Cool Goldman and Angry Whistleblowers

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


In 1869, an immigrant in Philadelphia who had "worked as a peddler with a horse-drawn cart and later as a shopkeeper," and who was in possession of a terrific beard, moved to New York, where he "hung out a shingle on Pine Street" as a commercial-paper broker. The firm that he founded grew over the years, becoming a massive success, and then an American institution, and then "a great vampire squid wrapped around the face of humanity." Oops! Too far. His successors are trying to dial it back a bit. "Please, call us Marcus," they say, getting back to their roots. "Goldman, Sachs & Co. sounds so stuffy":

After much internal discussion, the Wall Street firm has decided to call the retail banking operation Marcus — the first name of the company’s founder, Marcus Goldman.

Can you imagine if that works? If Goldman Sachs could cure years of accumulated distrust with a hipster-y name for its online retail bank? "You know," young people will say, "back when I associated Goldman Sachs with the last names of two guys, I thought it was a shadowy monstrosity that made money by cheating and gambling and that probably caused the financial crisis. But now that I associate it with the first name of just one guy, I realize how sweet and earnest it actually is, and I've opened a savings account." 

I guess that is how branding works, though. "Marcus is in line with recent start-ups taking on simple first names — think Earnest and Oscar — and is intended to convey a tech-era trendiness from a company that is not known for its youthful bona fides." (I have previously suggested that Goldman could change its name to Honest Dollar, after a startup that it acquired, but Marcus is even better.) And Goldman is in a sweet spot for the coming future in which banks will become tech companies: Its Marcus online loan product will compete with fintech startups like LendingClub and Prosper, and "could have a big advantage over those lenders because it is a bank with steady funding," but it is also pretty tech-savvy for a big bank, and "does not have the additional expense of branches around the country." 

Disclosure: I used to work at Goldman Sachs, and I have a savings account at its online bank. I mean: I have a savings account at Marcus! I have a Harry's razor! I have a Casper mattress! Life is great here, in the casual Friday of late capitalism.

Angry whistleblower!

There are two broad categories of corporate whistleblowers. There are people who report fraud because they participate in fraud and there is no honor among thieves. And then there are people who are a bit too pure for this fallen world, who see practices that their colleagues all think are fine but who are so scandalized that they immediately go to the authorities, the press, anyone who'll listen. Neither type is perfect. The first type is a little disreputable. If you have received two whistleblower awards for two unrelated frauds, or if you sue your former employer for letting you overcharge your own customers, I am sorry but I am going to lock up the silverware when you come over to my house. The second type can be a little exhausting, finding scandals everywhere they look.

The Securities and Exchange Commission has made a big push to pay whistleblowers for information, and one thing to notice is that that approach mostly appeals to the first sort of whistleblower. If you like committing fraud, you'll love getting immunity for that fraud and being paid for reporting it! The second type of whistleblower, though, is so pure that he will sometimes turn down the whistleblower award because it too is a scandal:

Eric Ben-Artzi, a former Deutsche risk officer, has told the SEC he is declining his share of a $16.5m payout — the third largest in the whistleblower programme’s history — awarded for information that led the agency to fine Deutsche Bank $55m last year. The SEC found Deutsche misstated its accounts at the height of the financial crisis by improperly valuing a giant derivatives position.

Mr Ben-Artzi said the fine should be paid by individual executives, not shareholders, and suggested the “revolving door” of senior personnel between the SEC and Germany’s largest bank had played a role in executives going unpunished.

He ... blew the whistle ... on his own ... whistleblower award. Here is an op-ed he wrote about it. (His lawyer, and his ex-wife, will still get their cut.) I love it. Needless to say, the mis-marking scandal that he exposed at Deutsche Bank was possibly a bit soft. Vicious fraud this wasn't; it was a series of accounting choices that offended Ben-Artzi's sense of purity. Which is clearly quite finely developed.

Tick sizes.

Here is a funny article about the Securities and Exchange Commission's tick-size pilot, where starting on October 3, some small-cap U.S. stocks will have to be quoted in five-cent instead of one-cent increments. The problem is that this requires people to update their software to stop quoting those stocks in pennies, and no matter how hard you try, you'll never get everyone to update their software:

There are signs the industry isn’t ready to comply. Although exchanges have spent years getting ready, tweaking their trading software to accommodate the shift, some investors have been caught off guard, according to an executive from Instinet LLC. That’s a problem because if a money manager’s orders don’t comply with the new rules, they will be immediately rejected.

The automation and electronification of the stock market have had a lot of efficiency benefits, but it also makes it sort of sludgily difficult to do stuff like this. In the old days, if you switched stock quotes from sixteenths to pennies, a guy would call and say "20-1/8," and you'd be like "no, dummy, pennies, remember?" And he'd be like "oh right fine 20.12 I guess." But now someone's computer won't be programmed for the wider tick size, and will bid $20.12 when it should be $20.15, and it'll be rejected, and the computer won't know what to do, and it will run around screaming its computer head off, and all the other computers will freak out too, and the whole system will go horribly haywire. And everyone knows this and no one can do anything about it, because the stock market is now just computers talking to each other, and the computers are perfectly rational but not quite reasonable.

Elsewhere in market structure, the recent bust of 46 alleged members and associates of La Cosa Nostra included the owner of Princeton Securities Group, a New York Stock Exchange floor broker. Humans aren't perfect either. 

Valeant lawsuit.

Shareholder disclosure lawsuits are sort of dumb because they work like this:

  1. A company does a bad thing and doesn't tell anyone about it.
  2. People find out and the stock drops.
  3. People who owned the stock before it dropped sue, claiming that the company should have told them about the bad thing.
  4. The company settles.
  5. The company writes a big check to the people who owned the stock before it dropped, which is effectively paid for by the people who still own the stock. (Many of them are the same people!)
  6. The lawyers take a cut.

I mean, you shouldn't do bad things, and if you do you should tell people about them, but this is not necessarily the most efficient enforcement mechanism. (Or maybe it is, I don't know, but it doesn't look efficient.) In any case, this sort of limits the appeal of these lawsuits for giant institutional investors, since they tend to own stock in a lot of companies both before and after the stock drop, so suing would just transfer money from themselves to themselves, with a stop for the lawyers to take a cut. But T. Rowe Price once owned 21.5 million shares of Valeant worth more than $4.5 billion at their peak, and then Valeant had a series of awkward revelations about its sales practices and accounting, and the stock plunged, and T. Rowe sold most of its shares (it's now under 1.5 million), so it's in a good spot to sue, and it did. Here is the complaint, and while its general outlines will be familiar to you if you have followed Valeant's crisis, it is like 200 pages long. So if you want a thorough, partisan overview of that crisis, it's a good source.


Here's a fascinating story about a guy who found weird stuff in the LendingClub data set. The basic weird thing he found is that some people seem to have taken out multiple loans, often at different interest rates, but LendingClub didn't really flag that for anyone, leaving a situation where borrowers were riskier than they seemed and equally risky loans had different interest rates. One bizarre subcategory of this is that some LendingClub employees took out multiple loans, apparently to boost reported volumes, which is very much a no-no. But most of it is just actual borrowers taking out loans, and the system just sort of not noticing.

Is that bad? One thing that I find generally interesting about LendingClub is that, broadly speaking, its promise is:

  1. We will give you a lot of data about our loans, but
  2. We will not give you the data that we don't give you.

In the last great wave of securitization -- private-label subprime mortgage loan securitization before the financial crisis -- there was some generic expectation that the banks underwrote the loans "correctly." There were specific representations about credit scores and loan-to-value ratios, but there were also fudge words about how the banks could waive those requirements if there were compensating factors. The idea was that the bank knew what it was doing, and was doing a good job. And then when it turned out the banks weren't doing a good job, there were crises and scandals and lawsuits.

I feel like platform lenders are more like: Here are 50 data points about this loan. Do what you want. If the loans perform well, then the data points were predictive. If they don't, they weren't. Either way, it was an interesting experiment. But that's it. There's no promise that they're the right data points. And if there's no field for "does this person have another LendingClub loan," and if that data point would have been helpful, well, sometimes that happens.

RIP Gawker.

Gawker Media has been sold in bankruptcy to Univision, which will keep operating most of its sites but which will shut Of course Gawker has done some bad things, and people have different and strongly-held opinions about it, but I just feel like: Do you read the internet? Do you notice that sometimes it's good? That it can be funny? Conversational and interesting and opinionated and bold and direct and joyful? A lot of that, everywhere, is Gawker's doing; it did more than any other site to create the style of the internet that we all read and write today. I have never worked at Gawker Media, have never really had anything to do with it (I went to a party once), but it just seems obvious to me that I would not be typing these words into this box had Gawker not existed. I'm sad that it won't any more. Anyway Nick Denton went out with a "Blade Runner" reference.

People are worried about unicorns.

People are pretty worried about Uber's plans to replace drivers with self-driving cars. "Uber Begins Its Endgame: Replacing Humans," is the Motherboard headline. Here is a Business Insider interview in which Uber Chief Executive Officer Travis Kalanick says, somewhat implausibly, that "in an autonomous world," the number of jobs for human drivers will go up. He also says, more plausibly, that "we are talking about a million lives a year being saved" in the move to self-driving cars. If you believe that self-driving cars will eventually be far safer than human-driven ones and save millions of lives (and I do), and if you believe that they will also eliminate jobs for tens of thousands of drivers (and I also do), what do you do with that information? I mean, it seems obvious enough to me what you do. It's a weird ideology that thinks that saving millions of lives is bad if it also costs some jobs. Like, you could just retrain the Uber drivers and give them jobs as murderers, and we'd still be better off on net.

In other Uber-is-evil news: "Judge rejects Uber’s $100 million settlement with drivers." And here is an Uber blog post about its new initiative that has a funny diagram:

I guess that is a little scary.

Elsewhere in unicorns, the SEC is investigating Hampton Creek for its alleged mayo spoofing. I mean, it's not a spoofing investigation; it's an investigation into whether Hampton Creek made misrepresentations about its sales while raising capital. (Because of the spoofing.) You might be surprised, by the way, to see a vegan mayonnaise company classified as a unicorn -- classically, a private venture-backed tech startup with a valuation of at least $1 billion -- but Hampton Creek "has raised more than $220 million" and "expects to raise another round of financing by early next month that will value the company at $1.1 billion," and Bloomberg calls it a "food technology startup," so that's good enough for me. Though the word "technology" there might be a bit gratuitous. 

And elsewhere in corporate cryptozoology, we talked yesterday about Bigfoot Project Investments, the pseudo-$10 billion company that is looking for Bigfoot, but I forgot to mention that a dog with a GoPro may have found him. I hope the dog has gotten some funding.

People are worried about stock buybacks.

Ooh no the hip thing these days is dividends:

Companies have long spent large sums on dividends and share repurchases to boost their share prices. Investors now are clamoring more for dividends as a result of plunging interest rates, and companies are loath to disappoint them despite slow growth that is pressuring earnings at many firms. The dividend yield on the S&P 500, or annual payouts as a share of the current price, has been steadily above the 10-year U.S. Treasury yield for most of 2016, after only occasionally doing so for decades.

I tend to take a boringly corporate-finance view of things, and don't really distinguish dividends and buybacks, but I feel like the People Who Worry About Stock Buybacks don't worry so much about dividends. They don't boost earnings per share, so they don't have the air of financial manipulation that buybacks apparently have.

People are worried about bond market liquidity.

This is not entirely accurate, but as a rough rule of thumb it seems fair to say that human traders like opacity, because it rewards them for accumulating information and developing relationships, while electronic traders like transparency, because the more information that is publicly available, the more of it they can crunch quickly to develop accurate prices and good trading strategies. Right now Treasury trades happen more or less without any disclosure, and the SEC is seeking comment on a Financial Industry Regulatory Authority proposal to require some disclosure, and some people think that's bad, while some people -- the ones with robots -- think it doesn't go far enough:

Citadel LLC, KCG Holdings Inc. and Virtu Financial Inc. are among firms with a message for the U.S. Securities and Exchange Commission: Regulators’ plan to shed more light on dealings in the $13.4 trillion Treasuries market needs to be more comprehensive.

The Finra proposal would, for now, just require reporting to regulators, not to the public. KCG wants to go further, "to require real-time reporting and immediate public dissemination of collected data."

Things happen.

Fight for Viacom Is Said to End With the Redstones in Control. Citigroup’s Last Proprietary Trader Walks Out the Door. Man Who Introduced Millions to Bitcoin Says Blockchain Is a Bust. Hedge Fund Och-Ziff Considered Sale of Part of Firm. Grand Plans to Revive Chicago’s Stock Exchange—From China. "Infamous rogue trader Nick Leeson, whose losses triggered the 1995 collapse of Barings Bank, has launched a free course for traders to learn from his 'vast wealth of knowledge and experience', as well as his mistakes." At Jefferies Equity Unit, Many Changes After Bad 2016 Start. "Credit Suisse Group AG is accusing a group of five investment bankers who left for Jefferies Group LLC in May of stealing confidential information and trying to coax former colleagues to join them." Ex-HSBC FX Trader Is Waiting for DOJ in Leafy London Suburb. Monte dei Paschi CEO, former chairman under investigation - source. CFTC Charges Deutsche Bank AG with Multiple Swap Reporting Violations, Related Supervision Failures, and Violation of a Prior CFTC Order. Mongolia "is staring at a full-blown balance of payments crisis." Donald Trump's tax plan "would provide negative tax rates for investments financed with debt, creating incentives for companies to pursue projects that wouldn’t make sense economically without the tax benefits." LinkedIn vs the Bots. Boozy seltzer. "NYC Parks stands firmly against any unpermitted erection in city parks, no matter how small."

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

To contact the editor responsible for this story:
James Greiff at