Money Stuff

Dow Records and Custody Deals

Also speed bumps, endowments, cartoons, Snapchat and unicornophagy.

The Dow.

The Dow Jones Industrial Average closed at 20,068.51 yesterday, its first close over 20,000, and so as custom demands there were parties and hats and approximately 20,068.51 articles about it. The articles fall into some recognizable categories. There is straightforward glee, or relief: Stocks are up! There is contrarian worry: If stocks are up, they might go down! There are articles explaining how the Dow is calculated, and other articles explaining how dumb it is that the Dow is calculated that way. There are personal-finance advice columns, and comparisons between the state of the Dow and the state of the economy, and worries about the distribution of stock wealth, and reminiscences of the 1980s. As with everything in financial news, there are articles blaming Goldman Sachs. There are -- and this was new to me, frankly -- a certain number of articles about how, while 1,000 is 100 percent of 1,000, it is only 5.3 percent of 19,000. (This one has a chart.) That is the sort of complex mathematical technicality that you have to master, if you are going to follow the Dow like a pro.

My favorite kind of Dow article, though, is probably the Dow Truther article, in which the reader learns that the Dow didn't really hit 20,000 for the first time yesterday. Last time the Dow flirted with 20,000, a few weeks ago, we talked about the Dow "theoretical high," which is found by adding the intraday highs of all of the Dow components; that crossed 20,000 all the way back in December. (I proposed my own pompously-named absurd Dow calculation, the Quantum Metaphysical Dow, which is calculated by multiplying the regular Dow by two; coincidentally, it crossed 40,000 yesterday.)

But that has nothing on this James Mackintosh column pointing out that the Dow is already at 30,000 because of, like, a century-old typo:

The official Dow record shows a drop of 24%—its worst-ever day—when the market reopened in 1914 after a four-month break because of the start of World War I. In fact, the market and the Dow rose that day, but the record was recalculated without any adjustment when the measure expanded from 12 to 20 stocks two years later. Because some of the new stocks added had lower prices, the new version of the average was pulled down.

If you correct for that and other early-days mistakes, you get a Dow value of over 30,000 as of yesterday morning. It is hard to know what to do with that information. Are you 50 percent richer than you thought you were? (30,000 is 50 percent more than 20,000, as Dow experts know.) Not ... not really? Unless you bought the Dow in 1913 and held it until now, which you didn't, for a number of reasons including the fact that American Smelting & Refining Company, Central Leather Company and Amalgamated Copper Mining Company are not Dow components, or companies, or even plausible company names any more. 

But you can learn something from this sort of Dow Trutherism. The Dow, as you are no doubt tediously aware by now, is dumb. It's a price-weighted index of 30 not especially representative stocks. It's a silly and outdated way to do an index, and if it didn't exist, no one would invent it today. But there's a reason it's hard to get rid of. The fact is that the Dow has the longest track record of any U.S. stock index, and for many purposes just having a lot of data points is important. If you want to do, like, Fibonacci analysis or whatever on 100 years of data, guess what, the S&P only goes back to the 1920s. If you want to do very long-term technical analysis, the Dow is the index for you. Except if that data is ... wrong? 


How was your day yesterday? The market was up; did you do any big trades? Bring in any business? Are you feeling good about yourself? Here, let me ruin that for you. Someone at JPMorgan Chase & Co. did a trillion-dollar deal yesterday:

J.P. Morgan Chase & Co. struck a deal to be the custodian for more than $1 trillion of BlackRock Inc.’s assets, poaching the business from State Street Corp., the companies confirmed Wednesday.

Don't you kind of wish you worked in custody services? "The bank is expected to make tens of millions of dollars in annual fees from the deal," which is nice but not record-setting or anything. My guess is that whoever did this deal will not end up being JPMorgan's biggest revenue generator this year. But he'll probably be the only person to bring in a trillion-dollar account. A trillion dollars! With a "t"! And then an "r," and an "illion"! And "the bank said the deal was one of the largest shifts of custody assets ever," emphasis added, because it suggests that there have been ... other, even bigger deals? Imagine going to your boss and telling her that you brought in a trillion-dollar account. Take the rest of the day off; you've earned it.

Try not to imagine being the State Street banker who lost the account, though. ("State Street Chief Executive Officer Jay Hooley said Wednesday that BlackRock approached the Boston-based custody bank as it sought to diversify its custodians. Hooley said that as BlackRock grew into a $5.15 trillion money manager it decided it needed another partner.") Try not to imagine waking up the next day and going out to pitch, what, a $10 billion account? Get 100 of those and you can almost fill the hole in your heart, and your P&L, that BlackRock left.

Market structure.

When IEX Group Inc. filed to become a public stock exchange, two controversial aspects of its application were the 350-microsecond "speed bump" on messages to and from its exchange, and its use of dark (non-displayed) "discretionary peg" orders that take advantage of the speed bump to avoid being picked off by fast traders. NYSE Group Inc., which runs the New York Stock Exchange, opposed both of those things. The discretionary peg, it told the Securities and Exchange Commission, would lead to a dark order book that "would not be performing one of the central functions of a registered exchange, which is to foster the price discovery process through display of orders." And the speed bump, NYSE predicted, would inspire dangerous copycats:

If IEX is permitted to include this programmed delay without including it in its rules, other exchanges could similarly add an intentional delay without changing their rules, without the benefit of any input from the Commission regarding what amount of systematic delay would be acceptable. NYSE believes this would be a step backwards in the development of the national market system

Or as I put it:

The big problem in the IEX application is that IEX are nice, so a lot of people want the SEC to let them do some new weird market-structure things. But the SEC is a general regulator, and if it lets IEX do those things, it sort of has to let everyone. And not everyone is so nice.

Well, NYSE was not wrong about the copycats! The latest exchange to announce a speed bump is ... oh ... hmm:

The owner of the New York Stock Exchange will rename its tiny NYSE MKT exchange the NYSE American and introduce a 350-microsecond delay on orders there, according to a statement Wednesday. NYSE American will also add a new order type called a discretionary peg -- which IEX pioneered -- a so-called dark order because it isn’t publicly displayed. Both new features closely mirror how IEX’s Investors Exchange works.

You know this is a cynical move because they renamed the exchange "NYSE American," like those Budweiser cans last summer.

I don't know? IEX has around a 2 percent market share in U.S. equities. Some of that comes from people who like discretionary pegs and speed bumps; some of it comes, perhaps, from people who have read "Flash Boys" and just think IEX are nice trustworthy people. NYSE American will presumably compete for that first category. "Some find that feature helpful," shrugged NYSE Group's chief operating officer. It just doesn't seem like all that big a category?

By the way, if you like market-structure controversies, you might enjoy this hilariously ill-tempered letter that NYSE sent to the SEC last week. The basic story is that NYSE filed to change some of its co-location services and fees, and some people (e.g. high-frequency trading firms) objected, feeling that they were getting rather gouged for data and access fees. And NYSE responded, in essence, that nobody's forcing all those HFT firms to pay for co-location:

Some of the comment letters contend that market participants effectively are required to co-locate with, or to subscribe to proprietary market data products directly from, an exchange. For example, the Wolverine Letter starts by stating that Wolverine Trading LLC is a proprietary trading firm and registered market maker, Wolverine Execution Services LLC is a registered broker dealer, and Wolverine Trading Technologies LLC is a technology provider, and continues by saying that “[a]s such, Wolverine is required to subscribe to the lowest latency NYSE market data products and services.” The Wolverine Letter then treats all its costs–including the optional cage surrounding its cabinets, power, cross connects, network ports and connectivity—as costs related to market access. However self-servingly it tries to characterize them, these listed costs, like rent and employee compensation and benefits, are simply costs associated with Wolverine’s business activities. These business activities and Wolverine’s business judgment—not the Exchange—determine the most effective way for Wolverine to select the products and services it uses.

That is ... sort of true? Wolverine, and other high-frequency trading firms, have a business model that makes them dependent on getting the fastest possible data from NYSE and other exchange operators. The value in getting the fastest feed is purely positional: Making the feed 10 microseconds faster for everyone doesn't matter much, but making the feed 10 microseconds faster for everyone but you destroys your business. That gives NYSE enormous leverage to just charge them whatever it wants (subject to SEC review). NYSE is correct that, if they don't want to pay, these firms can always just stop doing high-frequency trading. It is not an entirely satisfying response. But NYSE is shedding no tears for the HFTs:

These firms have chosen to build business models based on speed. The Exchange finds it interesting that such firms and SIFMA object to exchange fees, which are subject to Commission review and the requirements of the Act, but by and large do not disclose how much profit they or their members make from being co-located and using exchange market data products

Harvard layoffs.

"Harvard University’s endowment plans to outsource management of most of its assets and lay off roughly half the staff, in a radical overhaul of the way the world’s wealthiest school invests its money." Some of the departing teams will leave to start their own firms, with which Harvard will likely invest. And what do you make of this?

Some alumni and faculty have criticized Harvard for paying its traders too much for returns that have lagged Ivy League peers’. At the same time, others have questioned Harvard’s ability to attract top talent with pay that is less than what hedge-fund firms can afford.

You could build a simple model in which:

  1. Harvard can't pay in-house staff as much as hedge funds would pay them, because people would complain.
  2. Therefore it can't attract the best in-house staff.
  3. If it spins the in-house staff off into new independent firms and invests with them, it can pay them the market rate.
  4. Therefore they'll get better.

I don't know how well number 4 works. But for some of the departing Harvard employees, this might end up with Harvard paying them more.

Don't do insider trading.

Swedish telecommunications company Telia Co. AB decided, for some reason, to produce an employee code of conduct in the form of a series of adorable animated cartoons. (Ha, the reason is that "Telia had been in trouble with ethical compliance in the past," and is "facing $1.4 billion in corruption-related fines and is under scrutiny for blocking internet service to customers in Turkey at the request of the Turkish government," so the cutesy cartoons have a dark past.)

Here is the web page, which is admirably blunt. "Don't do this at work," it says at the top. "Don't launder money." "Don't do insider trading." If you click on any of the cartoons, you get an enlarged cartoon, though sadly the cartoony meme-ish headline language gives way to corporatese in the text. "Don't do bribes," says a cartoon of a shifty-looking green man catapulting bags of money to a faraway island. But then: "We act with the highest standards of transparency and integrity wherever we operate. By doing so we can ensure our actions and decisions are always in the best interests of customers, our business and society." Ah yes, the highest standards of transparency and integrity. What were you even thinking, little green man with a catapult?

Anyway it is a good web page as web pages, or corporate codes of conduct, go. Obviously I am particularly fond of "Don't do insider trading," though what I really want is a little cartoon man to tell people the Second Law of Insider Trading, which is: "If you have inside information about an upcoming merger, don't buy short-dated out-of-the-money call options on the target." That's the sort of thing that a cartoon could really drive home.  


We talked the other day about how social-media startups are selling essentially the same thing -- user engagement, wealth and coolness -- to both advertisers and investors, and how it is in some sense weird to start with the investors. The investors shouldn't really care about user coolness; they should care about advertiser money, and you might think that the best measure of a social network's ability to get advertiser money would be actually getting advertiser money. I realize this is hopelessly naive. Obviously social networks have to go public before they sell too many ads. People hate ads, and you don't want to drive them away before you lock in your initial public offering. Once you have the investors, and get all the people in the world on your network, then you can start spamming them with ads, because where else will they go?

It's a theory, but Snap Inc. seems to be parallel-tracking, since while working on its IPO it is also "in talks with the media-buying arms of several big advertising companies, including WPP PLC, Omnicom Group Inc., Publicis Groupe SA, and Interpublic Group of Cos., and is seeking ad-spending commitments of $100 million to $200 million for 2017 from each firm, according to people familiar with the discussions."

Elsewhere in Snap news, Facebook Inc., which previously copied Snapchat Stories in the form of Instagram Stories, is now copying Snapchat Stories in the form of Facebook Stories. When Snapchat does start spamming users with ads, where ever will they go?

People are worried about unicorns.

One worry that I have about unicorns is: What is the appropriate metaphor for a company that buys unicorns? If you are a startup, being bought is good! (Unless it's a down round, in which case it's bad.) But if you are a unicorn -- a mythical horned horse, that is, not a startup -- being merged into a larger entity seems uncomfortably close to unicornophagy. (Elasmotheriophagy? Try not to think too hard about this. "Being eaten," is the point.) "Roughly 24 hours before AppDynamics was set to sell its first batch of public stock at a valuation of $2 billion or so," writes my Bloomberg Gadfly colleague Shira Ovide, "bigfoot Cisco Systems Inc. swooped in with a buzzer-beater purchase offer that snatched it from the IPO market." She calls Cisco a "unicorn enabler" for this swoop, which will give investors in other tech startups some confidence in their valuations, but there are darker metaphorical interpretations. "Unicornivore," is Lex's term. Cisco may have validated AppDynamics's deliciousness, but it also ate it.

Things happen.

Jesse Litvak's jury "may be near a verdict, despite being deadlocked on two of the 10 securities fraud counts the defendant faced." Johnson & Johnson to Acquire Actelion for $30 Billion. BlackRock and Vanguard crush smaller ETF rivals. RBS to Take $3.8 Billion Charge Tied to U.S. Mortgage Probe. Dutch Regulator Accidentally Posts Soros’s Short Positions. My Bloomberg colleague Li Keqiang. Vodacom Said to Weigh $1.1 Billion Stake Sale to Black Investors. What does a "21st Century Glass-Steagall" mean? Oil Supermajors' Debt From the Crude Collapse May Have Peaked. Shkreli Wins Access to Law Firm Records to Defend Himself. Goldman’s $285 Million Package for Gary Cohn Is Questioned. Trump Hotels, Amid Calls to Divest, Instead Plans U.S. Expansion. "The US has been downgraded from a 'full democracy' to a 'flawed democracy.'" Swiss Watch Exports Have Worst Year Since Financial Crisis. RIP Harry Mathews. Wedding ring in sewer. Spider crowdsourcing.

Programming note: Money Stuff will be off tomorrow, back on Monday. See you then! 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.