Robot Funds and Consumer Relief

Also Wells Fargo, Uber, IPOs, satiety and repose.


There's a negative way to look at the news that BlackRock Inc. is laying off a bunch of stock-fund managers to focus more on computer-driven quantitative strategies, and a positive way. The negative view is that it's tough out there for an active equity manager. It's hard to outperform the index. People don't want stock-pickers, and are moving their money to index funds. There is a lot of pressure on fees. It's a declining business, which is why BlackRock is slowly backing away from it. As BlackRock's Mark Wiseman puts it:

“We’re in really rough seas, but BlackRock is an aircraft carrier,” Mr. Wiseman said. “Everyone else is in dinghies, and they’re bailing like hell.”

(Another negative way to look at it is that it's particularly hard to be an active equity manager at BlackRock. "The company’s active-equity funds have lagged behind rivals for years.")

The positive way is something like: Equity mutual-fund management is a solved problem. What happened here, in the big picture, is that BlackRock hired a bunch of people to figure out how to pick stocks, and they figured out how to pick stocks, and they wrote it down, and now they can go do something else. The computers will take it from here. That isn't quite how the people involved will perceive it -- it's not like the human stock-pickers were directly working on the computer programs that will displace them, and the quants who were working on those programs will stick around to manage and improve them -- but it feels broadly correct. "At the heart of BlackRock is a culture that embraces change and turns it into opportunity," says Larry Fink, BlackRock's chairman and chief executive officer, in the press release.

When I think about computers replacing humans as mutual-fund managers, I like to think about driving. Self-driving cars are, you know, pretty good, right? They're not going to replace human drivers next week, but it seems like they'll get there eventually. Intuitively, compared to driving, investing seems:

  1. Harder for a human, and
  2. Easier for a computer.

Driving is pretty easy! Lots of people can do it. And yet it requires constant instinctual processing of all sorts of visual information that is hard to codify into simple rules that a computer can manage. Meanwhile, active investing is really hard! Most people who do it can't beat the index. And yet it seems increasingly likely that most of the value that investment managers do add can be explained -- or at least imitated -- by looking at a bunch of numbers (historical stock prices, corporate earnings, etc.) and finding patterns in those numbers. Computers are really good at that sort of thing. (They have neural networks.) So of course robot-driven investment funds should be farther along than robot-driven cars, and fundamental equity mutual-fund managers should be at least as worried about their jobs as truck drivers are.

But in another sense this is pretty weird. Investing is not really about finding patterns in a bunch of numbers. I mean, that is what it's about, but not at a deep level. Deep down, it's about finding good ideas, about understanding what humans will want and who will be best situated to give it to them. It's about figuring out which product will be the next Facebook and which will be the next Friendster, about figuring out who will be the next Steve Jobs and who will be the next Ken Lay. That process is at least as human and instinctual as driving; it takes its inputs from the whole world, and parses them through a deep understanding of human behavior. It seems like it should be hard for a computer to do that.

But the computer doesn't have to. The market does that: People -- entrepreneurs and lenders and venture capitalists and, eventually, public-market stock investors -- make those evaluations, and express those evaluations in the form of prices, and ultimately those prices become the numbers that the computers examine to find patterns. The market is an amazing computer for turning all of those people's individual decisions into aggregated numbers. It just turns out that, if you plug an actual computer into the market, it makes better use of those numbers than the humans do.

Elsewhere: "Evidence That Robots Are Winning the Race for American Jobs." And: "America Must Divorce Dignity From Work."

Consumer relief.

In the years leading up to the financial crisis, big banks did some bad mortgage stuff, giving borrowers mortgages that they probably shouldn't have given them and then selling those mortgages to investors. This bad mortgage stuff had bad effects on the borrowers, who often couldn't pay back their mortgages when the housing market crashed, and on the investors, who didn't get paid back. The government went after the banks for doing the bad stuff. Strictly speaking, the government went after the banks for defrauding the investors, rather than for harming the borrowers. This makes sense: It is easy to explain how the banks defrauded the investors by taking money from them that was not paid back, but it is harder to explain how the banks harmed the borrowers by giving them money that they didn't pay back.

Nonetheless there was a strong intuition that the borrowers had been harmed, and so when the big banks reached their big mortgage-fraud settlements with the government, those settlements tended to include not just fines and repayments to investors, but also lots of money designated for "consumer relief." The idea was that the banks had harmed consumers by giving them mortgages that they couldn't pay back, and could fix that harm by modifying the mortgages -- writing down the principal or lowering the interest rate or whatever -- so that the consumers could pay them back. 

This generally sensible idea ran up against some practical problems. For one thing, some of the banks that did bad mortgage things never actually made mortgages. They bought up other people's mortgages, and packaged them into securities, and the securities were bad, and they contributed to the crisis and so forth -- but they didn't have any consumers to relieve. I was puzzled, for instance, by the Goldman Sachs Group Inc. settlement, which earmarked $1.8 billion for consumer relief, even though Goldman didn't have a consumer mortgage business. Goldman figured out a solution: It buys other banks' mortgages at a discount, and then modifies those. Fine. Weird, but fine. Even weirder, Deutsche Bank AG -- also not a big U.S. mortgage lender -- incurred $4.1 billion of consumer relief obligations, which it fulfills in party by lending money to investors who buy mortgages and modify them. Again, an odd way to punish banks for the financial crisis (Larry Summers calls it "a large systematic overstatement of the burdens borne by banks as they settled mortgage-related claims"), but consumer relief, sure.

But there's another practical problem, which is that the financial crisis was almost a decade ago. There just aren't that many busted subprime loans from 2007 that Deutsche and Goldman can buy up and modify. So, reports Matt Scully, Deutsche has another idea: It is considering "indirectly funding new loans to subprime borrowers."

The bank could do that by lending to companies that offer government-backed mortgages to borrowers that often have weaker credit, the person said. The loans would include Federal Housing Administration mortgages, which allow borrowers to make a down payment equal to as little as 3.5 percent of the price of their house. These types of loans, which leave taxpayers with most of the default risk, have become one of the main ways for subprime borrowers to get mortgages since the housing crisis.

Again: Fine! Subprime borrowers were harmed by the financial crisis, Deutsche Bank contributed to the crisis, and now Deutsche Bank will help them buy new homes. But it does seem a little far afield from the original idea of the mortgage settlements. That idea was that the big banks harmed consumers by giving them subprime loans, and should now help those consumers by writing off some of those subprime loans. Deutsche Bank might instead atone for its role in the subprime crisis by making new subprime loans.

Wells Fargo.

Look, I am not a consumer banker, but it seems to me that the right way to do that business would be:

  1. Give loans and bank accounts to people who want them, and
  2. Don't give loans and bank accounts to people who don't want them.

Obviously there are caveats -- you can't give a loan to everyone who wants one -- but it feels like a reasonable starting point. Wells Fargo & Co. seems to have started from the opposite end. Everyone knows about the fake-accounts scandal, in which Wells Fargo bankers realized that the easiest way to meet their account-opening quotas was by opening accounts for people who didn't want them, or even know about them. That resulted in a $110 million settlement with customers yesterday, with more perhaps still to come. 

But Wells Fargo also got in trouble yesterday for failing its obligations under the Community Reinvestment Act, "a 1977 law intended to promote lending in low-income neighborhoods." While it was giving accounts to people who didn't want them, it was also unfairly denying loans to people who did.

Well, not really. Its CRA scolding from its regulator, the Office of the Comptroller of the Currency, is sort of a weird document. Wells Fargo actually did a pretty good job of lending in underserved areas! ("WFB demonstrated lending levels that reflected excellent responsiveness to the credit needs in the majority of assessment areas," says the OCC, with lending market share exceeding deposit market share in most areas.) Instead, the OCC's list "of discriminatory or other illegal credit practices" that "adversely affected WFB’s overall CRA Performance Rating" consists mostly of older violations (including steering minority borrowers into subprime loans with high fees from 2004 through 2008), as well as the fake-account-scandal itself, which is bad, sure, but not a failure to lend in low-income neighborhoods. The OCC is harsh:

These findings reflect an extensive and pervasive pattern and practice of discriminatory and illegal credit practices across multiple lines of business within the bank, resulting in significant harm to large numbers of consumers.

I mean, you can see where they're coming from! Wells Fargo shouldn't have done this stuff, and the OCC is its regulator, and it would be a little weird for the OCC to just let an opportunity to yell at Wells Fargo slide. The best way to stop banks from doing bad stuff in the future is to embarrass them, over and over again, for the bad stuff they did in the past.

Still I feel like there is a lot of this in current discussions of banks, a sort of everything-is-also-everything-else mentality that makes it hard to talk clearly about the actual problems. Wells Fargo opened lots of fake accounts, which is bad. It also does a good job of lending in underserved areas, which is good. But its report card on lending in underserved areas is Bad, because a general cloud of Bad hangs over Wells Fargo, infecting everything that it does. Even the things that are good!

What's new in ride-sharing?

"Every strength, in excess, is a weakness," says Liane Hornsey, Uber Technologies Inc.'s chief human resources officer, in a moment of Pyrrhonian/Thielian skepticism. But what if her skepticism doesn't go far enough? What if, instead of every strength, in excess, being a weakness, every weakness, in excess, is really a strength? What if a lack of human-resources policies (Hornsey is pretty new), a flagrant disregard for the law, and a cruel corporate culture are the very things that made Uber the massive multinational company it is today? I guess we are saying the same thing.

Anyway Uber released its diversity report, and it's not that bad! "Compared with statistics at other technology companies, Uber’s diversity figures are not that different — and are modestly better than some." Still:

Only 36 percent of Uber’s work force is made up of women, while the technology jobs at the company — some 85 percent — are overwhelmingly held by men. In terms of racial composition, 50 percent of Uber’s employees in the United States are white and 31 percent are Asian, while 9 percent are black and 6 percent are Hispanic.

Hornsey "added that Uber was creating a task force to pinpoint major human resources failings." But Ellen Huet and Carol Hymowitz point out that "the more illuminating piece of data about diversity at Uber isn’t in its report at all": It doesn't report retention rates for different demographic groups.

One theme that I have sometimes seen in Uber coverage is that Uber represents not just a terrible bro-y culture that needs to improve to become an acceptable semi-bro-y culture, but also an almost blank-slate opportunity to start over and create an ideal Silicon Valley culture. (Farhad Manjoo: "This could be the start of a deep, long-term and thorough effort to remake a culture that has long sidelined women — not just at Uber but across the tech business, too.") I don't really get it: Intuitively, it seems to me that it would be easier to turn a very sexist company into a somewhat sexist company, or a somewhat sexist company into a not sexist company, than to turn a very sexist company into a not sexist company. That is just, like, geometry. But the Silicon Valley cult of Pyrrhonian skepticism has deep roots, and people can't resist the idea that the way to get an ideal corporate culture is to start with a terrible one.

Coincidentally yesterday Uber competitor Lyft Inc. also released its own diversity report, which went like this: "We’re woke. Our community is woke, and the U.S. population is woke. ... We’re not the nice guys. We’re a better boyfriend." I guess that is not a formal diversity report. It was a Time interview with Lyft President John Zimmer. Still you get the idea. The battle lines are clearly drawn. Uber: Weakness in Strength. Lyft: Your Woke Bae.

Elsewhere in ride-sharing, "Didi Chuxing, a company in China that last year bested Uber in that huge ride-sharing market, is in negotiations to get SoftBank of Japan to take part in a multibillion-dollar investment round." And elsewhere in tech-flavored car companies, "Tencent Holdings, one of China’s internet giants, has acquired a 5 percent stake in Elon Musk’s electric-car maker Tesla."

Why aren't there IPOs?

We talked about the declining attractiveness of public markets the other day, and I argued that it's not just a matter of increasing regulation: Instead, private markets have become more attractive, so there's just less need for even big multinational private companies to put up with the headaches of going public. But here is Steven Davidoff Solomon with a few more reasons, of which the most interesting might be that "the drop-off in activity is largely attributed to the disappearance of the small offering":

The market for new issues has moved toward liquidity and bigger stocks. Mutual funds prefer making big investments rather than small ones for liquidity and administrative purposes — lots of small investments simply require more people and more monitoring.

Twenty years ago, when 54 percent of initial public offerings "were considered small, with a market capitalization below $75 million in inflation-adjusted dollars," the stock market was a lot less concentrated in the hands of huge investment funds and "quasi-indexers." It is tough to make BlackRock, with its $5.1 trillion in assets under management, care very much about an IPO for a $75 million company. 

Food stuff. 

My Bloomberg View colleague Tyler Cowen is worried that we're spending too much time and energy on food, which makes us fat and sleepy, and not enough on music, which makes us sweaty and energized. "Food," he notes, "especially if combined with wine, encourages a state of satiety and repose." Which is nice individually but might be bad for society. "These days, it could be said that food is the opiate of the educated classes."

Elsewhere in food trends, with coconuts, "getting to a smoother mouth feel is one big technical area." At Olive Garden, shrimp scampi (500 calories) has overtaken chicken Alfredo (1,480 calories) as the best-selling dish. ("We started talking about the flavors first and the taste first," says Olive Garden's vice president of brand marketing about how it ... came up with ... the idea for ... shrimp scampi?) And: ASMR food.

People are worried about unicorns.

The hot thing for unicorns is not M&A any more; now it's IPOs again:

Until recently, startups could count on generous private funding, with the associated generous implied valuations, and avoid the perceived hassle of being accountable to public investors. If a company had both exit options on the table -- an IPO or an outright sale -- the sale option looked attractive.

The pendulum is starting to swing the other way, according to Lise Buyer, founder of IPO advisory firm Class V Group.

Me yesterday.

I wrote about David Einhorn's proposal to split GM's stock into two classes. Elsewhere, S&P Global Ratings and Moody's Investors Service don't like the idea.

Things happen.

Britain Heads Into the Unknown as May Signs Brexit Letter. EU Blocks Deutsche Boerse's $14 Billion Takeover of London Stock Exchange. Toshiba’s U.S. Nuclear Unit Westinghouse Files for Bankruptcy. The Future of Bitcoin Could Be Bitcoin Futures. Ex-Fox News CFO offered immunity from prosecution. Facebook continues to copy Snapchat features. Anbang, Kushner End Manhattan Talks as Lawmakers Cite Conflicts. Trump Said to Meet With Cohn on Thursday to Discuss Tax Overhaul. America's Central Bank: The History and Structure of the Federal Reserve. Does Doing the Same Work Over and Over Again Make You Less Ethical? Megan McArdle on Utah and economic mobility. The Great Nevada Lithium Rush to Fuel the New Economy. A hand can be a gun. Spiders could theoretically eat every human on Earth in one year. 

If you'd like to get Money Stuff in handy email 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

    Before it's here, it's on the Bloomberg Terminal.