Wednesday, April 09, 2014

The Wrong Way, and the Right Way, to Criticize Flash Boys

By Michael Dorf

Flash Boys: A Wall Street Revolt, is the latest book by Michael Lewis, and like his other work--e.g., Liar's Poker; Moneyball; The Blind Side; The Big Short--it's a great read. Flash Boys was previewed in last week's NY Times Magazine, and has been all over the news, so I will only summarize it very briefly here, before coming to my two main points.

Flash Boys tells how, mostly since the onset of the Great Recession, stock trading has become systematically rigged so as to take money out of the pockets of investors and put it into the pockets of "high frequency traders" (HFTs) and the proliferating exchanges and big banks that (literally) rewired the trading rules so as to advantage HFTs.  Lewis explains that the HFTs use superfast computers and connections to deploy various arbitrage schemes that add no value to the market. The scheme that gets the most attention in Flash Boys is electronic front-running.

Here's how it works. Let's say that a big institutional investor places an order for 100,000 shares of Microsoft stock at $40/share. The broker will route the order to multiple exchanges, ostensibly so that the investor gets the best prices, but really so that the HFT algorithms can "see" the order first on one exchange (or in a "dark pool" run by one of the Wall Street banks) and then, using their lightning speed, zip around to the other exchanges to buy stock at $40/share (or less) before the institutional investor gets there, then sell it to the institutional investor at a higher price, such as $40.01 or whatever. By doing this sort of thing constantly, the HFTs are able to exact a small "tax" on just about every large transaction, and thus siphon off billions of dollars annually.

Flash Boys recounts how Brad Katsuyama of Royal Bank of Canada began to notice that something was awry in the markets and eventually traced the problem to HFTs. He then assembled a team of traders and technologists to create a new exchange, IEX, that is designed so as to deny HFTs the opportunity to extract rents from investors. Joe Nocera noted recently in the NY Times that Katsuyama wasn't the first person to try this and that Lewis wasn't the first to write about it, but it's fair to say that the spotlight that the new book shines is considerably brighter than the attention previously paid to the issue by the mainstream media.

Now, on to my observations.

(1) The Unsuccessful HFT Counterattack on Flash Boys

Flash Boys has barely been in print for a week, but predictably, the backlash against Lewis has already begun. I have no deep knowledge of this subject matter, but I did talk to people who know a lot more than I do, and they confirmed my basic impressions. Based on the obvious flaws in what his critics are saying, it looks to me like Lewis is right and his Wall Street/HFT critics are wrong. I've scoured the internet for the responses of the HFT industry, and found that the main talking points of the defenders of HFT are weak.  (For a good summary with links to other sources, I recommend this Bloomberg Businessweek article.) Here are the main points, followed by my analsysis.

--High-frequency trading has dramatically reduced costs to investors compared to the old method of filling orders by hand.

This claim is ubiquitous but it conflates two very different phenomena: computerization and high-frequency trading. Lewis is not against computerization, nor are the heroes of Flash Boys. IEX itself is a fully computerized exchange. The claim of Flash Boys is that most of the HFT activity in the electronic exchanges serves no legitimate investment purpose, and can only be seen as a way of exploiting microsecond-or-shorter-length information advantages so as to extract rents from investors. Lewis and IEX can grant that today's investors keep more of their money than investors on the old manual exchanges did, but their point is that they are now unnecessarily giving up a chunk of cash to HFTs.

--High-frequency trading increases market liquidity.

The book pretty clearly demolishes this claim. Microsecond-by-microsecond buying and selling does not add liquidity. In lots of other circumstances, middlemen serve a very valuable economic function by bringing together sellers and buyers who otherwise would not have found each other. Flash Boys allows that daytraders and the like could serve that function in a stock market that is otherwise illiquid. But he also explains how the vast majority of HFT activity does not add liquidity; insofar as it increases bid/ask spreads, in fact it prevents some trades that would have occurred in its absence. The closest thing I could find to an attempt to rebut this among the Lewis critics was stuff like this from a Financial Times review:

--"Lewis counters [the claim of liquidity enhancement by saying] that HFT creates volatility and a mind-bending amount of activity, not liquidity. But he carries the reader so firmly toward this conclusion that one ends up feeling a bit manipulated, and wondering whether his paean for IEX, the new exchange, is really justified."

Did you see what the reviewer did there? He doesn't actually offer any reasons or evidence against Lewis's view. He just says he feels manipulated by the strength of the narrative.

--High-frequency trading is not as lucrative as Lewis contends.

I came across three versions of this claim. One is that Lewis's numbers are wrong. I don't have the expertise to judge this, so I'll just note that if that's true, then it's hard to see why the people whose business is threatened would feel so threatened. The second version states that, even before IEX opened, HFTs were making less money than they did just a few years ago. Assuming this is true, that may be because the banks have figured out how to capture some of the rents that were formerly going to HFTs or because the competition among HFTs is so fierce that they incur high costs (such as the cost of ever-faster hardware and straighter electronic paths between exchanges). If so, that would not mean that the rents extracted from investors have diminished. The third version of this claim compares the total amount of money made by big Wall Street players to the amounts made by HFTs, and notes that the former is a substantially larger number. At best, this means that the HFTs are effective parasites, extracting enough from the investor hosts to feed themselves but not so much as to kill the hosts.

--Flash Boys and IEX (for example, in this admittedly weird video on its website) give the misleading impression that HFTs prey on mom and pop investors but tactics like front-running only work against large orders.

Oh come on. Most small-time investors who own stock do so through large funds, such as retirement funds, that are the prey for HFTs.

--A good Michael Lewis story involves a virtuous outsider (Billy Beane; Brad Katsuyama) challenging the conventional wisdom of the complacent insiders (baseball scouts;  HFTs and the Wall Street banks) but the HFTs are themselves outsiders, not insiders.

Lewis acknowledges that HFTs are not Wall Street insiders, but in any event, so what? The HFTs can be both outsiders and villains.

Accordingly, the Wall Street rebuttal of Flash Boys strikes me as quite weak.

(2) The Real Problem with Flash Boys

Flash Boys is thus an entertaining read and largely impervious to the criticisms from HFTs and Wall Street. But it is nonetheless problematic in a very different sort of way.  Buried within Flash Boys is a strongly anti-regulatory bias.

Flash Boys ends in a state of uncertainty. IEX is open for business and drawing serious customers, including, for a few days at least, Goldman Sachs, but its future is not guaranteed. The Wall Street banks (including Goldman) could still freeze IEX out. If Katsuyama and IEX fail, the book implies, then that will be that.

But doesn't this overlook an obvious alternative? Surely the SEC or Congress could change the rules so that publicly traded companies must be traded on transparent exchanges that use rules like the ones that IEX uses. And indeed, some industry insiders are already worried that state and federal regulators will will move under existing laws--spurred on by the pressure generated from the media blitz around Flash Boys.

Yet in Flash Boys itself, the government appears chiefly as a malign force. A sub-plot that is only tangentially related to the IEX story involves the federal (and then state) prosecution of Sergey Aleynikov, a Russian programmer who innocuously took some largely useless code with him after he left Goldman Sachs but was inexplicably tried and convicted for doing so. It's not clear what the Aleynikov story is even doing in Flash Boys, except to show that a bunch of ignorant jerks run the government. The SEC makes a brief appearance in the book when Katsuyama visits, but despite the interest from a few old-timers, the younger SEC staff side with the HFTs--and Lewis explains their position by surmising that they plan to go through the revolving door to work for HFTs after their brief government stints are over.

Indeed, even if the government wanted to do the right thing, Lewis seems to think it simply can't. The world of HFTs itself was launched by a 2005 reg that responded to earlier shenanigans, which in turn exploited opportunities created by an earlier set of regulator interventions, and so on, back to the earliest days of financial regulation. As Lewis sees things, any time the government intervenes to cure one problem, it creates opportunities for the better-resourced financial community to do even more damage.

Is that a sensible view of the world? Sometimes it is.  The late great tax law scholar Marty Ginsburg once wrote that "every stick crafted" by the IRS "to beat on the head of a taxpayer will, sooner or later, metamorphose into a large green snake and bite the Commissioner on the hind part." Financial regulation is a similar domain, where firms and individuals with a lot of money at stake will do their darndest to exploit whatever openings the law leaves.

But Marty was exaggerating to make a point. Some regulations are harder to evade than others. And the possibility of perverse consequences hardly counts as a reason for regulators to throw up their hands and invite anarchy. Unless, that is, you are an economic libertarian.

Despite his willingness to go after Wall Street, Lewis does seem to be an economic libertarian. In Flash Boys, he says that salvation will come, if at all, from market forces let loose by Katsuyama and IEX. In Boomerang, a travelogue sequel to The Big Short, Lewis more or less endorses the Tea Party view of what went wrong in the leadup to the financial crisis--casting profligate Greek borrowers and American municipalities as the chief villains in a morality play that doesn't exactly let the banks off the hook, but sees their worst sin as one of stupidity and short-term thinking.

That's also the main story line of The Big Short: The morons tasked with running the big Wall Street banks were too shortsighted to see that their proprietary portfolios were full of worthless CDOs and synthetic CDOs that would come a cropper as soon as the real estate bubble popped, while a handful of clear-eyed outsiders had the foresight to find ways to short the market, even when a looming depression created the risk that a successful bet might go unpaid for want of solvent counter-parties. That was a great story, but it overlooked the role of deregulation--especially the repeal of Glass Steagall--in turning mega-banks into the sorts of institutions in which people could get rich by issuing crap loans, even as they thereby imperiled the broader economy and the banks themselves.

Conversely, if Lewis didn't have so much faith in the beneficence of the market (if only it would work its magic as God intended), then he might regard the "toll" that HFTs extract from transactions as a kind of private Tobin tax on financial transactions that discourages constant trading and encourages buy-and-hold or "value" investing. I wouldn't endorse this view myself. If it's a good idea to impose a tax on financial transactions, then I'd prefer to see the proceeds go to more socially beneficial purposes, but it's telling that Lewis never even considers the possibiilty that the "investors" who are being ripped off by HFTs are themselves not exactly adding value by allocating capital to those who can profit the most from it.

Don't get me wrong.  Michael Lewis is a wonderful writer and he has a knack for finding people whose individual stories provide a window on larger events. But that window tends to open on a particular view of the world in which the government as regulator is either absent or part of the problem. Such stories exist, of course, but so do stories of successful regulation. I hope Lewis sees fit to tell one of those one day.