Money Stuff

Commissions and Chatbots


The central, awkward question in the debate about the Department of Labor's fiduciary rule for brokers who provide retirement advice is: Is bad retirement advice worse than no retirement advice? Like here is a simple hierarchy of things you could do to save for retirement, from best to worst:

  1. Save for retirement in an efficient portfolio of index funds with very low fees.
  2. Save for retirement in a mediocre product with very high fees.
  3. Not save for retirement.

These are not all the options -- option 0 is putting your retirement savings into brilliant venture investments, or winning lottery tickets; option 4 is putting all your retirement savings into a Ponzi scheme, or losing lottery tickets -- but they are fairly common ones. So if some slick-talking hustler shows up at your place of employment and talks you into option 2, has he done you a favor, or done you harm? The answer depends on what you would have done if he hadn't shown up. If you were on your way to Vanguard to buy index funds when he waylaid you, he has moved you from option 1 to option 2, and made you poorer in retirement. If you were on your way to blow your paycheck on lattes at Starbucks, he has moved you from option 3 to option 2, and made you richer in retirement. The context is key.

Here's a story about the slick talkers who go out and sell retirement annuities to teachers for Axa Advisors LLC, who "start by explaining how a teacher’s state pension works, how the tax-advantaged 403(b) operates and what sort of gap might have to be filled with savings to help maintain the teacher’s standard of living in retirement," and then move on to pitch "Axa's Equi-Vest variable deferred annuity," a product with a 460-page prospectus and "a total annual cost that can range from 1.81 to 2.63 percent." The brokers "can earn roughly 5 to 7 percent of the total amount teachers deposit in their 403(b)’s for the first year," and "are paid more to sell annuities than to sell mutual funds." If you think of these brokers as fiduciaries for the teachers, then that sounds like a conflict of interest, since they get paid more to sell a product that charges the teachers higher fees. (Axa says that the higher commissions "reflected the complexity of selling annuities," and that the annuity "offers guarantees not available in mutual funds or index funds.") But there is no particular reason to think of the brokers as fiduciaries for the teachers, except that it would be sort of nice if they were.

But if they weren't selling products with high fees that pay for high commissions, would they be out hustling at faculty meetings to sign teachers up for annuities? And if they weren't, what would the teachers be doing with their money?

“Many school employees would have never taken the initiative to open a retirement account if I had not been there,” said Mr. Jenkins, who left Axa in 2014. “The fees that were built into the annuity product paid for a field staff of agents to go into the schools and reach out to people. I feel good about what I was able to do for them.”

Financial products are "sold, not bought," as the saying goes, and the key element of Axa's pitch seems to be not the stuff about the variable annuity, but the explanation of how much money the teachers will need in retirement. If you crack down on people selling expensive products, it's not obvious that they'll switch to selling better cheaper products. 

Elsewhere, Morgan Stanley has decided that, despite the fiduciary rule, it will still offer commission-based retirement advice:

Morgan Stanley, one of the country’s largest wealth-management firms, said Wednesday that it will allow its nearly 16,000 brokers to continue offering individual retirement accounts that charge investors per-transaction commissions rather than fees as a percentage of an investor’s assets. The approach could mean lower costs for clients that don’t do much trading and could help Morgan Stanley retain and poach advisers who rely on commission-based business.

Merrill Lynch has gone the other way, getting rid of commission-based retirement accounts. The rough math here has to be something like:

  1. Fee-based accounts charge, say, 1 percent of assets per year or whatever.
  2. Some customers in commission-based accounts don't trade much or buy fancy products, so end up spending less than 1 percent per year.
  3. Other customers in commission-based accounts trade a lot and buy stuff with high commissions, so they end up spending more than 1 percent per year.

In a commission-based world, category 3 subsidizes category 2. In a fiduciary world, category 3 shrinks -- it's harder to push customers into high-commission products -- and so category 2 gets harder to justify, so you might just give up and collapse everyone into category 1. (Which is bad for the people who were in category 2, who now pay more.) But category 3 doesn't vanish entirely; Morgan Stanley can still make money on commissions if it promises to act in its customers' best interests, discloses its conflicts, takes only "reasonable" compensation, and probably sets aside a bit more money for fiduciary-rule lawsuits.


Why would you want this:

This week Bank of America, MasterCard and several financial start-ups announced new tools — known as chatbots — that will allow customers to ask questions about their financial accounts, initiate transactions and get financial advice via text messages or services like Facebook Messenger and Amazon’s Echo tower.

I am a hopeless old person, but I have a limited set of questions about my financial accounts -- "how much money is in them," that sort of thing -- that I answer by logging into my accounts and looking at the numbers in them. Natural-language processing and people skills just don't strike me as that important in a bank's electronic interface. Nor does texting. But I grew up on the internet back when there was an internet, and it seems natural to me that when I want to do a bank thing I should go to the bank's website. (Or, fine, app.) Now that there is only Facebook, it must seem weird to young people not to do their banking through Facebook Messenger. I guess you have to go where the people are. Meanwhile the next evolution in bankbots will be empathy:

Critics of chatbots have said that even with artificial intelligence, bots will continue to lack the intuitive and empathetic understanding that humans need when dealing with difficult problems, such as financial decisions.

Are ... bankers ... suffused with intuitive and empathetic understanding, though?

Elsewhere in robots:

Welcome to the world of multifactor investing. It’s the vanguard of money management and a place where no answer is too outlandish to the question, “How many different investment themes can you stuff into an exchange-traded fund?” Following the success of smart beta -- those hybrid ETFs that add a robot’s idea of stock picking to passive indexes -- managers are pitching a new type of diversified portfolio effectively run by a committee of robots.

And why not? A couple of months ago I jokingly sketched a future in which the move from active to passive investing would proceed first by the rise of index funds (a crude algorithm to allocate capital to whatever companies are biggest), and then by the rise of smart beta (a more complex algorithm to allocate capital based on factors that have historically predicted good performance), and then by the rise of smarter and subtler algorithms that allocate capital based on ever finer algorithmic judgments about what data is likely to predict performance. Multifactor is a step along that path, subtler than smart beta because it uses more factors and tries to have good timing on their weightings.  It's "the vanguard of money management" now, but in like a year the cool thing will have even more factors and even more subtlety, until eventually we are reading about ETFs that fully solve the socialist calculation problem by considering and optimizing all the data in the economy to allocate capital to its best use.


There are some companies that announce earnings and you are like, well, that is the thing about that company. It is an entity that sells goods and services, and hopes to sell them at a profit, and when it tells you what that profit was, that gives you a lot of insight into how good it was at selling its goods and services and managing its business. It is all very ... normal. It must sound nice to Deutsche Bank:

Deutsche Bank AG’s surprise third-quarter profit was overshadowed by Chief Executive Officer John Cryan failing to dispel concerns that uncertainty tied to a U.S. settlement will continue to linger.

I sometimes joke that banks are mostly random number generators for paying regulatory settlements, but this isn't just the settlement stuff, it is an overall existential malaise. "These are good overall results, but they are lacking strategic information," says an investor. Meanwhile Cryan, who is nicknamed "Mr. Grumpy," sent a memo to employees saying that "we aim to be more ambitious in headcount reduction." And: "Deutsche Bank Said to Review Its Valuations of Inflation Swaps." And: "Pimco’s Idea for Deutsche Bank Bonuses: Pay in Risky Bonds."

In other bank news, "Barclays Plc said profit rose 35 percent in the third quarter as revenue from fixed-income trading surged to the highest in more than two years."

And in other earnings and strategy news, Tesla Motors Inc.'s plan to acquire SolarCity Corp. is controversial among shareholders, but it is also the heart's desire of CEO Elon Musk. So he has given the shareholders a surprise present to butter them up: profit

Tesla Motors Inc. posted an unexpected profit and said it expects to get through the rest of the year without raising cash, easing one concern for investors as it expands production of electric cars and prepares to acquire money-losing SolarCity Corp.

How thoughtful.

Blockchain blockchain blockchain.

We talked yesterday about's fun strange blockchain experiment, which features a central intermediary that keeps a ledger of transactions, but sort of saves that ledger to the bitcoin blockchain every now and then. Aaron Brown e-mailed:

One aspect of the Overstock blockchain IPO that's interesting is they're using the blockchain for validation rather than convenience. A trusted intermediary implies one that is both honest and competent. If you don't trust the competence of the central intermediary, you use blockchain to record transactions quickly and accurately, but count on the intermediary to slowly get his records into synch. So you use Bitcoin to move money around quickly without all the cumbersome rules and delays and taxes required by the banking system, but you keep most of your wealth in the bank so you are not exposed to hacking or Bitcoin collapse for more than a few recent transactions.

If you don't trust the honesty of the central intermediary, as Patrick Byrne absolutely does not, you use the banking system for transactional convenience, but keep all your money in Bitcoin to protect it from inflation, government or legal system seizure, taxes and other unpleasant stuff.

He concludes: "Most financial blockchain ideas seem to be for transactional reasons, doubting the competence of the financial clearing system, but trusting its honesty. Overstock is the first idea I've come across that does the reverse."

Universal proxy.

Voting in corporate democracy is so pervasively, comically dumb that a Delaware judge has called to move it to the blockchain because even that would make more sense. One dumb thing about it is that when there is a proxy fight, management sends you a proxy statement with an attached proxy card (ballot) listing its board nominees, and the activist insurgent sends you a proxy statement with a ballot listing his nominees, and you clip out whichever ballot you like more and mail it back. The ballots are different colors, so you don't get confused. This is not how ballots work in, you know, U.S. presidential elections. Regular ballots list all the candidates, and you check which ones you want. (You have to check which ones you want on the corporate proxy cards too, it's just that they only list one set or the other, so the checking feels a bit redundant.) But the Securities and Exchange Commission plans to fix that:

The Securities and Exchange Commission today voted to propose amendments to the proxy rules to require parties in a contested election to use universal proxy cards that would include the names of all board of director nominees.  The proposal gives shareholders the ability to vote by proxy for their preferred combination of board candidates, similar to voting in person.

Super! It's worth remembering that this isn't really how it works anyway, since lots of institutional proxy voting is electronic. But isn't it fun that the SEC is modernizing and rationalizing the snail-mail option for voting in corporate elections? In 2016?

People are worried about covered interest parity.

Here's a covered interest parity sort of worry, from Gordon Liao at Harvard Business School:

I document economically large and persistent discrepancies in the pricing of credit risk between corporate bonds denominated in different currencies. The discrepancies amount to 50-100 basis points on trillions of dollars of debt notional. I relate this violation of the Law-of-One-Price (LOOP) in credit markets to violations of covered interest rate parity in the foreign exchange (FX) market. The two seemingly unrelated LOOP violations are closely aligned in the time series and the cross-section of currencies. I develop a model in which LOOP violation and limits of arbitrage in one market spill over to the other market.

Here is the paper.

People are worried about non-GAAP accounting.

One big complaint that people have about non-GAAP accounting is that it is used to juice executive pay: If executives are paid, not based on earnings under generally accepted accounting principles, but based on earnings that they get to adjust to make them look better, then they'll get paid too much. The counter-argument is basically that companies use non-GAAP metrics because they are in some sense more real than the GAAP ones; they do a better job of capturing sustainable, economic income than mechanical application of GAAP would. And you want to reward executives for creating sustainable income. But this seems bad:

Executives of EpiPen maker Mylan NV are unlikely to suffer a reduction in their pay from the company’s recent $465 million settlement of allegations that it improperly overcharged Medicaid for the lifesaving drug.

That is because Mylan historically has calculated executive pay using a nonstandard measure called “adjusted diluted” earnings, which excludes the costs of such litigation settlements, the company’s regulatory filings show.

I mean, I get that litigation settlements are non-recurring, except at banks. But they also seem particularly within the control of executives. Like an ideal executive compensation system would pay an executive for what she can control, ignore random fluctuations that don't reflect her own work, but also very much dock her pay when her company is sued and has to pay hundreds of millions of dollars for misconduct. 

People are worried about unicorns.

Yeti Holdings Inc., which makes coolers, has been working on an initial public offering at a valuation of about $5 billion, which would have been an excellent addition to stock market cryptozoology. Yeti would leave the Enchanted Mountain, have a nice shave, and try to make it in human society. But Yeti may pivot back up the mountain:

The company, which filed for an IPO in July, is currently weighing an equity sale to a select group of investors or the issuance of debt to finance a dividend payment to its current owners, according to people familiar with the matter.

Now is the point in Money Stuff where traditionally I would ask for someone to draw me a yeti unicorn, but I actually think I'm fine without one. 

Elsewhere Snap Inc.'s IPO will be big, raising "as much as $4 billion" and valuing Snapchat at $25 billion to $35 billion or more. And "ZTO Express Inc., the Chinese delivery service that gets most of its business from Alibaba Holding Group Ltd., raised $1.4 billion in the biggest U.S. initial public offering this year."

People are worried about bond market liquidity.

Here is the Bank of England's Bank Underground blog:

Does market liquidity risk affect bond returns more seriously in stress times than in normal times? Does a higher cost of funding for dealers in stress times cause bond returns to be more sensitive to liquidity shocks? Focusing on Euro-dominated investment-grade corporate bonds, we conduct a quantitative analysis in a regime-switching model, and confirm a ‘yes’ answer to both of these questions.

Most of the time when official government agency blogs write about bond market liquidity worries, they tend to come up with "no" answers, so this one is noteworthy.

Things happen.

Harvard Called ‘Lazy, Fat, Stupid’ in Endowment Report Last Year. Platinum assets frozen on allegations of 'plundering' over $200 million. Rhode Island to withdraw capital from 7 hedge funds. "Senior banking officials should attest each year that their companies are free of criminal fraud and civil abuse, said Christy Goldsmith Romero, special inspector general of the Troubled Asset Relief Program." UK posts stronger than expected GDP growth of 0.5%. BoE seeks details of UK exposure to Deutsche and Italian banks. David Rubenstein interviews Lloyd Blankfein. Reaction to AT&T-Time Warner Deal Shows Presidency’s Growing Reach. ICAP-Tullett Deal Said to Face Regulator Query on Licensing. Twitter data is used to suppress Saudi dissidents and catch underage drinkers in Texas. Hotel CEO openly celebrates higher prices after anti-Airbnb law passes. Alphabet Creating Stand-Alone Self-Driving Car Business. The Real Problem with Appraisal Arbitrage. JPMorgan Said to Block FT Site on Concern Over Shared Content. Hamptons Home Prices Fall the Most Since 2013. Brooklyn Nets Guard Jeremy Lin Donates $1 Million to Harvard. Female law students are more ethical than male ones. Banana Surprise. The New Must-Have Yard Accessory: a Feral Cat. Featherless penguin at SeaWorld given special wetsuit. Noel Gallagher Spent Years Moving Furniture Around to Make Liam Believe in Ghosts.

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

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