Money Stuff

Index Rules and Analyst Fatigue

Also Goldman trading, email pranks, Bitcoin Cash and GAAP complaints.


In recent years, founders of big tech companies have realized that they can take their companies public while still maintaining control in perpetuity, by setting up dual- or even triple-class stock structures and issuing low- or even nonvoting stock to the public. A lot of people have worried about this, but now they can stop worrying: That trend is over. I mean, Facebook Inc. and Alphabet Inc. and Snap Inc. have already done it, but don't expect too many more, because S&P Dow Jones Indices has announced that the S&P 500 and some of its other big indexes "will no longer add companies with multiple share class structures." Existing multiple-class companies (like Alphabet and Facebook -- but not Snap, which never made it into the S&P 500) will stay.

The S&P 500 is probably the biggest driver of passive demand for U.S. stocks, and S&P's decision should more or less end low-vote initial public offerings in the U.S. (Was the last one Blue Apron Holdings Inc.?) Why would you do an IPO if you can never make it into the index? Why would a banker advise you to? Why would investors buy your stock, if they can't flip it to index funds? The consequences of a low-vote IPO have always been sort of vague and theoretical; now they are real and immediate.

When FTSE Russell announced last week that it would exclude low-vote stocks from its Russell indexes, I had mixed feelings. On the one hand, given how much investors complain about nonvoting stock, it's nice that now maybe they'll just stop buying it. On the other hand:

What happened here was not that the pickiest and most careful investors scrutinized Snap closely and made the bold decision not to buy its stock. Instead, the very least picky possible investors -- the index funds, whose mandate is to buy all the stocks in the market -- are the ones who won't be buying Snap. It makes no sense: In a reasonable world, you'd expect the passive funds to be passive buyers of whatever the market provides, while active funds would make active decisions about what the market should provide. In our actual world, though, the passive funds have all the power, and the active funds are constrained to follow their lead. 

That's even more true of S&P's decision. The key governance debate of the modern market -- whether companies should be allowed to raise equity capital without giving their shareholders voting rights -- was fought and decided, not at the Securities and Exchange Commission or in Congress, and not by investors themselves making independent decisions about what governance protections they needed, but by a couple of for-profit index providers. 

By the way, some S&P indexes will still allow nonvoting stock:

The S&P Global BMI Indices and S&P Total Market Index are broad market indices intended to represent the investment universe. ... S&P DJI has determined that the methodologies for these indices should not consider governance arrangements when selecting the universe of constituents. Therefore, the methodologies for these indices are not being modified.

This is entirely sensible. S&P divides its indexes into indexes that represent "the investment universe," the entire market, and those that represent a selected subset of that market. If you want to buy the whole market, you get everything, good or bad, voting or not. If you want to buy the selected subset, you get a subset selected based on the S&P's own criteria. It is free to set those criteria, and those criteria are naturally going to be biased in favor of some notion of goodness. I mean, the basic criterion for the S&P 500 is bigness, and there were already requirements about public float and net income designed to weed out undesirable companies. Now there will be voting requirements too.

But not all of the consumers of the S&P 500 think that way. In casual usage, the S&P 500 is "the market," and it is easy to confuse it with "the investment universe." If you are investing in an S&P 500 index fund, or comparing a fund's performance to the S&P 500, it's probably not because you have given some thought to the ability of analysts at S&P Dow Jones Indices to pick good stocks, compared it to the ability of analysts at Fidelity or Capital Group to pick good stocks, and decided that S&P is the best. It's because at some level you thought that S&P Dow Jones Indices wasn't picking stocks, that the stocks in the S&P 500 were just there, that S&P 500 membership was a fact of nature rather than a decision. This is a good reminder that it doesn't work that way.

Elsewhere in index criteria: "Governance through Shame and Aspiration: Index Creation and Corporate Behavior in Japan."

It's hard work being a research analyst.

The job of a research analyst, oversimplifying dramatically, is to come in to work, think about whether people should buy a stock, and then put out a report saying whether they should buy the stock and what its price should be. Some days, though, things get a little hectic, and the analyst has to think about whether people should buy two different stocks, or even three or more, and put out two or more reports saying whether they should buy the stocks and what their prices should be. You might imagine that it's harder to analyze two stocks than it is to analyze one, and you would be right. In particular, it's harder to analyze the second stock, because you are tired from analyzing the first, so you are more likely to analyze it lazily and get the answer wrong:

Analysts cover multiple firms and often issue several forecasts in a single day. We find that forecast accuracy declines over the course of a day as the number of forecasts the analyst has already issued increases. Also consistent with decision fatigue, we find that the more forecasts an analyst issues, the higher the likelihood the analyst resorts to more heuristic decisions by herding more closely with the consensus forecast and also by self-herding (i.e., reissuing their own previous outstanding forecasts).

That is from  "Decision Fatigue and Heuristic Analyst Forecasts" by David A. Hirshleifer, Ben Lourie and Siew Hong Teoh of the University of California, Irvine and Yaron Levi of the University of Southern California. It is delightful, and full of somewhat speculative insight into what an analyst's workday might be like:

Yet another alternative explanation—one that we cannot rule out entirely—is that analysts choose to structure their workday by first working on forecasts for which they have high-quality information relative to the consensus. This would explain both the higher accuracy of early forecasts and the lower tendency in such forecasts toward herding or self-herding. However, it is not obvious why analysts would follow such a work strategy. It may make sense for an analyst to prioritize making forecasts for firms about which the analyst has better information. However, this could just as easily entail making a well-informed forecast at the end of a workday, deferring the ill-informed forecast for the start of the next workday.

The most famous decision-fatigue study, which the authors of this paper cite, is the one about how parole boards are more likely to rule favorably just after lunch. Obviously one wants to know whether the same is true of research analysts -- are they more likely to issue a Buy rating after a good lunch? -- but it is harder to tell. Lourie told me by email: 

We know the order of the forecasts and when they were issued, but can't know for sure how much time it took to write. There is a clear decline in the number of forecasts issued during lunch but no statistical difference to separate those that are issued straight after lunch. 

How's Goldman Sachs doing?

Not great! It had "first-half trading revenue that trailed its major banking rivals—a first since Goldman went public in 1999." One oversimplified interpretation of this is that Goldman has, relative to its peers, focused on being a purveyor of high-margin premium products: less "we will buy 100,000 shares of stock for you," more "we will build a bespoke derivative for you." And the modern world is increasingly suspicious of premium financial products: Derivatives are "weapons of mass destruction" after the financial crisis, and paying up for fancy trades is disfavored in a market dominated by index funds.

A steadily rising stock market has also pushed investors away from risky investments and toward simpler products ​where giant banks such as J.P. Morgan Chase & Co. dominate.

Goldman’s trading desk, by contrast, has been more geared toward hedge funds—which have been less active as they face outflows and more trading moves to exchanges.

“The business changed around them,” said James Mitchell, an analyst with Buckingham Research Group. “If you’re a hedge fund and you’re worried about investors pulling money next quarter, are you really going to be asking Goldman to build you a five-year yen swap?”

Disclosure: I used to work at Goldman Sachs Group Inc., building high-margin premium derivatives, so I am quite biased here. It seems to me that there was once a whole economy of fun in the financial system: Fun hedge funds would trade fun products at high margins, which made it fun to work at the banks, even if sometimes it ended in crisis. Boring index funds will trade boring products with boring banks at low margins, which is probably an improvement for society, but something is lost aesthetically.

Elsewhere: "Fidelity Tries to Out-Vanguard Vanguard With Latest Fee Cuts." And the New York Fed's Liberty Street Economics blog asks: "Were Banks ‘Boring’ before the Repeal of Glass-Steagall?" (The answer is no.) 

Email prankster.

There is a guy in the U.K. who emails bank chief executive officers pretending to be their board members, tries to goad them into saying stupid things, and then triumphantly publishes the results. In the past I have found his project a little tiresome, because he is always so pleased with himself, and there is always so much hand-wringing about What It Means For Email Security, but in fact the bank CEOs always respond with bland professional politeness and the results are never that funny. Bank CEOs, it turns out, just keep it appropriate over email.

But then the prankster moved on from bank CEOs to Trump White House officials, who have no ability whatsoever to keep it appropriate over email, and got much funnier results. For instance, he pretended to be Reince Priebus and emailed Anthony Scaramucci during the minute or two when Scaramucci was the White House communications director:

The very real Scaramucci responded: "You know what you did. We all do. Even today. But rest assured we were prepared. A Man would apologize."

Fake Priebus wrote back: "I can't believe you are questioning my ethics! The so called 'Mooch', who can't even manage his first week in the White House without leaving upset in his wake. I have nothing to apologize for."

Actual Scaramucci responded: "Read Shakespeare. Particularly Othello. You are right there. My family is fine by the way and will thrive. I know what you did. No more replies from me."

What character from "Othello" does Scaramucci think he is? The pillow?

In other Scaramucci news, the sale of his SkyBridge Capital stake to HNA Group Co. is still on. (And "is not contingent on Anthony Scaramucci's job," an HNA spokesman found it necessary to say.) And: "Anthony Scaramucci erroneously listed as dead in the new Harvard Law alumni directory."

Blockchain blockchain blockchain.

Happy Bitcoin Cash Fork Day! The basic story is that the bitcoin community can't get their act together to agree on how bitcoin should work, so they're forking it into two different blockchains. If you own a bitcoin as of this afternoon, then you will continue to own it on the old Bitcoin blockchain, and you'll also own it on the new Bitcoin Cash blockchain, meaning that you'll in effect have one Old Bitcoin and one Cash Bitcoin. That's nice for you! The Cash Bitcoin is apparently worth a few hundred dollars in the gray market. One imagines that once Old Bitcoins go ex-Cash, the price of Old Bitcoins should drop by the value of the Cash Bitcoins, but honestly who knows; you should not expect bitcoin markets to operate sensibly.

Another fun question is who gets the Cash Bitcoins: If you own your bitcoins yourself, you participate in the fork, but if you own your bitcoins through an exchange or other centralized intermediary, then the intermediary gets the Cash Bitcoins, and can just go ahead and not give them to you. ("Several exchanges, such as BitMEX, Bitstamp and Coinbase, have said they will not support or allow trading of Bitcoin Cash on their exchanges, which means investors holding bitcoins on these sites will not receive any new tokens.") 

Elsewhere, here is a paper from law firm Baker McKenzie and blockchain company R3 titled "Blockchains and Laws. Are they compatible?" Traditionally when a title consists of a yes-or-no question, the answer is "no," and it's funny to imagine that being the case here. "No, sorry, once we build a blockchain, law will become obsolete, and human society will run entirely on code." I guess their answer is actually yes, though the paper is about boring data-protection-law considerations for bank back-office blockchain consortia, rather than the more exciting securities-law considerations for blockchain initial coin offerings. (I suspect those really aren't compatible with laws.)

Oh and: "Ignition Creative Teams With Jack Abramoff In Reality Series About Bitcoins."

People are worried about non-GAAP accounting.

No actually here is Miles Johnson worrying about GAAP accounting, arguing that companies should instead be evaluated on non-GAAP measures of return on invested capital:

Return on invested capital measures the profit a company is generating on every dollar of capital ever invested in its business, and is far trickier to calculate than simple published accounting earnings. It requires trawling back through hundreds of pages of footnotes in annual reports to adjust back all the capital ever invested into a business. New Constructs, an investment research company, bases its entire practice on the superiority of ROIC over the frequent shenanigans found in GAAP earnings. Not only do accounting rules change over time, rendering crude historical comparisons inaccurate, there are multiple ways large US-listed companies can use adjustments and other tricks to massage their quarterly numbers.

You could imagine a hierarchy of accounting regimes:

  1. An ideal accounting regime that accurately reflects the economic realities of a particular business and allows its users to fully understand how that business works and what its prospects are.
  2. Generally accepted accounting principles, which are sort of arbitrary human creations that fail to capture many important realities, but which are at least consistent across companies.
  3. Just making stuff up so the numbers look good.

Obviously the first one is hard to achieve. If the alternative to GAAP is just making stuff up, then that is bad, but it's not like GAAP is perfect on its own.

People are worried about unicorns.

Ello is not a unicorn, but it is a social network that was briefly dear to my heart, so I enjoyed this postmortem about how it died from a surfeit of early positive attention that it wasn't ready for. I mean, not died exactly -- it still exists! I'm still on it! -- but faded into the obscure corner of the Enchanted Forest where the fizzled startups go. It is quiet there, and a bit sad, but there are those who appreciate its mournful charm more than the gaudy attractions of the successful unicorns. Also I bet Ello has less boardroom drama than Uber.

Things happen.

Wells Fargo Customers Sue Claiming Insurance Scheme Was Scam. Banks May Face Extra $50 Billion Capital Needs After Brexit. Chinese Banks’ Dash for Capital Gets Under Way. Fledgling quant funds seek to disrupt Wall Street. Struggling Stock Pickers Eye Quant Tools to Gain an Edge. Venezuela Bonds Fall After U.S. Announces Sanctions. Europe Becomes Hedge-Fund Hotspot as Economic Recovery Takes Off. No Bubble in Stocks But Look Out When Bonds Pop, Greenspan Says. 1MDB Misses $603 Million Payment to Abu Dhabi Sovereign Wealth Fund. Scarcity of financial intermediation? "Democrats are saying, with increasing clarity, that they want to overthrow a legal paradigm that’s existed for about 40 years and which held that consumer welfare — typically as measured by consumer prices — is the sole relevant metric for making antitrust policy." Ferrari Wants ‘Utility Vehicle’ in a Plan to Double Profit. Millennials are leaving New Jersey. Jellyfish May Be the Snack Food of the Future. 

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.