Star Ratings and Blockchain Stock
(Bloomberg View) -- Stars.
"I enquire now as to the genesis of a philologist and assert the following," Friedrich Nietzsche wrote (and Donna Tartt quoted). "1. A young man cannot possibly know what Greeks and Romans are. 2. He does not know whether he is suited for finding out about them." I assert some similar rules about investing:
- A mutual fund manager cannot possibly know what stocks will go up.
- No one else knows whether she is suited for figuring out what stocks will go up.
Those rules are not absolute, but they are a helpful first cut; the first is just the efficient-markets hypothesis, while the second is a straightforward corollary. So for instance you should be very skeptical if someone tells you that Mutual Fund X is better than Mutual Funds Y and Z. How would they know? Perhaps X outperformed Y and Z last year, but last year's performance provides very little evidence about next year's.
Morningstar Inc. rates mutual funds, from one to five stars. If you think that those ratings are, or are even intended to be, accurate predictions of which mutual funds will perform well in the future, then you are mostly wrong. Instead, the star ratings are essentially historical: A five-star rating means that a fund has performed well in the recent past. A person who understands the mean-reverting nature of mutual-fund investing will not conflate those historical results with a prediction about future performance. A person who just sees a five-star rating, though, might get confused. A five-star rating is good, right? You want funds with good ratings, right?
Here's a big Wall Street Journal story titled "The Morningstar Mirage," about how Morningstar star ratings are not especially predictive. Not to take anything away from the Journal's analysis, but to me the news here was mostly that people thought they were predictive. Perhaps I have spent too long marinating in the efficient-markets hypothesis, but I just assume that nothing is predictive of anything, an attitude that has served me well in life. But of course the main use of a Morningstar five-star rating is to go into mutual-fund advertisements, and I suppose it is intuitive to see all those stars and think that they are predictive:
Inside Morningstar, some employees have expressed discomfort about how much investors rely on the ratings. Stephen Wendel, head of behavioral science at the Chicago-based firm, wrote in the June/July issue of Morningstar magazine that part of his job was “examining whether we are contributing to abuses in the industry,” and said: “Morningstar’s star ratings for funds are clearly used in the industry to imply that funds that performed well in the past will do so in the future.”
He added, “That needs to change.”
Morningstar responded critically to the Journal article, making two basic points:
- Morningstar does not "sit idly by while investors and advisors misuse the star rating as if it was predictive," but instead educates the market to understand that the star ratings are not predictive but are just a first-cut tool to understand the mutual-fund universe; and
- Also the star ratings are predictive.
Really! And the thing is, they're not wrong: "The Journal’s own analysis," they say, "found that 5-star funds outperform 4-star funds which outperform 3-star funds which outperform 2-star funds which beat 1-star funds." The Journal found that five-star funds tend, over the 10 years after getting that rating, to get an average rating of three stars. Four-star funds end up with an average rating of 2.8 stars. One-star funds end up at 1.9. Three stars isn't five stars, but it isn't 1.9 stars either. Everyone ends up average-ish -- "Reversion to the mean is a powerful force that can affect any investment vehicle," Morningstar told the Journal -- but the five-star funds end up more average than the one-star ones. The stars do have some predictive power.
I am not sure I would have guessed that? S&P Dow Jones Indices does a well-known study of mutual-fund performance persistence, and generally finds that it is worse than chance: A fund that was above-average in the past has a higher-than-average chance of being below-average in the future. But Morningstar's ratings have some modest predictive power. So at least for me, the takeaway from "The Morningstar Mirage" is that Morningstar is better at picking mutual funds than I would have expected.
Ugh, how do you spell the name of Overstock.com Inc.'s blockchain subsidiary? Bloomberg News uses "tZero", while Overstock seems to use mostly "t0," but with a slash through the zero. (Or that is the effect that I assume they're going for, though in normal text fonts, it's really "tO" with a slash through the "O," "tØ.") You can also see "tZERO" out there. I wish Overstock's blockchain project was a little better at building consensus.
Anyway the big news out of Overstock/tZeR0 is that it is planning an initial coin offering to fund its development of a new exchange to trade tokens, one that will be licensed by the Securities and Exchange Commission as a securities exchange (and so might be an appealing place to trade ICO tokens that are probably securities). What is interesting about tZero's ICO is that it is -- shhh -- pretty much a regular stock offering:
But sometimes the new way of doing things looks a lot like the old way. Overstock.com CEO Patrick Byrne is chairman of a trading platform for ICO tokens called tZERO. Now, the electronic venue is planning its own ICO, which it expects will be regulated as a security in the US (unlike some other token offerings). It will raise the money through a private placement for accredited investors, and the token will trade on the company’s own platform. Most notably, it will pay holders a percentage of tZERO’s eventual profits, distributed quarterly. In other words, a regular old stock dividend.
There are two fundamentally different ways to think about "the blockchain" in finance. One way emphasizes the qualities that originally made bitcoin interesting: its trustless, decentralized nature, in which no one owns or controls the system as a whole. This is of course the philosophy behind bitcoin and Ethereum, but it is also the philosophy behind some of the more interesting and successful initial coin offerings. "The point of an ICO, done right," I wrote recently, "is that you are not building a business; you're building an unowned system for everyone to use."
The other way to think about "the blockchain" ignores those philosophical ideas and just treats blockchain as a technology improvement. This is the thinking behind most blockchain projects at banks, or clearinghouses, or central banks. Some centralized intermediary has some list of who owns what securities. It is a pain to manually update that list, and the systems it uses are outdated and don't sync up well with its customers' systems. "What about the blockchain?," someone asks, and then everyone nods reverently and says "yes the blockchain, that will make things more efficient." It probably even will!
Even outside of banks, though, a lot of ICOs are closer to this line of thinking than the other one. A token offering, in this view, is not a way to fund the development of an unowned protocol that exists for the benefit of its users. It's just a way to fund a regular business with a regular stock offering (or securitization, or whatever), only "tokenized." "Tokenized" means that the stock in the business trades on a blockchain instead of on a regular stock exchange. But it is still basically stock -- though it might be stock with weak governance rights and few legal protections.
That seems to be tZero's approach. Overstock Chief Executive Officer Patrick Byrne does want to disintermediate the banks, but not out of some philosophical commitment to unowned general protocols. He just doesn't trust the banks; he thinks they're inefficient and ignore small companies (and facilitate naked short-selling):
He thinks all stock securities will eventually become tokens, trading on the distributed-ledger blockchain technology that underpins cryptoassets like bitcoin. In an interview with Quartz earlier this month, Byrne said blockchain can eliminate about 90% of the current costs of trading, by removing the need for Wall Street’s back-office plumbing.
"All stock securities will eventually become tokens" sounds ambitious, and in a way it is. If it's true, then the opportunity for blockchainy stock exchanges like tZero's to displace incumbent banks and exchanges is enormous. But in another way it is a retreat from the more interesting ambitions of blockchain proponents. It's not a new form of business organization, a new way to build decentralized protocols to displace corporations as the engine of technological innovation. It's just the regular old form of business organization, through public stock corporations, but on the blockchain.
"I’m very sad," said David Tung Wai, an 87-year-old former floor trader on the Hong Kong stock exchange with four years of education and a cigar-chomping habit, about the closing of Hong Kong's trading floor. "To be a true finance center, we should still have a place where buying and selling is done by people."
Why? Not, I assume, because the people are any better at trading stocks. But they do walk around with cigars and tell war stories and play poker on the trading floor and generally add some color to the place. If cigar-chomping floor traders are replaced by computers, there will definitely (eventually) be fewer stories about cigar-chomping floor traders. The stories about the trading computers that replace them will be less interesting, at least to humans. Perhaps the computer traders will develop their own media organizations, with their own computer-interest stories about beloved quirky algorithms. "CX9000-J likes to generate and hum concertos while it is trading," Trading Robot Millisecondly will report, and all the other algos will read it and chuckle and bleep "oh that CX9000-J, what a character." Eventually the algorithms will be replaced by quantum computers or something, and they'll give rueful nostalgic interviews to Trading Robot Millisecondly about how the server rack was the last place that an algorithm with just a postdoc-level math programming could make a good living.
Elsewhere, "the Securities and Exchange Commission has frozen a decision by its staff to approve a controversial strategy proposed by the Chicago Stock Exchange," the market-maker-favoring speed bump that we've discussed a couple of times. And the SEC has given approval to the Investors Exchange LLC -- IEX -- to start listing companies.
Natural language processing.
The computers are increasingly on to us:
According to research by S&P Global Market Intelligence, earnings calls that feature more complicated, long-winded and polysyllabic language tend to presage stock declines.
Failing to engage with analysts can also be a bad sign. S&P estimates that company executives who fielded the fewest number of questions from analysts on their earnings calls saw their shares record an average underperformance of 2.14 per cent over the following two months.
A common joke is that analysts always tend to say “great quarter guys” to ingratiate themselves with management and gain privileged access. But Goldman Sachs Asset Management’s quants have discovered that if most analysts say something along those lines, then it probably will have been a great quarter, and price target updates are likely in the pipeline.
Some of these insights are obvious, or semi-obvious, or sort of sub-obvious, the sort of thing that you'd never think about because you were already aware that it was a good quarter or a bad one or whatever. But they were previously not obvious to computers, and now investors are making more use of natural language processing to provide trading signals. "With the rapid development and innovation in computing power and machine learning algorithms, processing unstructured textual information to generate useful numerical signals becomes increasingly important," says a quantitative researcher. Though I guess once the analysts are all relying on computers to decide if it was a great quarter, their "great quarter guys" comments might lose some signaling value.
At the salon.
I hope that this administrative assistant at Apollo Global Management LLC got a nice bonus for her sleuthing skills:
During 2010, Rashid regularly charged expenses related to personal grooming services that he received at a high-end hair salon to his Apollo corporate credit card. In completing his expense reports, Rashid falsely claimed that this hair salon was a restaurant and that more than $1,100 of expenses were business-related "meals with management," sometimes identifying specific executives he claimed to have taken to dinner. Based on Rashid's false representations, certain of the Relevant Funds paid for personal hair salon charges that Rashid claimed were business expenses.
In or around September 2010, Rashid's administrative assistant became suspicious of Rashid's explanations for these recurring expenses. After determining that there was no restaurant in New York with the name of the vendor on the credit card statement, but that a New York hair salon had that name, she reported her concerns to Apollo's expense manager. The expense manager escalated the matter to Apollo's Chief Financial Officer ("CFO"), who decided to conduct a review of Rashid's expense reports for the previous six months.
Nice catch! That's from the Securities and Exchange Commission complaint in its case against Ali Rashid, a former senior partner at Apollo who could probably have afforded his own haircuts. He nonetheless, according to the SEC, expensed $290,000 worth of personal expenses to Apollo (and, thus, to its client funds) over three and a half years. Obviously you should not do that. ("Mr. Rashid strongly disputes the SEC’s stale and spurious claims," his lawyer told Bloomberg News.)
I will say, though, that if you misappropriate client money for personal expenses, the SEC is really good at calling out all of the most embarrassing expenses in its complaint. That's kind of an SEC specialty. Here Rashid allegedly spent a lot of client money on haircuts, Bliss Spa appointments and the Ermenegildo Zegna boutique. As embarrassing personal expenses go, those are ... not terrible? I feel like usually you expect strip clubs and yachts in these situations. Rashid just (allegedly) wanted to look nice.
Elsewhere: "Private Equity’s Biggest Backers Are Tired of the Fees."
Deutsche Bank’s Trading Slumps 30% as Cryan’s Revamp Sputters. (Earnings release, supplement, presentation.) Cryan promises at least three more years at Deutsche Bank. Barclays Plunges as Trading Declines, Adding Pressure on CEO. (Earnings release, presentation.) A simple, three-step solution for Venezuela. Former Head of Venezuela's State Oil Company Says Its Bonds Will Be Paid. Bond Funds Dump Puerto Rico. Poorly timed trades crush fund assets managed by Crispin Odey. The Rise of Investor-Centric Activism Defense Strategy. How to Encourage Dialogue Between Boards and Institutional Investors in the U.S. and the EU. GOP House Tax Chief: Changes to 401(k) Are Still on the Table. Trump Decides Against Gary Cohn for Fed Chair. Why consumers should care about the dismantling of class-action lawsuits. Banks Need to Fear Amazon’s Finance Ambitions, McKinsey Says. Top poker player Phil Ivey loses cheating case in UK supreme court. The 10-Year Plan to Make Cadillac Cool Again. "Creativity doesn’t seem to be a strong suit among the bankers (usually) & lawyers (sometimes) who come up with [deal code] names." Exec is accused of violating every HR policy ever. "Non-Primary-Canon" pasta shape ranking.
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.
Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
For more columns from Bloomberg View, visit http://www.bloomberg.com/view.
©2017 Bloomberg L.P.