing/paying market makers into their dark pool. They are using publicity and Michael Lewis to try to convince retail investors to provide free liquidity for their backers.
I guess "for free" was a bad phrase. I meant that IEX is hoping for retail investors to make up for the lack of market makers who get paid via the bid/ask spread.
The article explains why it makes sense. The institutional investors' trades don't match up that often. If you allow HFT's then they will buy/sell on other exchanges (or in other dark pools) that match the other side of your (barclays in this example) dark pool's institutional investors trades.
You don't charge the HFT firms since they allow you to charge your other investors.
Whatever part of the newspaper it appears in, the article provides a nice illustration of data presentation. Use of colours, sampling, even the choice of axes and time-scales.
It's ironic that the "dark" in "dark pool" is now being interpreted in a negative manner, synonomous with "opaque".
The reason such pools are dark is to reduce the market impact of large orders. If an institutional investor (e.g. a pension fund) were to announce to the world "We are going to sell a shitload of Apple shares tomorrow!", the Apple share price would fall, as everyone front-ran the trade. That would result in a lower price for the pension fund when it actually got around to selling its shares. Regulators actually recognise the advantages in reducing market impact and they specifically allow delays in reporting large, off-exchange trades between clients and broker-dealers to the market for that reason.
Dark pools were conceived of as a mechanism to allow institutional investors to trade off-exchange (in theory, with one another), so that they didn't have to post bid orders publicly on the "lit" exchange. The reason they're called "dark pools" is because you can't see the order book, so you don't get to see big orders being added to the book (and take advantage of the knowledge that someone's just placed a big order, to front-run them). In other words, you don't know how deep the pool (i.e. the liquidity on the order book) is - just like you can't see how deep a dark pool of water is.
As someone who works in financial markets, it's weirdly fascinating to watch terms like "dark pools" and HFT end up being used as epithets, almost, by the media.
It's a bit like how the word "hacker" came to be regarded by many as describing a computer criminal.
That author seems to think that this was okay behaviour by Barclays, because the clients had choices and could think for themselves(ie caveat emptor)
I do not see how lying to your clients about the type of pool you have is ok. If it is necessary to have some HFT operators in your pool then do not be a chicken and come out and say so. Is honesty not the best policy anymore?
You can be getting the best deals(fills in this case) and still feel like getting a bad deal. It is human psychology.
Let's say you find the best deal on a specific car for $10,000. Dealer claims that you are not paying any dealer markup over retail. However, later you find out that there was a special promotion from manufacturer, where dealer got the cars for $1,000. Even though you realize this was not something you could get yourself you would still feel miffed that you were not given a larger discount.
I admit my example doesn't have very realistic numbers.
I think he pretty clearly comes down against false advertising and giving up your fiduciary duty to your clients. Those are the real issues in the court case.
His main point, and one that seems to be missing in most articles about dark pools, hft, etc. is that buy side investors are as sophisticated (or should be if they are to get away with charging their crazy management fees) as sell side participants. The whole reason they have high paying finance jobs is to provide to their clients the service of making sure they are getting the best execution they can.
Strangely enough, I had never heard of a "darkpool" until today. I bought Flash Boys at the airport bookstore earlier, and read it on the plane just now. And tonight I find a reference to darkpools on the HN front-page. Hmmm. Truth really is stranger than fiction sometimes.
Anyway, FWIW, if anybody here hasn't read Flash Boys by Michael Lewis, it's a pretty interesting read that covers some ground related to the content of this article: HFT, dark-pools, etc. I understand that it's not without some controversy, but I found it damn interesting all the same.
As someone who has worked in the industry, Flash Boys is literally the worst intro you could find into electronic trading. Dark Pools by Scott Patterson is much, much better (and even it gets basic facts incorrect).
Agreed, as someone who worked in HFT and thus has a positive bias towards it, I found Dark Pools to be an extremely informative and factual historical account of the industry even if the author's opinions disagree with mine.
Flash Boys however I can't even get through because the author backs an extremist negative position with completely incorrect facts. I fill with seething rage every time I realize how much more public exposure Flash Boys gets.
I wouldn't say he backs an extremist negative position. In fact, if you read carefully, he is quite adamant in his assertion that HFT is great. As long as it is in the service of large buy side institutions he has no problem with it. If on the other hand, HFT firms dare to upend the relationship with traditional ibanks, then he gets upset.
Well, if it has nothing else going for it, it was interesting enough to pique my curiosity and make me want to read more. To my way of thinking, any book that accomplishes at least that, is somewhat worthwhile.
You can't trust people. Even supposedly-trustworthy people working for supposedly-trustworthy household-name financial institutions like Barclays. What then can we trust? Technologies like BitCoin are predecated on the idea that we can trust mathematics and peer-reviewed logic. Are these mechanisms inherently more trustworthy than individual humans and human institutions? If this is truly the case, then the argument for financial intermediation to be founded on a similar technological basis is an exceptionally strong one. Anybody else interested in following this rabbit hole to see where it leads?
No he doesn't. Either through ignorance, incompetence, or malice he wrote a pretty terrible book about dark pools and HFTs. Dark Pools by Patterson is much better and even it misses on lots of details.
[1] http://www.bloombergview.com/articles/2014-06-26/barclays-no...