Dark Pools – TwoMorePages Book Review

Dark Pools – TwoMorePages Book Review

This is hands down the best financial book I have read to date. It was even better than Flash Boys, by my favorite financial author Michael Lewis, the book which prompted me to read further to understand HFT (high frequency trading) better. Indeed, I’ll go so far as to say that it is even better than Michael Lewis’s The Big Short , which everyone (including me) loves.

Compared to Flash Boys, It did a much better job of illustrating the problems associated with high frequency trading; why I, as a normal investor should care; and, best of all, helped to illuminate how HFT came about in the first place.

Major takeaways

What was extremely surprising was learning about the ironic nature of what happened: practices that were originally put in place to help the average investor by punishing and taking advantage of rent-seeking intermediaries (market makers and specialists on the floors of the NASDAQ and NYSE) eventually morphed to become the the very practices which now prey on average investors, acting as an unseen intermediary that drives costs up. At least before, you knew you were getting fucked because you were crossing large bid/ask spreads from market makers; the fucking was transparent. Now, you’ll put in an order to buy or sell across what looks like a thin spread, but then transact at worse than you expected. EVERY. TIME.

The other surprising, but extremely relevant takeaway from this book was the revelation that it’s not necessarily the speed in which HFT firms trade that is the problem. That’s what most people initially conclude: “Oh, well nobody can think in terms of fractions of seconds. That’s how HFT firms are taking advantage of me.” Rather, it is the special order types that HFT are allowed to use, combined with their speed, that gives them their alpha, their edge over you and me and any other investor. Instead of being limited to normal market orders, or limit orders, like you and me, HFT get to use special orders that most people don’t even know about: ones that allow the to provide liquidity only when they want to (when they’re going to make money risk free), and step out of the way all the other times whenever shit is going down (like during the flash crash).

Less surprising, and covered in Flash Boys by Michael Lewis, was the revelation that HFT firms basically front run all of us any time that we make trades. Because of delays between one exchange and another, they can tell when say, Fidelity is purchasing a lot of shares of a specific company (for their customers presumably), and can then pull their original offer to sell at $X , and then put in a new offer to sell at $X + some amount. So then Fidelity ends up paying more for the stock than they would have otherwise.

So, why do I care?

Now, that all may mean fuck-all to you in isolation. “Who cares if they make money scalping pennies here and there? I just index and buy every once in awhile through my 401(k). I’m too small for HFT to care about, and I trade so infrequently that it shouldn’t matter” is an initial thought that I had when reading this book.

But here’s why you and I should care: We’re the ones getting fucked. Let’s take an example: Your 401(k) buys stocks of Google. Mine does inherently, since I purchase shares of the S&P 500 broad market index, and google is part of that index. Your orders, plus mine, plus those of several thousand others all pool at our custodian, Fidelity. The Fidelity fund manager needs to buy say 10,000 shares of Google and sees on his trading screen that there are currently 50,000 shares of Google offered (available to buy) at $X. So he puts in his order, thinking that he’ll buy 10,000 shares at $X.

Well, because of the tiniest delays in communication between exchanges, HFT firms can detect that there is buying demand in the market once Fidelity (my 401(k) custodian) puts in that buy order. In that fraction of a second between when the buy order first hits an exchange, the original offers to sell at $X disappear, and re-appear at $X + $.01 or $.02 (or some other arbitrary adder). This happens again and again until the orders eventually fill at those higher prices. So even though Fidelity thought it would be able to buy at $X with plenty of liquidity (since, after all, that’s what the exchanges showed), it actually isn’t able to.

Well, our 401(k) purchase order is basically a market order, so we sweep the prices up, paying say $X + $.10 in the process, let’s say. Well, now we just paid $.05 more than fair value, more than we had to. All these little things add up.

In a world where costs very much matter (especially if you are an boglehead / indexer like I am), these costs are HUGE. My expense ratio in my S&P 500 fund at Fidelity says I am only paying .05% . BUT, if you factor in the slippage (how much more we paid for Google in this example over $X), our expense ratio actually ends up being MUCH MORE than we bargained for, which costs us, you and me, a lot in the long run, especially when factoring in compounding gains.

The story

Okay, so we addressed why you should care. Why else would you read this book? Well, for one, it is actually entertaining. Much like Michael Lewis, Scott Patterson does a really good job of illustrating a story when making his points. He does a great job of starting from the beginning, and then building on the reader’s building knowledge base. If you didn’t know anything about HFT when you started, you wouldn’t be at a disadvantage at all. The way he illustrates what is happening, both through his characters and the actions and motivations of said characters, is extremely easy to understand.

And remember, these aren’t fictional characters in a book he’s writing. These are real people. These were real stories. These things really happened.

Josh Levine and Island

The story of how Island got created takes up about ⅓ to ½ of the book, and for good reason. Island is basically the backbone that started HFT. It was the first matching engine that took people completely out of the buy/sell matching equation. Whereas before, you’d have to pass through a human broker or market maker in order to buy/sell anything, and run into human errors (or greed) along the way, Island gave people a way to instantly match buyer and seller with no middleman in the way to skim profits off the top. It was also perfectly scalable so that as more people got into the market of computerized scalping, the matching engine wouldn’t slow down, as the old NYSE or NASDAQ often did.

I do find it very interesting that Josh Levine basically made Island because he saw an extremely inefficient system where middlemen (market makers and specialists on the floor of exchanges) were skimming tons of money from regular people by basically standing in the middle of a buyer and seller and taking a fee. He thought that was inefficient and dumb, and sought to create a way where buyer and seller could find each other without having to cross large bid/ask spreads along the way. He was a strong proponent of decimalization, so that stocks could trade at $X.01, or $X.23 instead of only $X + ⅛ , $X + ¼, etc. I’m only 30, so I can’t remember living in a world where I’d have to buy stocks in increments of ⅛. That seems archaic and bizarre to me, and Josh Levine is a big reason why I don’t live in a world like that. That’s crazy.

But I digress. Basicaly, Levine sought to make a more efficient system, much like I used to try to do and continue to do when I encounter what I think are stupid and inefficient systems at work. And HE DID IT! And along the way, he tried to solve problems, like how to get more liquidity on Island. The result? The maker-taker fee system that the book goes on to describe as a major problem with HFT trades. It seemed to make so much sense at the time; whoops. Unintended consequences.

The story of Island to me is so interesting, both in describing how it came to be, how it unseated the business model of major exchanges like the NASDAQ and the NYSE, the problems it faced as it grew, and ultimately, what happened to its original senior management as it grew up and got bought. Moreover, the description of Josh Levine’s motivations, how he wasn’t motivated by how to make more money, but by how to mold an inefficient market into an efficient one (and bring down entrenched intermediaries in the process), was absolutely fascinating to me. I loved it.

Haim Bodek and special order types

Aside from Josh Levine, the only other person that could be considered a protagonist in this story (if it were fictional, which it’s not) would be Haim Bodek. His story is an interesting one in the book, showing how he ran a very successful HFT firm where he learned how the traditional limit / market orders that basically everyone uses will always lead to getting taken advantage of. His conclusion that you had to know about secret order types that the exchanges don’t publish info about in order to make money (either as a HFT or as a lay trader) is pretty damning.

Because of the special order types that HFT firms use, Haim Bodek concludes that all that visible liquidity is fake. They can step out of the way and cancel their order in the smallest fraction of a second between you submitting your buy/sell order, only to sell to you at a higher price (or buy from you at a lower price) than you had originally shown.

That’s precisely the problem that Michael Lewis describes in his book, Flash Boys. His protagonist trader notices that every time he tries to transact, he transacts at a worse price than he expects.

Bodek’s main conclusion is that it’s not the speed in which HFT firms can trade that is the problem; it is the secret order types that are, that allow HFT to basically make risk free profit at the expense of everyone else.

Many of Haim Bodek’s critics may attack him, saying that he ran his HFT into the ground because he couldn’t keep up with the times, and now his attacks on the industry are just sour grapses. Well, I don’t buy that. Any criticism I see of him tends to concentrate on him as a person and not on his arguments. In fact, the next book I’m reading is by Haim Bodek himself, where he tries to illuminate the problem and explain in further detail what Scott Patterson touched on in this book.


I used to be a proponent of HFT, saying things like “They provide lots of liquidity. See the bid/ask spread on that stock? It’s less than a penny, because of HFT. In the past, that would have been much larger.” Well, turns out I was wrong, and this book is a great tool to illustrate exactly why. More than that, it describes the history behind HFT and the problems inherent with it in an extremely easy to understand and entertaining way. It’s not a textbook that tries to explain things to you in a dry fashion; you learn through the eyes of the players in the story as it develops.

This is 100% the best non-fiction book that I’ve read so far this year. Kudos, Scott Patterson, for shedding light on something that isn’t really that easy to understand. I can’t imagine how hard it was to get the info to write this book.