Hit Stock and Private-Equity Conflicts

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

Hit stock.

In controversies over high-frequency trading in the U.S., people often talk about "front-running," by which they mean something like this:

  1. You want to buy 500 shares of stock.
  2. You see 100 shares available for sale on Bats, 100 on Nasdaq, 100 on the New York Stock Exchange, etc.
  3. You buy the 100 shares on Bats.
  4. With your buying now well underway, you take a little break, have a snack, stretch your legs, and plan to come back and do the rest of your buying in, like, a millisecond or two.
  5. I see your 100-share purchase on Bats, and realize that you are planning to buy 400 more shares elsewhere.
  6. I leap into action while you're relaxing, and buy up those 400 shares on Nasdaq and NYSE and so forth.
  7. Then, when you finally get around to buying them, they're only available from me -- and at a higher price.

I have stylized this description a little bit, but the important thing to notice here is the logical leap in step 5. I saw you buy 100 shares, and correctly intuited that that was just the start of your buying, so I raced ahead of you to buy shares elsewhere and push up the price. But I didn't know your plans to buy the other 400 shares. I just noticed a general pattern, which is that when someone buys 100 shares on Bats, there's a decent chance that there will soon be a buyer for 100 shares on Nasdaq. Sometimes there won't be. But my job is to notice patterns like that and to act on them, as long as they're more profitable than not. Often the hardest part is acting on them: It's nice to notice that buying on one exchange tends to correlate with buying on other exchanges, but it's only profitable if you can get to the other exchanges before everyone else does.

Anyway here's a lovely story about what I guess I will call latency arbitrage on the Shenzhen stock exchange? Only instead of fleeting cross-exchange discrepancies caused by latency differences between exchanges, the pattern here is huge and obvious:

All of this year’s initial share sales in China have been valued at 23 times earnings or less, versus the median ratio of 68 for stocks in the Shenzhen Composite Index, according to data compiled by Bloomberg and China Finance Information Network. And every one of the 99 new listings in Shenzhen has climbed by the daily maximum on its trading debut.

Chinese initial public offerings are systematically underpriced, because their prices are limited by regulation, so they always pop on the first day of trading. If you can buy in the IPO, you can make a guaranteed profit. If you can't -- and allocations are tight -- then the trick is to buy on the Shenzhen exchange before the stock hits its daily limit up. But everyone knows this, so the stocks go limit-up pretty quickly, so you really have to show up at 9:15 a.m. on the dot if you want a chance. The trade is referred to as "hit stock," and it has all the fun and controversy and colocation of a U.S. latency arbitrage trade:

Under the system started in June, the Shenzhen exchange allows brokers to gauge their transmission time (or latency in the industry jargon) and place orders slightly before the opening auction begins, accepting those that arrive at 9:15 a.m. or later. That means a buyer from Beijing, 1,200 miles from Shenzhen, could theoretically beat a rival located right next to the exchange, the person said.

Of course, as with U.S. high-frequency trading, the guaranteed profits here aren't a real guarantee; they're just based on observed patterns. One day a Shenzhen stock will go down instead of up after its IPO, and the arbitrage will vanish. Elsewhere: "Superstitious Chinese investors jittery as black swan sculpture appears in Beijing financial district."


If you run an alternative asset manager, like a hedge fund or a private-equity fund, you are dependent on outside investors to supply your capital (and pay your fees). Many managers have tried to reduce that dependency. The best way is of course to get so rich that you only have to manage your own money, but there are others. You can set up a public permanent-capital vehicle, for instance, which sells stock to the public and never has to return the money. Another popular approach is to own an insurance company. The insurance company gets its money from insurance premiums, a perhaps less fickle source than big endowment managers. And then it has to invest the money somewhere. And there you are. You are good at investing money, and you own the insurance company. It's a natural fit.

The captive-reinsurer approach is a hedge-fund favorite, but here is a story about Athene Holding Ltd., "one of the top fixed-annuity providers in the U.S.," which is partly owned by Apollo Group Management LLC private-equity funds, and which filed for an initial public offering valuing it as high as $7.8 billion. "Apollo helps oversee one-fifth of Athene’s investments, including a slug in Apollo-related funds and businesses -- casino companies, a cruise line and real estate," and Athene "accounts for about one-fourth of the asset-management fees earned by" Apollo. 

That makes the relationship between the two firms close enough to have been questioned by regulators, investors, bankers and lawyers who’ve worked with Athene who asked not to be identified discussing clients.

“This is characteristic of many private equity firms pushing limits, here into incestuous financial relationships, with the general goal of increasing aggregate fees to the firm,” said Lawrence Cunningham, a professor at George Washington University who specializes in corporate governance.

The prospectus makes the case for the value of those conflicts:

We have historically relied on AAM to efficiently reinvest large blocks of invested assets we have acquired. AAM’s investment professionals have developed an intimate knowledge of our liability profile, which is long-dated and predominantly surrender charge protected. This knowledge serves as the foundation of our asset management strategy by enabling us to take advantage of our generally illiquid liability profile and identify asset opportunities with an emphasis on earning incremental yield by taking liquidity risk and complexity risk, rather than assuming solely credit risk. Through AAM and Apollo, we are able to source, value and invest in these high quality assets to drive and target greater investment returns. 

I suppose if you are investing in Athene, you are not just buying a company that is good at selling annuities. You're also buying, indirectly, Apollo's investment management. As is often the case with conflicts of interest, this might just be one you want.

How's Theranos doing?

One weird thing about the unraveling of Theranos Inc., the Blood Unicorn (Elasmotherium haimatos), is how few of its investors have sued. Or even complained. There's a lawsuit from Partner Fund Management LP, filed last month, saying what you'd expect it to say: that Theranos told PFM that its product worked, PFM gave Theranos money, and then it turned out that Theranos's product never worked. That's the sort of lawsuit that you would expect in a dispute about whether a private company was raising hundreds of millions of dollars under false pretenses. But PFM wasn't Theranos's only investor. Do the others just not feel misled? Were they never promised more than what Theranos actually achieved? Did they ignore the glowing news coverage and focus on more sober disclosures from the company itself? Are they just embarrassed? Or what?

Here's a new lawsuit against Theranos that doesn't particularly answer those questions. It's a class action complaint on behalf of Theranos investors, but more particularly on behalf of Robert Colman, who "purchased Theranos securities in September 2013 through a member interest in Lucas Venture Group XI, LLC," and Hilary Taubman-Dye, who "purchased Theranos securities in August 2015 at $19.00 per share as Series B units of a SharePost affiliated fund, Celadon Technology Fund VII, LLC." In other words: The plaintiffs here aren't particularly Theranos investors. They're second-hand investors in Theranos, participants in funds, and all they apparently knew about Theranos is what they read in the press. And so the complaint reads like any public securities fraud lawsuit, basically just quoting news articles. It provides no insight into what I think is the interesting question of Theranos, which is: What were they telling their investors in private?

Here's one part-answer from the Wall Street Journal:

Some Theranos investors have said they were largely left in the dark about how Theranos spent their money, receiving little information from the company for years.

As its troubles deepened this year, Theranos began providing investors with more details, according to some investors.

That's from an article about how "Theranos Inc. received much of its funding from high-profile private investors who weren’t part of the ecosystem that typically backs startups," including Rupert Murdoch and Riley Bechtel. 

Social credit.

Here's a story about how Facebook and Amazon "are beginning to compile digital records of social and financial behavior to rate creditworthiness." Soon they will "draw on bigger, combined data pools, including a person’s internet activity." "Algorithms would use a range of data to calculate" an overall rating, "which would then be used to determine all manner of activities, such as who gets loans, or faster treatment at government offices or access to luxury hotels."

Hahaha no I'm kidding, it's not about Facebook and Amazon, it's about "more than three dozen local governments across China," who are pioneering what will eventually be a thoroughgoing national system of algorithmic social control that will use traditional credit-rating information, online shopping habits, "interactions with other internet users," traffic-law compliance, and whatever other inputs it can get to dole out rewards and punishments. There is a terrifying meat-grinder-esque diagram. 

But didn't you kind of believe that it was about big U.S. internet companies? I mean, U.S. credit card companies are already working on using "interactions with other internet users" as an indication of credit. The data is there. It presumably has some explanatory value. Someone is going to exploit it. Governments can exploit it as a tool of social control, or private companies can exploit it as a tool of profit-making, but I suspect the results in either case will be pretty similar. I suppose it is comforting to hope that a nominally communist country like China is a little ahead of the U.S. in its pursuit of this particular dystopia, but it is hard to be sure.


Here are a paper and related blog post by Xiaoding Liu of the University of Oregon finding that "an increase in the firm’s corruption culture is associated with an increase in the incidence of earnings management, accounting fraud, option backdating, and opportunistic insider trading," which I suppose is straightforward enough. What is surprising is how Liu measures "corruption culture":

To measure corruption attitudes for company insiders in the U.S., I apply the epidemiological approach as described in Fernández (2011), which is based on the key idea that when individuals emigrate from their native country to a new country, their cultural beliefs and values travel with them, but their external economic and institutional environment is left behind. Moreover, immigrants not only bring their beliefs and values to the new country, they also pass down these beliefs to their descendants (e.g., Guiso, Sapienza, and Zingales, 2006). Thus, relevant economic outcomes in the country of ancestry can be used as proxies of culture for immigrants and their descendants.

"Insiders’ country of ancestry is identified using their surnames," and then those surnames are correlated to a corruption index for the relevant countries. So if your name is Somali, the paper suggests, your company is more likely to commit accounting fraud than if you have a Danish name.

Is that something you would trade on? Like, would you look at a company's senior managers, try to guess their ethnicity from their last names, and then use it as a signal of the likelihood that the company is committing fraud? (A weak signal, sure, but remember, "if there were signals that made a lot of sense that were very strong, they would long ago have been traded out.") With this and the social credit stuff, it feels sometimes like we might be getting a little too good at collecting data, and at extracting weak but useful signals from it. Here is Izabella Kaminska:

Funny, in that sense, that so many tech-enthusiasts view AI as the ultimate solution to the economic problem. After all, if AI is left to develop independently it risks becoming a psychopath. On the hand, if we agree human coders must be put in charge of moderating AI’s discriminatory instincts, we may just be recreating a system which amplifies a specific subset of human biases and prejudices on fairly tyrannical grounds.

Some Trump.

How can you figure out what Donald Trump will do as president? Shortly after the election, I noted that his campaign had served as a canvas for the projection of wishes: The working class thought of him as the candidate who'd bring back jobs, the Ku Klux Klan thought of him as the candidate of white nationalism, and you had the odd spectacle of investors watching a candidate who talked about banning Muslims, deporting immigrants, starting trade wars and imprisoning his opponent -- and interpreting all of that as "pro-business."

Based on early cabinet picks and so forth, though, Matthew Yglesias argues that we should "have taken Trump both seriously and literally": He seems to have really meant the stuff about immigration crackdowns, but he also seems to have really meant what he said in his mostly-unread policy statements about financial deregulation. So I guess some good news and some bad news, for business. "The best information about how he will govern is still the literal text of his formal proposals," says Yglesias, though I might argue for the literal text of his tweets instead.

But you can look elsewhere if you really want! For instance, here are Myron Scholes and Ash Alankar:

Market prices do convey important information about changing risks. For example, option prices suggest that Mexican assets are expected to deliver larger gains than losses, implying Trump won’t seek to impose headline-grabbing sanctions on the country. Although less pronounced, options market indicators are similar for China, Japan and emerging markets.

In short, the options market does not appear to view Trump as a protectionist but rather as someone who understands the value and importance of global trade.

Look, I am a pretty efficient-markets kind of guy. If the options market thinks that Trump is planning to increase trade with Mexico, who am I to disagree? Who is Trump to disagree? And yet that ... doesn't sound like what he's said? Perhaps the options market knows his plans better than he does. (That is called "strong-form market efficiency.") Or perhaps the options market is not above projecting some of its own wishes onto Trump.

Elsewhere, Andrew Ross Sorkin wishes that Trump would "voluntarily agree to hire what is known as a 'corporate monitor,' an independent overseer with unfettered access to your organizations who will provide regular reports to the public about any possible instances of conflicts." (He suggests Kenneth Feinberg.) And: "Donald Trump Son-in-Law Could Face His Own Conflict-of-Interest Questions." And: "Trump’s Win Is a Gut Check for Women in the Workplace." And Trump this morning called for "loss of citizenship" as a punishment for exercising First Amendment rights. Perhaps he doesn't mean it literally!

People are worried about unicorns.

I mean, they're obviously worried about Theranos. Elsewhere, Om Malik is worried about Silicon Valley's "distinct lack of empathy for those whose lives are disturbed by its technological wizardry." Lee Fang is worried about ties between Peter Thiel, Donald Trump's Defense Department transition team, and Palantir Technologies, the Spy Unicorn. Stripe Inc.'s founders are now the youngest Irish billionaires. And: "Electrolux, the Swedish appliance maker, is exploring starting an 'Uber for laundry' in which consumers would use their own washing machines to wash other people’s clothes."

People are worried about bond market liquidity.

Is it time to retire this? Will the Era of Trump leave bond market liquidity worries behind? I can project my own hopes here too, you know. Anyway, while you ponder that, here's an article about "Liquidity concerns in a weak market." It's the usual sort of thing, but it's from Malta. Even on an island in the Mediterranean, people are worried about etc.

Things happen.

Time Inc. Is Said to Turn Away From a Takeover by Edgar Bronfman. Samsung Electronics Is Heading for a Split. Testimony Shows Anthem and Cigna at Odds Over Proposed Merger. The Secret, Dangerous World of Venezuelan Bitcoin Mining. Monte Paschi Starts Crucial Bond Swap as Italy Worries Mount. Monte dei Paschi’s Future Hangs on Sunday Vote. In OPEC’s High-Stakes Poker Game, Iran and Iraq Call Saudi Bluff. China prepares clampdown on foreign M&A in war on capital flight. Also a clampdown on Australian real estate. Dodd-Frank Rules on Business: Up for a Rethink. Cooperman Clients May Get Stuck With Part of Omega’s Legal Bill. Bentley Flying Spur or Rolls-Royce Ghost: Which Fantasy Car Is for You? New £5 notes contain animal fat, says Bank of England, drawing anger from vegetarians. This Woman Pickpockets 'Hamilton' Fans Outside Theater. Pangolins at the university. Cat boulders.

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