Money Stuff

Robots, Research and Taxes

Also mattress wars and bitcoin fanatics.

Will a robot do your job?

Do you think C-3PO would be a good investment banker? He'd be able to do discounted cash flow calculations in his head, would remember all the relevant comparable deals and their multiples, and could give you a precise estimate of the probability that you would achieve your targeted synergies. He knows all the languages, which would help with cross-border deals. On the other hand he is a big nerd. He is probably awful at golf. He'd be a bummer at the closing dinner. Investment banking is a business of trust and relationships, and I could definitely see him missing deals to a less comprehensive but more fun human investment banker.

Of course C-3PO's biggest disadvantage is that he is fictional. The more realistic question is, do you think that the phone-chat robot who does customer support for your cable company would be a good investment banker? Every deal would quickly break down as the clients shouted "representative representative representative."

Will robots take all the jobs at investment banks? Well here's one "top executive at a major Wall Street bank":

“It’s, um,” he says, shifting tone and making clear he can’t speak publicly. That question is gnawing on him, he confides, because he has kids. “I would want them to pick their careers very carefully. I think AI is going to eliminate most jobs. That’s a private view. I think we’re just starting to feel that.”

The precedent here is the hollowing-out of equities and currencies trading floors: Once upon a time, you needed armies of humans to trade stocks, with each of them responsible for tracking orders and deciding what price to show for a handful of stocks. Then people realized that you could reduce those traders' heuristics to algorithms, and then one computer could do that for all of the stocks, more quickly and reliably than the humans do. And now a lot fewer people trade stocks.

That precedent is obviously relevant. But investment banks trade things with a wide range of liquidity and customization, and intuitively, the computers are going to take over the most standardized things first and take a lot longer on the weirder ones. Equities and currencies: Very liquid, very fungible, very susceptible to computers. Bonds and swaps: A harder nut for the computers to crack, with fewer central order books and more high-touch client interaction, but no one really doubts that they'll get there. Companies, though -- mergers and acquisitions -- are almost totally illiquid; there are no order books at all for the merger market. Transactions that have been, or can be, reduced to order handling and pricing are ripe for robot takeover. Transactions that are one-off, that are about bringing humans together to make entirely new combinations, are harder. 

I used to work at Goldman Sachs Group Inc. building equity derivatives for corporate clients. That job exists because computers have taken over equity trading. You can build equity derivatives that require complex dynamic hedging precisely because you can program a computer to do the hedging, and because computerized trading reduces the transaction costs of that hedging. If you automate the building blocks and make them cheaper, you can build more complicated structures with them. The robots take over the dumb routine stuff and the humans move on to higher-value activities, as they do in pretty much every other industry that has ever been automated. 

Also I mean:

Inder Thukral, co-founder of Kognetics LLC, is making machine-learning algorithms that ease life for investment bankers working around the clock on mergers and acquisitions.

“It’s the spirit-crushing work that it’s replacing,” he said in an interview. Dealmakers should focus on only the most important stuff, he said. “Searching through 6,000 documents is a computer’s job, so let it do it.”

One possibility is that instead of needing 100 analysts working 100 hours a week, more fully automated banks will only need 50 analysts working 100 hours a week. But another obvious approach would be to keep the 100 analysts and work them only 50 hours a week. Surely there are hours to be cut from the investment banking schedule. 

That said, here's a roundup of what Wall Street jobs are going to be taken over by robots, and a lot of human-interaction-type activities are definitely being targeted. "Banks are working on cataloging trader and salesperson conversations to create profiles of clients to help better anticipate their desires," and it's a bit tough on a salesperson who has a computer looking over her shoulder, reading and listening to her conversations, so that it can one day replace her.

Elsewhere, here is a story about how much more fun it is to be an auditor now that there are computers. "Technology takes the tedium out of auditing," is the headline: a happy story about auditors' jobs getting better, rather than a sad story about their jobs getting scarcer. The big accounting firms are adding staff. "A large part of the role of an auditor is interactive and people-facing, and it requires judgment," says a Deloitte partner. "These are the kinds of human abilities not likely to be replaced by automation."

Research prices.

As everyone expected, the banks that are required to charge explicit prices for research by European regulations will also be allowed to charge those prices by U.S. regulations. That was not necessarily a given -- "brokers say SEC rules technically prohibit them from selling stand-alone analysis unless they register with the agency as investment advisers" -- but the Securities and Exchange Commission "is expected to provide formal assurances that it won’t object if brokerages break out the cost of market analysis for their European clients, rather than bundling it together with other services." 

The moves may disappoint investor advocates and some officials at U.S. pension funds because they’ve been urging the SEC to to go even further. They want the regulator to allow what’s happening in Europe to also happen in the U.S. -- a system where investors can easily buy research from one brokerage, while paying another to execute trades.

The European rules, which will go into effect in January, are about protecting investors from sneaky bundling; the concern is that fund managers overpay banks for trading execution (with their investors' money) in order to get access to "free" research, and the solution is to force the managers to pay for the research directly (either with their own money, or with the investors' money but in a fully disclosed way).

The U.S. rules, which went into effect last century, are about protecting investors from paying for bad or conflicted research. One would hope that that would not be a serious concern for big institutional investors, who can choose what research they like. The European rules are adapted to the actual state of the market in 2017, in which people mostly invest through funds and in which hiding the ball on fees is the big threat to their returns. The U.S. rules are adapted to the state of the market in the 1930s, and are focused on protecting retail investors who choose to buy individual stocks based on the advice of their brokers. Perhaps the conflict will spur the U.S. to rethink that focus, but more likely the SEC will just issue some no-action letters and call it a day.

Meanwhile what will the prices be? "Goldman Sachs Group Inc. has asked some clients to pay $30,000 a year for up to 10 of their staff to access basic research through its analyst portal once the MiFID II rules come into force in January, according to people with knowledge of the matter," but the prices are all over the place:

Goldman’s package compares with the fee of about $40,000 a year that UBS Group AG proposes charging some clients for about five users to access basic equity research. JPMorgan Chase & Co. plans to charge as little as $10,000 a year for some customers to access its equity research portal, the lowest price to emerge so far.

And here's a list of various options that run up to $455,00 for a "gold" package from Barclays Plc. Would you have expected that? One odd thing about investment banking, compared to many other businesses, is the rigid standardization of its fees, which are often set by reference to lists of precedent transactions rather than by analyzing the costs or value of actually doing the transactions. Nobody wants to distinguish themselves based on price; better for the entire industry to be a premium business. I wouldn't have been surprised if every bank ended up with more or less the same fee structure. 

A tax trade.

If you are very rich, there is no reason to pay taxes on most of your income. I mean, sure, in some loose sense the law requires it, but in practice you can build yourself something like a consumption tax system. If you need to spend your money, on consumption, then it will need to be income to you -- it will need to come into your bank account -- and so you will probably need to pay taxes on it. (This is not legal or tax advice, and of course there are popular approaches that involve making this money a loan rather than income in order to defer taxes.) But if you are very rich you will have no immediate need to spend most of your income, and if you are well advised you can often find ways to park it in some nebulous non-place so that it is not taxed as income to you.

If you later end up spending it, well, sure, income. But if you approach the end of your earning years and realized that you have just plain earned too much, then you can give the money away to charity without ever paying taxes on it. Here is a nice Wall Street Journal story about how George Soros has moved some $18 billion of his money to his Open Society Foundations. (Disclosure: After high school, I had a summer job at a debate camp in the Czech Republic sponsored by the Open Society Institute, true story.) And here is a Bloomberg News story on the tax reasoning that partly explains the move:

Money managers have until the end of the year to pay taxes on fees they earned from investors in offshore funds and had deferred payment on. Many are now turning to charitable donations -- including to their own foundations -- to help offset the levies coming due.

Soros earned money "offshore," didn't pay taxes on it, and now can move it to his own foundation -- whose investment activities he still partly controls -- to continue not paying taxes. Really, this is fine. Income that is given away to charity is supposed to be tax deductible, and in a sense Soros never got to enjoy this money: It sat in its nebulous non-place for a while, and then went to his foundation, without ever passing through his pockets. In another sense, though, I mean, I don't know, would enjoy having billions of dollars, even if I couldn't spend all of it. And it does end up looking, at a high level, like a consumption tax regime for the rich, where they are taxed only on what they need rather than on what they earn. Since most people need most of what they earn, and pay taxes on all of it, this can seem a bit regressive.


This story about the war among online mattress companies and the bloggers who review their products is absolutely wild. One thing that I think about a lot is the stereotype that Wall Street is less ethical, less honest, more conflicted than the rest of the business world. Is that true? Or is it more that expectations of Wall Street are higher, that we expect brokers and bankers to be trusted advisers rather than self-interested salespeople? (Of course, the brokers and bankers usually bring that on themselves, by holding themselves out as trusted advisers.) But I tell you what, mattress salesepeople are not subject to any fiduciary rules. Mattress bloggers can get away with conflicts of interest that would get securities analysts barred from the industry. "It pains us to see you (or anyone) recommend a competitor over us," one mattress company chief executive officer told a blogger. "I am confident we can offer you a much bigger commercial relationship." Imagine what the SEC would do with emails like that.

Hey so how are we doing.

Yesterday was not a great day for some fringe financial figures. Marc Faber, a popular recurring character on financial television, said a lot of really racist stuff in his newsletter, throwing in an "I am not a racist but" for good measure. He was quickly kicked off three corporate boards and banned from several television channels, including Bloomberg's, so let's never talk about him again. I am sure someone else can step up to tell you to buy gold. 

Meanwhile, former White House communications director and hedge-fund marketer Anthony Scaramucci has moved on to his next act. "How many Jews were killed in the Holocaust?," asked the Twitter account of Scaramucci's somewhat-existent media company ScaramucciPost innocently, offering a choice of responses from "less than one million" to "more than 5 million." It is just asking questions here. The post was deleted and blame assigned to a ScaramucciPost employee, who -- naturally -- "apologize[d] if anyone was offended by the Holocaust poll." Scaramucci eventually tweeted that he was "pained imagining that my post led anyone to believe I am giving comfort to Holocaust deniers," expressed his "full support" for the employee, gave the correct answer to the poll (six million, oddly not one of the original choices), and also apologized "if anyone was offended." Maybe they should do a Twitter poll to see if anyone was offended?

Let me say something here. These people are not important. They are not titans of industry or allocators of capital or government officials or accurate analysts of markets. Perhaps they used to be, but now they just trade on their outrageousness, and their burning desire to remain relevant is going to drive them to take increasingly desperate measures. Maybe just ... stop, with them? Like, you don't have to have them on your television show. You don't have to go to Scaramucci's conference, or to the launch party for his next pretend media venture. You don't have to -- and of course I am guilty here -- pay attention to their provocations, or parse their coy retractions. They will not go away of their own accord, but if we ignore them then that is almost as good.

Blockchain blockchain blockchain.

Here's the story of a man who left his children homeless so he could concentrate on trading dogecoin:

They have moved to a campsite in the Netherlands, where they're waiting for bitcoin to really take off.

It's only been a few months, but the 39-year-old father of three says he doesn't regret a thing. "We were just like – sell it, sell it, what can we lose? Yeah, we can lose all the material stuff. Yeah, we can lose all our money. Yeah, we don't have three cars anymore. We don't have the motorcycle anymore. But in the end, I think we, as a family, will still be happy and just enjoying life."

What can we lose? Literally everything we have? Sure, let's do it! 

Elsewhere, last week was the "Most Exciting Electroneum Week," which is good news if you sold all your possessions to buy electroneum. And: "New UWS Restaurant Will Let You Pay With Bitcoin, And Only Uses Pink Himalayan Salt." And here is a "Primer on Virtual Currencies" from the U.S. Commodity Futures Trading Commission's fintech unit, "LabCFTC." 

People are worried about unicorns.


SoftBank’s efforts to become a dominant investor in the US have run into roadblocks in Washington, where national security regulators have yet to clear at least three of the Japanese conglomerate’s deals struck since Donald Trump became president.

These aren't the $100 billion SoftBank Vision Fund's big investments in private U.S. technology companies, but still, the unicorns hoping for some of that largesse might be a little worried by this news. What if their SoftBank money gets held up for national security reasons too? Isn't there a national-security case for having a giant and fully valued unicorn sector?

Things happen.

Why Investors Can’t Get Enough of Tajikistan’s Debt. J.P. Morgan to Buy Payments Firm WePay in First Major Fintech Acquisition. Top Credit Suisse owner backs part of break-up plan. Stock Exchanges Question SEC’s Plan to Revamp Trading. Wall St banks boosted by lending to rich clients. This Is Why Rich Asian Clans Are Pouring Money Into Tech Startups. A Shunned Industry Makes a Comeback in Ireland—Banking. Hedge funder’s case against Preet Bharara dismissed. Man did not enjoy human-resources job at Google. Bull runs into stroller in Brooklyn.

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.