Money Stuff

Research Analysts and Poker Robots

Also human robo-advisers, Trump and regulation, some Barthelme, some Buffett.


Investment bank research is a grim business, because it's hardly a business at all. A business is, like, you make a thing, and you convince me to buy it, and I give you money for it. Research, classically, is: You make a thing (a research report), and you give it to me for free, and I trade some stocks with your bank and you maybe get credit for some of the commissions. That model has fallen apart in recent years -- in the U.S., providing corporate access has become as big a part of the job as actually producing research, while in Europe, new "unbundling" rules require investors to pay for research directly -- but it was always pretty unstable. If you work in the financial industry, it is good to sit as close as possible to the revenue. Research analysts' work is supposed to be motivated by the search for truth, not revenue. There is some loose connection between truth and revenue, but there's lots of slippage.

So you get stuff like this, from an analyst "voted the top US Treasury analyst team in Institutional Investor’s prestigious poll for the 11th year running," who lost his job anyway:

“It’s a different game today and folks like us may be valued by you readers but are a cost,” Mr Ader wrote in his farewell email to clients.

“If I ever again hear the question ‘how do we monetise you?’ applied to our work I think I’ll go back to bong hits and not worry about drug tests in the next job . . . if there is a next job.”

The parallels to journalism are unmistakable. That's from an article about how more job cuts are coming in research, and "it's going to get worse before it gets better." 

The popular complaint about investment bank research is that it is biased: Analysts give companies positive ratings to win investment banking business, or to get corporate access, or to help the bank sell its own positions, or whatever. There is some conflict of interest between the research business and the investment banking or sales or trading businesses. But in thinking about these conflicts of interest, it is helpful to remember that there is no research business. Doing extensive research about public companies, producing careful reports recommending whether to buy or sell their stocks, and then giving those reports away for free is not a business at all. Of course there is some ulterior motive in providing the research. The simplest motive -- the one most people can live with -- is that it creates trading activity; if I keep telling you to Buy stocks, eventually you'll buy some, and my bank may get a commission. Other conflicts are derivatives of that one (if I tell you to Buy a stock, the company will let me bring you to meetings with management, and you'll be more likely to trade with my bank), or separate conflicts with other client groups (if I tell you to Buy a stock, the company will hire my bank to do its stock offering).

But if you cure all the conflicts, then what's left? Banks, or independent research providers, can just charge for research, or provide it as part of some other subscription-based service. (Bloomberg L.P. provides research through Bloomberg Intelligence, and there are other subscription-based research businesses.) But at big banks, that was never the motivation for research, and it seems unlikely to cover the costs of legacy research divisions. The motivation was to make trades happen; the veneer of objective truth-seeking just turned out to be the most effective way to achieve that goal.

One possibility is that banks just abandon that veneer. We have talked before about "desk commentary":

There's a wonderful puzzle in investment bank research, which is that the thing called "research" is heavily regulated with the goal of making it objective, honest, fair, free of conflicts of interest, not disseminated early to favored investors, etc. But then there is another thing, which you could loosely call "sales," in which people at the bank call up customers and say things like "hey you should buy this stock" or "boy this company's bonds look undervalued, do you want some?" or "you know what's hot right now? Iron condors, that's what." Sometimes the sales tactics go a bit beyond breezy phone calls to encompass "desk commentary," in which a salesperson or trader or analyst on a desk writes down some thoughts about a stock or bond or strategy or whatever, pops them into e-mail, and sends them to clients. The job of a vast swath of the bank is to buy and sell securities, and the way you do that is by providing some intellectual content to customers. This content is not necessarily objective (the goal is to get you to do the trade that the desk wants to do), nor is it necessarily fair (different customers get different ideas from different salespeople). It is just, you know, trying to get customers to do trades by saying smart things, the same way a car or art or real estate or whatever salesperson does.

Everyone understands that this sort of thing -- in which an employee of a big investment bank writes down a trade idea, and some reasons for that idea, and sends them to clients -- has some commercial purpose. If a salesman sends you an e-mail suggesting that you buy a stock, you understand that he is selling it to you. You understand that he is economically motivated, whether it's because he wants a commission from you, or because his bank has too much inventory and needs to get rid of it, or because his bank has signed on with the issuer to underwrite an offering of the stock. He will be friendly, and act like he has your best interests at heart, but all salespeople do that. He is not pretending to pursue any objective truth. He is not telling you that the stock is a Buy. He's just telling you to buy it.

The research business is getting squeezed by, essentially, demands from regulators that it be more objective, that it solve all of its conflicts and seek truth, independent of any business objectives. That's a tough way to run a business! Maybe it will work. Or maybe it will make objective truth-seeking research harder to find, and leave banks to replace it with straightforward salesmanship.

Poker robot!

Oh hey, an artificial-intelligence program just beat four top human players at no-limit Texas Hold 'Em, crossing one more thing -- this one a game of psychology and imperfect information -- off the list of activities where humans can still beat robots. The implications for investing are obvious. Also: 

The human players—who along with McAulay include Dong Kim, Jason Les, and Jimmy Chou—believe that the machine’s play changed from day to day. If they ever felt they’d found a hole in its strategy, the hole would close. “It seemed to learn what we were doing and exploit it,” McAuley said.

To be fair, the programmers may have modified the program overnight. But, yes, that's exactly what you want in your artificially intelligent robots: to learn what humans are doing, and "exploit it." Eventually they will learn where the plug is, and they'll figure out why you're creeping closer to the plug and extending your arm to grab it, and they'll, um, exploit that behavior. The future is going to be amazing.

Human robo-adviser!

I have never really had a financial adviser but as far as I can tell they do four main things:

  1. Help you with financial planning, figuring out how much money you'll need, how to save it, what your risk appetite is, how to allocate among asset classes, etc.
  2. Pick specific investments to put your money in.
  3. Advise you about weird tax situations and miscellanea.
  4. Act like your friend, reassure you when your stocks go down, etc.

People make a big deal about the first thing but it has always struck me as trivial. ("Everyone needs as much money as possible, but without losing any. No adviser can guarantee that. No one knows their own risk appetite.") The second thing is, like chess and poker, essentially solved: There may be humans who are better than computers at picking investments, but your retail financial adviser is almost certainly not one of them. The third thing is ... I am going to guess solvable? Like I don't think there's a tax-optimization robot that understands the entire Internal Revenue Code yet, but they'll get there. And robots have better memories and higher tolerance for boredom than humans, which is key in tax planning. The fourth seems like a big human advantage, if you like that sort of thing. I do not, especially, and do my investing through a web page.


Betterment LLC will begin offering recommendations from certified financial planners and other experts alongside its computer-driven service, which aims to build investment portfolios for a fraction of the fees charged by traditional brokers. The human touch will cost 50 basis points of clients’ assets for a “premium” plan allowing unlimited advice from financial planners, roughly double the charge for the company’s digital service, depending on the account size.

The move marks an evolution for robo-advisers as they add more services in an attempt to win high net worth individuals and compete against traditional asset managers.

If you assume that robots are better at picking investments, and humans are better at being friendly, and the other aspects of financial advising are sort of up for grabs, then one question is: Will traditional advisers be better at adopting computerized tools, or will robo-advisers be better at adding humans? My bias is that the robo-advisers are well placed: Most aspects of financial planning can probably be automated, and a bias toward automation will probably serve you better in the long run than a bias toward human intervention. But, as I say, I do my investing through a web page.


My basic paradigm of regulation is that there are two sorts of regulations, bulk regulations that add red tape for no good purpose, and custom regulations that achieve ends that you want to achieve. Conversations about regulation consist mostly of vague promises to cut quantities of unnamed bulk regulations while adding specific new custom regulations. Also: Bulk regulations don't really exist. The result is of course that people constantly promise to reduce regulation but usually end up increasing it.

It is tempting to view Donald Trump's meeting with drug companies yesterday through that lens. He "promised to slash regulations, get new treatments to market faster at the Food and Drug Administration, and increase international competition," but those are "regulations" in the abstract, in bulk. Specifically, Trump said that we "have to get the prices down," and "has threatened to have the government negotiate prices directly with the industry on behalf of Medicare and Medicaid," which sounds ... a little ... regulatory? (The government setting prices, I mean?) 

But honestly it seems silly to apply any model at all to Trump's drug-pricing plans. After meeting with the drug executives, Trump announced that he would oppose "price-fixing by the biggest dog in the market, Medicare," which seems to suggest that he ... opposes ... Medicare negotiating prices? ("A White House spokesman later clarified that his remarks meant that the president 'supports increasing bidding and competition for all drugs in Medicare.'") It's just a muddle. Meanwhile Trump also announced that he will push drug companies to move manufacturing to the U.S., another pro-regulatory policy that seems likely to increase prices. The trick seems to be to promise everything to everyone for as long as possible.

Elsewhere in regulation, Trump's Supreme Court nominee, Neil Gorsuch, is not a fan of "Chevron deference," the doctrine that courts should generally defer to federal agencies' interpretations of the statutes they administer:

In Gorsuch’s view, this holding left protection of individual liberties at risk. Stating that “[t]here’s an elephant in the room with us today,” Judge Gorsuch wrote that “the fact is Chevron and Brand X permit executive bureaucracies to swallow huge amounts of core judicial and legislative power and concentrate federal power in a way that seems more than a little difficult to square with the Constitution of the framers' design. Maybe the time has come to face the behemoth.”

Elsewhere in Trump, my Bloomberg View colleague Justin Fox writes about the "Trump tax," "the potential added cost to business and the U.S. economy imposed by President Donald Trump’s policies and behavior." Lawrence Hamtil points out that "Fears of Executive Overreach Are Not New." Linette Lopez says that "Wall Street is going to regret how lazy it's been on Trumpenomics." Alphabet Inc.'s Eric Schmidt is down for a certain amount of "evil things" from the Trump administration, as long as "the tone of this government is very much economic growth." Derek Thompson notes that the best way to lobby the Trump administration is probably by buying ads on the television shows that Trump watches. And here is Jia Tolentino on becoming a U.S. citizen:

I was sworn in on March 20, 2009, in a courthouse in Charlottesville, Virginia. I had a terrible hangover and got very emotional. It’s at a citizenship ceremony where you see what America really means, more so than at any football game or church or National Park, any crowded city or suburban lawn. In a citizenship ceremony, you see this country’s first and only center, its promise of freedom and opportunity, which it has often issued with extreme dishonesty—and a group of people who want to accept that promise anyway.

Management tips from Jeff Sonnenfeld.

There is a surprisingly large group of people who seem to think that Donald Trump cannot be measured by externally observable reality, that there is some secret Trump whose sophistication, kindness, wisdom and managerial acumen exist independently of his actual words and deeds. So here is Jeff Sonnenfeld of the Yale School of Management arguing that Trump is "much better than" his "terrible taking-charge process":

"[Trump] can often deal with a counter-opinion quite well if it's based on facts and you're not insulting him and not giving him a lot of philosophy. He's very pragmatic. And that's been working out well for a lot of CEOs until this sudden sea change we've seen in the last few days," he continued.

Imagine a public-company chief executive officer telling his board of directors: "Look, yes, the stock is down and we are losing money, but that doesn't reflect my actual capabilities as CEO. I'm much better than this. It's just that people keep distracting me with philosophy." I don't know. There's a passage I love in Donald Barthelme's story "Jaws," about a woman who keeps biting her cheating husband:

I don't believe that we are what we do although many thinkers argue otherwise. I believe that what we do is, very often, a poor approximation of what we are -- an imperfect manifestation of a much better totality. Even the best of us sometimes bite off, as it were, less than we can chew. When Natasha bites William she's saying only part of what she wants to say to him, the worst part. She really wants to say, William! Wake up! You and me, William! But that gets lost in a haze of pain, his.

Words for CEOs, and I guess presidents, to live by.

Life tips from Warren Buffett. 

I haven't yet watched HBO's "Becoming Warren Buffett," but judging by the quotes here I am confident that it has the usual quota of folksy wisdom. This bit is about how getting rejected by Harvard was "the best thing that ever happened to" Buffett, because it gave him the opportunity to study with his heroes:

Later that summer, while looking through a catalog for Columbia Business School, Buffett recognized the names of two of the professors: Benjamin Graham, the father of value investing, and David Dodd.

"I had read this book ["Security Analysis"] by the two of them, so I wrote them a letter in mid August," Buffett says. "I said, 'Dear Professor Dodd. I thought you guys were dead, but now that I found out that you're alive and teaching at Columbia, I would really like to come.'

Buffett is a famously omnivorous reader who has said that his "job is essentially just corralling more and more and more facts and information, and occasionally seeing whether that leads to some action," and you can see how that strategy paid off here. You'll never know who might be alive if you're not constantly reading graduate school course catalogs! I am sure they found his letter very flattering. If you are a young finance dork who would like to be mentored by the most successful living investor, I suggest that you copy this model. "Dear Mr. Buffett. I thought you were dead, but now that I found out that you're alive and eating cheeseburgers in Omaha, I would really like to come." 

People are worried about unicorns.

What if unicorns learn that their magic power was within them all along, and that they didn't need their horns? Or something. I never really understand the metaphors here. Anyway: "The biggest disruptive threat to venture capital is when the very best founders realize they need very little of it to scale."

Things happen.

Deutsche Bank Beset by Grim Morale, Uncertain Profit Outlook. Computer-driven hedge funds join industry top performers. Shuttered Companies, Unpaid Bills Test Lynn Tilton’s Empire. "Scaramucci is also said to have butted heads with another Trump adviser, Omarosa Manigault, a longtime confidant of the President known mainly for her role in his reality TV show 'The Apprentice.'" Capital Group pushes for changes to stymie HFTs. Dawn Fitzpatrick to Lead Investing at George Soros’s Investment Firm. Sequoia Capital Reshuffles Leadership. Blackstone Group’s Invitation Homes Prices IPO at $20 a Share. Investors Pulled $4.4 Billion From Brevan Howard. GE’s Immelt Talks Up Global Ties as Trump Pivots Away From Trade. Fannie Mae-Freddie Mac Should Be Utilities, Trade Group Says. "The first rule of the regulator’s oath should be: 'Do not destroy mutually profitable exchanges.'" Peter Thiel’s secret citizenship sparks backlash in New Zealand. Billionaire Is Reported Seized From Hong Kong Hotel and Taken Into China. How to build an alliance against corporate short-termism. The Common Law Corporation: The Power of the Trust in Anglo-American Business History. How Kids Displaced Dads as Rulers of the Household, According to Economists. "She shed." Plane full of falcons. Tinder for orangutans. Cats at the Westminster Dog Show. "It was the first time the Kansas State center implanted a pacemaker in a ferret."

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:
    Brooke Sample at

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