Market Milking and Research Troubles

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.
Read More.
a | A

Milking markets.

Navinder Singh Sarao is the guy who was arrested in 2015 for spoofing the U.S. equity futures market and perhaps contributing to the flash crash of 2010; he pleaded guilty in November and is awaiting sentencing. One fun fact about Sarao is that, before his arrest, he spent a lot of time complaining to securities regulators, sometimes in all caps, about market manipulation by high-frequency traders. "I don’t like the HFT arena and have complained to the exchange numerous times about their manipulative practices, please BAN IT," he told the U.K. Financial Conduct Authority. I once said that Sarao "seems like a recognizable type," "the sort of day-trader who whines a lot about how markets are rigged against him and how he'd be making so much more money if it weren't for those evil high-frequency traders."

Except that Sarao was making a lot of money. "By the time Sarao was arrested in April 2015, he had about $50 million tied up in investments around the world," in vehicles with subtle names like the "NAV Sarao Milking Markets Fund." And then it ... disappeared:

A review of Sarao’s investments from 2005 to the present day, based on dozens of interviews and thousands of pages of documents, reveals another twist in an already remarkable story. Navinder Sarao, the trading savant accused of sabotaging the world’s financial markets from his bedroom, may himself have been the naïve victim of what his lawyers portray as a series of cons that stripped him of almost every cent he earned. 

The Bloomberg Markets story features among other things a system that "allowed gamblers to bet on movements in currencies and securities using an interface that looked like an online casino, with a roulette wheel and buttons for 'higher' and 'lower' instead of red and black." (It was called "MINDGames, short for Market Influenced Number Determination games.") But I want to highlight three things that are apparently true about Sarao:

  1. He was a jaded cynic, and thought that everyone else in the market was cheating.
  2. He was a naïve mark who seems to have been cheated out of most of his money.
  3. Also he was cheating.

It sounds weird, put like that, but it fits the type. No one's as easy to con as a conspiracy theorist, and no one is as likely to become a con man himself. If you spend all your time congratulating yourself on seeing through the lies to the deep conspiracy that runs the market, you'll just have less energy to devote to making sure that your accountant isn't stealing all your money. If you are on an epic quest to battle evil and expose corruption, you won't be looking out for pickpockets. And if you think that the system is rigged and everyone else is cheating, then it's easy to justify cheating yourself. I feel like there is a lot of this in financial markets, and elsewhere, and it explains a lot. Everyone worries that everyone else is cheating, so they just make it come true.

Research.

There is a well-known conflict of interest in Wall Street investment research, which is that, if a research analyst thinks a company is bad and gives it a Sell rating, then the company will get mad at the analyst's bank and won't hire it to do investment banking business. People mostly think that Wall Street banks are evil, and so this conflict of interest is often described as something nefarious that the banks do: They inflate their ratings, giving Buys to companies that are really Sells, in order to win investment banking business. This, I think, is mostly not true, though it used to be, and though there are other considerations that do lead banks to give lots of companies Buy ratings.

But the conflict obviously exists. It's just that it's probably best thought of, not as a nefarious thing that banks do, but as a sad and awkward thing that is done to banks. The research analysts really do try to be objective and independent and unmoved by outside pressures. And then they sometimes lose clients for the bank. 

So here is a Wall Street Journal story about how JPMorgan Chase & Co.'s research department downgraded Indonesian equities to underweight, which caused the Indonesian government to protest, and ultimately led to JPMorgan losing its primary-dealer status in underwriting Indonesian government debt. "In Indonesia, and other parts of Asia, government officials’ expectations for positive research are unabashed and unapologetic," and the church/state separation of research and investment banking hasn't really caught on:

Western diplomats met Indonesian finance ministry representatives to explain the division between research and investment banking and were told: “Yes, but they are the same. They are J.P. Morgan,” according to a person briefed on the conversation.

If you strain really hard, you can read this as a nefarious story about JPMorgan's conflicts of interest. "Higher-level managers began to sign off on Indonesia research," and the rating was eventually bumped up from underweight to neutral. But JPMorgan still hasn't gotten the business back. And the analysts downgraded Indonesia, not -- obviously! -- to win the country's underwriting business, but because they thought investors should be underweight Indonesian stocks. The system worked as well as you could hope for, if you want research independence.

But it worked rather poorly for JPMorgan! One obvious question here is: Why bother writing equities research on Indonesia? Sell-side research is a tough business, creating a labor-intensive product that is then mostly given away free to clients with an uncertain return in the form of possible future commission business. And even that model is under threat from new regulations. Research is having a hard time justifying its value to banks these days, just on the basic level of: Do banks get enough business from their research reports to justify the analysts' salaries? And then you add in the fact that sometimes the analysts cause the banks to lose important clients. Research creates lots of risk, but it's not so obvious where the profit is.

Miscellaneous beta.

One way to think about the stock market is, like, "is Twitter stock a good long-term investment?" or "do I trust Wells Fargo's management?" or whatever. But the more modern, exciting and fun way to think about the stock market is to forget that company names, and companies, and managements, and businesses even exist, and to just view the stock market as a set of statistical properties. You can invest by picking a range of stocks based on factors and correlations, without ever thinking about what the underlying companies do

Obviously the factor that you'd like to select for here is "stocks that are going to go up a lot in the future," but the analysis is not yet quite that sophisticated. You can do some other things though, which are less useful but also impressive in their ways. For instance, if you want to replicate the performance of the Standard & Poor's 500 Index, that's easy to do. You just buy all the stocks in the index. But you can also get pretty close without doing that. For instance, let's say you want the performance of the S&P 500, but you don't want to own any Goldman Sachs Group Inc. stock. You could just buy the other 499 stocks, in their index weights. Or you could buy those stocks, but re-weight them to give a bit more weight to the stocks that correlate the most with Goldman Sachs. The result would track the index closely, you'd expect, but it wouldn't quite be the index. You'd get the same return as everyone else, but you'd have the quiet satisfaction of knowing that you'd done it in a slightly different way.

That seems to be the idea behind OpenInvest:

Users click through a series of menus to create an “issue profile,” checking boxes to select investment themes—such as gender equality or reduced carbon emissions—as well as groups of companies to exclude. The preset screens lean left. Users can nix weapons manufacturers, tobacco companies, and even those whose executives have backed Donald Trump.

Based on those preferences, OpenInvest creates a basket of more than 60 stocks that both jibes with its customers’ wishes and should, the company says, track the broader market. It balances factors such as size, sector, and each stock’s sensitivity to the market’s ups and downs. OpenInvest says it’s still passive because beating the market isn’t a goal.

Obviously I love that last sentence. "We pick individual stocks, but it's still passive because we're not trying to pick good stocks." But that's not really what they mean. They mean that they are applying modern statistical investing techniques to achieve a very modest goal: not to choose factors that might outperform (momentum, whatever), but to match the index performance without matching the index's list of stocks. I suppose you might want to do that for social-responsibility reasons, but there's also something aesthetically appealing about it. It's too easy to match the index by buying the index; matching the index by buying different stuff adds a pleasingly higher degree of difficulty and a slight Borgesian flavor.

Elsewhere, here is Cliff Asness and AQR on "betting against correlation": "These findings are a strong indication that leverage aversion is indeed an important part of the low-risk effect." And here is a story about (alleged!) closet indexers:

Schroders, Fidelity International, JPMorgan, Henderson and Amundi have been named in a list of 80 investment companies that have potentially sold funds that charge high fees for active management but closely mimic their benchmark.

Gary Cohn.

I enjoyed these two profiles, from the New York Times and the Wall Street Journal, of Gary Cohn, the former president of Goldman Sachs who is now the director of the National Economic Council and a rare cosmopolitan technocrat in the Trump White House. (Disclosure: I too used to work at Goldman.) One thing about cosmopolitan technocrats is that they tend to think in terms of reality. So, for instance, did you know that building things costs money? From the Times:

Mr. Cohn also argued that the bold infrastructure projects that Mr. Trump envisioned would need private-industry partners, those people said, in order to avoid weighing down the government with costs.

That got Mr. Trump’s attention.

The president-elect turned to the other people in the room — his son-in-law, Jared Kushner; his chief strategist, Stephen K. Bannon; his chief of staff, Reince Priebus; and Steven T. Mnuchin, his campaign’s chief fund-raiser and Mr. Trump’s nominee to be Treasury secretary — surprised that his infrastructure ideas had such a potential downside.

“Is this true?” Mr. Trump asked the group, according to those people. Heads nodded. “Why did I have to wait to have this guy tell me?” he demanded.

As Cohn puts it, diplomatically, to the Journal:

“When you’re dealing with the economy, which is my realm, you have to be pragmatic,” Mr. Cohn said in an interview Wednesday. “You have to be realistic to what’s going on in the world and you have to be willing to adapt. I think that’s my job, to advise the president on what is the right solution.”

The implication there is that there are other areas of national government where you don't have to be pragmatic or "realistic to what's going on in the world." I guess I have to ... hope that that's right?

In addition to his pragmatism, Cohn is also an expert in that other well-known speciality of Goldman Sachs, the close analysis of texts:

“In D.C., no one when you meet them says, ‘Nice to meet you,’” Mr. Cohn said. “They say, ‘Good to see you.’ That’s because they pretend like they might have met you before.... Everyone says, ‘Good to see you.’ Well, I know you can see me.”

It's true that Cohn's visibility is well known, though to be fair these people are not just noting his visibility ("I see you"), but also approving it ("Good to see you"). 

Elsewhere, Senators Elizabeth Warren and Tammy Baldwin are wondering if maybe Goldman Sachs had a hand in pushing Trump's financial-regulatory executive orders. They sent a letter to Chief Executive Officer Lloyd Blankfein "calling on him to disclose the extent to which Goldman employees were involved in the drafting of two recent Executive Orders that will directly benefit the company," and "to fully disclose any lobbying related to these executive orders and detail the profits Goldman Sachs expects to gain as these orders take effect." I suppose they are kidding? You can read the executive orders. They do not read like the work of Goldman's crack hermeneutics squad. Also it would be difficult to quantify how much profit Goldman will make from new "principles" of financial regulation like "make regulation efficient, effective, and appropriately tailored." 

Scott Alexander on cost disease.

It's good:

Imagine if tomorrow, the price of water dectupled. Suddenly people have to choose between drinking and washing dishes. Activists argue that taking a shower is a basic human right, and grumpy talk show hosts point out that in their day, parents taught their children not to waste water. A coalition promotes laws ensuring government-subsidized free water for poor families; a Fox News investigative report shows that some people receiving water on the government dime are taking long luxurious showers. Everyone gets really angry and there’s lots of talk about basic compassion and personal responsibility and whatever but all of this is secondary to why does water costs ten times what it used to?

People are worried about unicorns.

Here's how "say on pay" works at U.S. public companies:

  1. The board decides how much to pay management.
  2. It asks the shareholders if that's okay with them.
  3. Approximately 98 percent of the time, the shareholders say yes.
  4. If they say no, the board pays management that amount anyway, because the say-on-pay vote, while required by law, is nonbinding.

Here's how public shareholders of Snap Inc. will vote on real matters for binding shareholder votes, like mergers and board elections:

  1. They won't.

So it is not a huge surprise that "investors buying into Snap Inc.’s much-anticipated initial public offering won’t have any say on how much the company pays its executives." Snap's investors won't have a say on anything. And no public-company investors have much of a say on pay. Say-on-pay votes aren't a practical mechanism for investors to participate in the running of their companies by deciding how much to pay management; they are opportunities for symbolic participation, designed by lawmakers to maybe shame executives a bit. As I said last week, Snap's decision to go public with nonvoting shares is a practical one, designed to prevent proxy fights and hostile takeovers and other real risks of voting stock. But it also gets "rid of a lot of the symbolic baggage of shareholder democracy, the stuff -- advisory say-on-pay votes and shareholder resolutions -- that has little practical effect but that makes shareholders feel loved and included."

Meanwhile, investors are skeptical about Snap anyway:

“The argument here is, ‘We’re going to build this huge audience and monetization will follow,’” said Rett Wallace, chief executive at Triton Research LLC, whose firm collects and analyzes data on companies. He added that before looking at Snap’s prospectus, many investors were hoping for answers about how to make money off Snapchat’s growing user base. Now there is a question about whether Snap can build that huge audience, he said.

The trick in social media seems to be: Always talk about the other thing. If investors are worried about monetization, talk about user growth. If they're worried about user growth, talk about monetization. Of course they're always worried about both, but perhaps you can distract them for a while.  

Elsewhere, here is a story about a Twitter fight between Mark Cuban and Donald Trump that seems to have started with Cuban criticizing Trump for talking about a tech bubble.

People are worried about bond market liquidity.

It is widely believed that all of the problems of bond market liquidity can be solved by creating a universal all-to-all electronic bond trading platform, which is why there are 128 of them:

As of January this year, 128 bond trading platforms were available to fixed income market participants as traders seek new technology to improve connectivity and electronic trading.

This boom in innovation has seen traders readily embrace electronic trading and a variety of alternative protocols offered to meet bond market needs.

However, IOSCO explained this has led to fragmentation and difficulties for those trading the bond market, highlighting the importance of connectivity.

That number seems to keep going up. There is a well-known XKCD on the subject.

Things happen.

How BTG came through the ultimate stress test. Tarullo says Trump’s Wall St principles ‘good starting point.’ Bank Investors Cheer Tarullo Resignation. America’s Biggest Creditors Dump Treasuries in Warning to Trump. Hedge funds gamble on Mt Gox bitcoin payout. Blockchain gang. Steve Young Is an Athlete Who’s Actually Good at Finance. Your potential Trump-Fed candidates. China’s Weibo eclipses rival Twitter’s market cap. Swiss Voters Reject Corporate Tax Overhaul Plan. Samsung Heir Faces New Questions About Role in Political Scandal. Former A.I.G. Executives Reach Settlement in Accounting Fraud Case. "The idea that pizza might not be considered a vegetable was, apparently, un-American." Teach a miner to code. Mostly Prior-Free Asset Allocation. "How about a limiting case for the signaling approach, whereby you invite a famous person, and simply make him or her submit to audience questions, with not even a chance to respond?" "Those with surnames further back in the alphabet did worse in high school, in college, and in the job market early in their careers." Inside the 'world's first' five-star cat hotel. "West Australian dolphins may be using blowfish - or 'blowies' - to get high." Newsletter Writers Are the Most Bullish Since 2004.

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 mlevine51@bloomberg.net

To contact the editor responsible for this story:
James Greiff at jgreiff@bloomberg.net