Banking Villains and Trading Computers

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

Wells Fargo.

One of the most frustrating things about modern bank scandals is that they are not particularly good human dramas, lacking small casts of easily identifiable heroes and villains doing obviously heroic or villainous things. The Wells Fargo & Co. scandal, for instance, involved 5,300 mostly low-level bankers fired for creating fake accounts. They make bad villains: You couldn't get all of them on a stage, or even remember all of their names, and their misdeeds are mitigated by the fact that they were making $12 an hour, afraid of losing their jobs, and faking accounts because they thought it was expected of them.

Of course Wells Fargo also had a chief executive officer who made tens of millions of dollars while this was going on, and who eventually resigned over the fake accounts after being shouted at in Congress for a while, but he is not a wholly satisfying villain either. Like, he didn't open any fake accounts. It seems unlikely that he ever told anyone to open any fake accounts; certainly there's no evidence that he did. It doesn't even seem like he had much reason to want the fake accounts, which after all were not profitable. He clearly missed some red flags -- like employees telling the board about all the fake accounts! -- but inactivity does not make for compelling drama. A villain whose villainy consists of sitting around while other people do bad things isn't that scary. That's "Seinfeld," not "MacBeth."

But it does kind of feel like someone at Wells Fargo has to have done something more dramatic, right? You wouldn't expect 5,300 employees to stumble independently on the idea of making up fake accounts. And the fake accounts were especially concentrated in certain areas, like Southern California. So you could look into the managers in areas where the fraud was concentrated, trace their paths across the country, and draw some conclusions about what happened and who was responsible for it.

This Bloomberg News article does exactly that sort of forensic tracking of managers, with fascinating results, identifying individual Wells Fargo stars who popped up in suspicious places. But here again the conclusions are sort of implicit and negative:

After Wells Fargo & Co. executive John Sotoodeh handed off more than a hundred branches in Southern California to a colleague in 2009, problems surfaced quickly.

His successor, Kim Young, addressing rumors that some employees were opening bogus accounts, called an introductory meeting with staff and warned she wouldn’t tolerate misconduct. Within a few days, managers recall, sales crumbled across her new turf.

It's not that he told employees to open fake accounts. It's that she told them not to, and sales fell. You can draw conclusions! What was he telling them? But you can't quite see the dramatic scene where the manager calls the employees into a room and says "boys, I've thought up a fraud." It remains subtle, implicit, just out of reach.

My favorite story is that at one point, some of Sotoodeh's Southern California bankers went to the southeastern U.S. to train some bankers from Wachovia, which Wells Fargo had acquired. "Soon after the coaches left, the Wachovia staff discovered debit cards in their mailboxes for Wells Fargo accounts they hadn’t requested, one person said." The debit cards just appeared there, magical, unbidden, with no human agency. Even after carefully following the senior managers in charge of Wells Fargo's bad sales practices, the story still feels like one of cultural diffusion, of anonymous ancient peoples moving across the land and leaving archaeological traces of their passage, like stone tools and animal bones and unwanted debit cards.

Elsewhere, here is a Planet Money episode about how Wells Fargo allegedly retaliated against employees who complained about the fake accounts by firing them and noting on their regulatory U5 forms that they had been creating fake accounts. Which is horrifying. But I wonder how many of the 5,300 people fired by Wells Fargo for creating fake accounts now claim that they were actually fired in retaliation for refusing to create fake accounts? How can you be sure what's true? The 5,300 fired bankers are a sad anonymous mass, but at least we knew they did something illegal. Now maybe we ... don't? Maybe that was a lie too? Everyone I've seen or heard interviewed about the scandal describes seeing people create fake accounts, but not doing it themselves. Wells Fargo created 2 million fake accounts, but it's weirdly hard to find even one person who actually did anything wrong.

Hedge funds etc.

Active bad, passive good; hedge funds bad, low-cost strategies good; humans bad, computers good; etc.:

Star traders are losing clients after years of poor returns in a near-zero-rate environment, with managers finding it tough to read economic indicators, predict markets and get an edge in an era of widespread access to information. Investors are turning to model-driven funds in the hope that machines, detached from all emotional bias, are better placed to make money or protect their capital should markets turn volatile.

Here's a claim that I see a lot:

Investors could be ditching their human money managers just when they’re needed most, Gesualdi said, with the U.S. Federal Reserve expected to increase interest rates and spur volatility in global markets. “This is usually the best environment for macro managers,” he said.

Okay so look. It seems reasonable to think that hedge funds/active managers/humans outperform passive strategies/whatever in certain environments and underperform in others. It also seems reasonable to think that investors are hopeless trend-followers and always move into passive strategies after they outperform, meaning that they flee active managers just as those managers are poised for a comeback. But it is a stretch to go from those plausible claims to talking confidently about how hedge-fund (active, etc.) outperformance is just around the corner. I can believe those cycles exist; I find it harder to believe that they are easy to time.

But if you do think that you can time them, surely that is a market opportunity. You start a meta-fund that indexes when indexing is the best choice, uses complex hand-crafted human hedge fund strategies when that is the best choice, and unerringly switches between those strategies at the optimal time. (You could charge 2 and 20 in the hedge-fund times and just hibernate, or work as a bartender, in the indexing times.) That seems like a good product! I'd buy that. But just grumbling that you have a product that works sometimes, but not now, and that people should invest in it now because one day it will work again, seems a bit self-serving. 

Elsewhere, "Vanguard Group smashed its annual record for inflows last year. This year, it surpassed that record in 10 months." "Kentucky plans to pull more than half of its investments in hedge funds in the coming three years." "Former SAC Trader Launches Company Focused on Big Data." Here's Dan Loeb on stock-picking and "quantamental" techniques. And here is famed quant Emanuel Derman with some tweets comparing the free pens given away in bank branches

Speed bumps.

Today's fun fight in U.S. equity market structure is over the Chicago Stock Exchange's proposal to add a 350-microsecond speed bump to its market to protect displayed orders. The idea is that you can post or modify a limit order immediately, but if you want to trade with a resting limit order, you have to wait 350 microseconds. And during that 350 microseconds, the resting limit order can change, and you'd be out of luck. When we talked about it a few months ago I called it "sort of a blind last look" for market makers on the Chicago exchange: When prices change, the market makers get 350 microseconds to change their minds about their quotes before they have to trade on them.

This is controversial, and the Wall Street Journal has an article about the controversy today. It is a high-frequency-trader-versus-high-frequency-trader fight -- basically Virtu Financial LLC likes the speed bump and Citadel Securities and Hudson River Trading LLC don't -- and it has no moral dimension, though you can add one if you want. If you like the speed bump, you can say that it protects legitimate quotes against "latency arbitrage"; if you don't, you can say that it is "unfairly discriminatory." 

But there is one pleasingly cynical reading of the speed bump proposal. This is the technical version of it, from Hudson River Trading's comment letter:

While quotes that can be adjusted due to the LTAD will be less accessible and allow liquidity providers to display tighter quotes and larger quoted sizes, as Protected Quotations, they will result in CHX receiving a greater portion of the Securities Information Processor (“SIP”) market data revenue. CHX notes that prior to these “unusual messaging patterns,” its Time-weighted Average CHX At The NBBO in SPY relative to the total NMS Size At The NBBO in SPY was 44.36% (“Quote Market Share”) and its share of volume in SPY was 5.73%. CHX does not note the unusual disparity between its Quote Market Share and its actual market share. It is important to note that CHX encourages market participants to increase its Quote Market Share by sharing a portion of the SIP Market Data Revenue with the participants that contribute to its Quote Market Share.

CHX appeared to benefit from its geographical distance from the other equities exchanges. This geographical distance may have allowed CHX participants to quote at the NBBO in large size and adjust quotes before orders originating in the NY/NJ area reached CHX. This is consistent with CHX’s high Quote Market Share relative to its share of volume prior to a market participant beginning to trade with CHX’s displayed liquidity and the fact that when a participant began trading with it, the Quote Market Share dropped dramatically. With the LTAD, CHX market participants could again increase Quote Market Share while the LTAD would allow them to adjust their quotes before they became liable to trade.

One important way that stock exchanges make money is that they contribute market data to a consolidated feed, which is sold to trading firms. If the Chicago Stock Exchange often displays the best price on SPY -- the SPDR S&P 500 exchange-traded fund -- then it gets a big share of that revenue, even if people don't actually trade SPY on the Chicago exchange that much. Which they might not do because it takes a long time for their orders to get from New York to Chicago. So market makers could quote a lot on Chicago (which had the best quote 44 percent of the time), but not actually trade a lot (they only traded 5.7 percent of the time), and get basically free market-data revenue without much risk. Then someone started actually trading with the market makers' quotes before they could cancel them -- this is what the exchange calls "latency arbitrage" -- which made this strategy less profitable. So the exchange proposed the speed bump. "CHX’s Mr. Kerin confirmed the speed-bump plan would boost the exchange’s market-data revenue but said that was 'not a driver' behind the proposal."

Elsewhere in market structure, former Commodity Futures Trading Commission member Bart Chilton has some thoughts about the CFTC's proposal to get access to trading firms' source code: "Like the Blob, if this proposal were a reality, it would become acutely carnivorous throughout government."

Blockchain blockchain blockchain.

Here is a Tabb Forum opinion piece about internal blockchains, a concept that I find weird but keep seeing:

On the buy side, in particular, transactions and position balances of securities and cash are tracked in multiple applications and used for different purposes .... Middle-office staff often start their day by ensuring that asset holdings and cash balances in the order management trading system used by portfolio managers to calculate asset allocation reflect the most up-to-date positions. Unfortunately, performing these off-line and manual reconciliations and valuations, and fixing data for reports, diverts staff away from performing their core function: generating returns.

Obviously it would be better to have one integrated computer system that worked fast and was good at keeping track of all of your stuff. "By utilizing a single, immutable source of transactional data, as espoused by proponents of blockchain and distributed ledger technology, the need for multiple internal reconciliations and the risk of trade or reporting errors is vastly reduced." Equally obviously, you can take out the phrase "as espoused by proponents of blockchain and distributed ledger technology" there. There's nothing "distributed" about a single ledger kept by a single firm to keep track of all of its stuff. You don't need the cryptographic innovations of bitcoin's blockchain to just keep one list of your stuff in one place, instead of lots of lists in lots of places.

But if you go to asset managers and say "I will sell you one database that keeps track of all your stuff, which can then feed into all the other applications you use to trade and monitor that stuff," no one will be that excited. (For one thing: Now you have to integrate all those applications with the new database!) But if you instead come in and say: "Blockchain. Blockchain. Blockchain," the managers will jump up and down, whoop with glee, throw papers like confetti, tear down the walls of the conference room in frenzied excitement at the mere word. "Blockchain! Why didn't you say so? We'll take seven." 

Ink.

Here's a story about a guy who pleaded guilty to check fraud:

“The defendant’s commitment to bank fraud is extreme,” said Assistant Brooklyn U.S. Attorney Allon Lifshitz.

The prosecutor added: “Indeed, he has even tattooed the logo of Bank of America — one of his victims — on the inside of his left elbow reflecting the pride he takes in his crimes.”

No one will mess with him in prison, what with his Bank of America tattoo. Don't you kind of wish this was a thing in white-collar crime? Like insider traders getting tattoos of the stocks that they insider trade? Wells Fargo tellers getting "noname@wellsfargo.com" tattoos? Come on, show your extreme commitment to white-collar crime. 

Cyber tronc box.

Here is a Bloomberg Businessweek profile of Michael Ferro and Tronc (or tronc), which is pretty amazing even by the high comedic standards that apply to things about Tronc (or tronc). Ferro wouldn't talk to the Businessweek reporters, so to get insight into his personality and business practices they interviewed rapper Lupe Fiasco. Why not? Also there is a sentence that says: "And yet, until recently, Ferro was on the verge of laughing all the way to the bonc, as it were." Also the bonc might have accepted ... bitcoin?

In early 2014, according to Josh Metnick, former chief technology officer at Wrapports, Ferro got excited about a new product, which he wanted to name tronc—borrowing a word that dates back to the early 20th century, when hotel and restaurant workers would collect service fees in a “tronc,” a fund that would later be communally distributed. The idea, Metnick says, was to create a kind of digital “tronc box,” whereby newspapers would be able to collect micropayments from the readers they were serving far and wide across the web. At around the same time, in February 2014, the Sun-Times became the first major newspaper in the country to accept bitcoin in exchange for access to its pay-walled stories. The cryptocurrency experiment was short-lived, and the cyber tronc box never got off the ground. 

People are worried about unicorns.

Here's a fascinating Bloomberg Decrypted podcast in which a startup founder just ... confesses very explicitly to committing securities fraud? He says that he "cooked the books" and lied to his investors to persuade them to put up more money to keep his company afloat, and it worked, and eventually he sold the company to Twitter, and the investors made money, and everyone was happy, and he's a hero, and he's even got a book out recounting this story. There seems to be a popular belief that securities fraud doesn't count if it's done in early-stage startups. "If this had been public companies and we were on Wall Street, we would all have gone to jail," says the founder. Okay! The law is not ... really that different? This is not legal advice, of course. If you're planning to lie to your investors, and memorialize it in a book and a podcast, you're on your own.

Elsewhere: "Blue Apron Didn’t Get A State Processed Food Registration For Three Years." "Updated Uber App Will Connect Your Calendar With Your Ride." And Slack, the meta-unicorn that invests in other tech companies while also running a workplace chat service, has new competition from Microsoft. (On the chat side, I mean. Microsoft already invested in other tech companies.) Slack took out an ad in the Times to say hey.

People are worried about bond market liquidity.

Here you go: "Euronext and Algomi to Launch New Trading Facility to Improve Liquidity in Pan-European Corporate Bond Trading." 

Things happen.

Fed Signals It’s on Track to Raise Interest Rates in December. High Court delivers blow to UK’s Brexit plansCredit Suisse Drops as One-Time Gain Fuels Quarterly Profit. SocGen Rises After Profit Beats Estimates on Trading Increase. Facebook Defies Social Media Gravity With User and Profit Growth. Ex-BlackRock Fund Manager Mark Lyttleton Pleads Guilty to Insider Trading. Hexagon CEO was under arrest during results call, investors unaware. How Sociopathic Capitalism Came to Rule the World. In the future, we will all be rental serfs. Can Banks Make Wall Street Sexy Again for Millennials? Gawker settled with Hulk Hogan. David Duke at Oktoberfest. The Only Article You Need To Read About Why Trump Voters Are Angry. Anthony Weiner checks into cybersex addiction rehab center. Dumb chatbots. The Bros Who Met Their BFFs on Bumble. "Nerds of all kinds see the mind-calming potential in an Excel sheet, the possibility of order, the potential to spend soothing hours programming automatic cell coloring while a big project bubbles away in the back of your brain." The Cubs won. A Russian Bear Probably Didn’t Officiate A Wedding.

If you'd like to get Money Stuff in handy e-mail form, right in your inbox, please subscribe at this link. Thanks! 

This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.

To contact the author of this story:
Matt Levine at mlevine51@bloomberg.net

To contact the editor responsible for this story:
James Greiff at jgreiff@bloomberg.net