ADVERTISEMENT

Merrill Lynch’s Secret Stock Deals

From March 2008 through March 2013, Bank of America Merrill Lynch had a secret way to get you the stock you wanted.

Merrill Lynch’s Secret Stock Deals
The silhouette of a man talking on a mobile phone. (Photographer: Aaron M. Sprecher/Bloomberg)  

(Bloomberg View) -- Masking.

If you want to buy stock and you are a customer of Bank of America Merrill Lynch, you give BAML an order and it gives you back the stock. You might not care where it gets the stock, but you might; if you are a serious enough stock trader you probably do. There are a bunch of places where it could get the stock, but there are two big obvious categories. One place is: the stock exchanges. BAML could take your order to buy 100 shares of XYZ, go out to the New York Stock Exchange or Nasdaq or wherever, buy 100 shares, and deliver them to you. In the simplest form, if a stock is quoted on the major stock exchanges with a highest bid of $10.00 and a lowest offer of $10.01, and you send BAML an order to the effect of "just buy me 100 shares," BAML will go buy 100 shares at the offer -- $10.01 -- and deliver them to you for $1,001, plus a commission.

Another place is: its own inventory. BAML is a "broker-dealer," not just a broker; in addition to going out and finding stock for you, it keeps some stock lying around, and can use that stock to fill your order. BAML could take your order to buy 100 shares of XYZ, look into its warehouse, notice that it already has 100 shares of XYZ, and sell them to you. The price will again be determined by the public market price; if the market is $10.00 at $10.01, BAML might sell you the 100 shares for $1,001, plus a commission.

Should you care about which of these approaches BAML follows? I mean, people do, endlessly, for reasons that can be a little tedious to hear about. If BAML gets your stock from the stock exchanges, you might care a lot about which stock exchanges it goes to and in which order, because buying stock on one stock exchange may send a signal to high-frequency traders that you might be buying more, and may push up the price of the stock. So you want BAML to route to stock exchanges in a careful way so as not to push up the price. If on the other hand BAML sells you the stock from its own inventory then you don't have to worry about exposing your order, though you might worry about other things. (For instance: If BAML sometimes sells you the stock from its own inventory, and sometimes goes to the stock exchange, is it making those decisions based on what's good for you or what's good for it?)

In any case, from March 2008 through March 2013, BAML had a secret third way to get you stock:

Beginning in March 2008, BofAML entered into agreements with "electronic liquidity providers" ("ELPs") to execute a portion of BofAML's institutional client orders. Those ELPs, which changed over time, included Citadel Securities, D.E. Shaw, Madoff Securities, Knight Capital, Getco (which later merged with Knight Capital to become KCG, now owned by Virtu Financial), Two Sigma Securities, Sun Trading, and ATD (then a division of Citigroup, now owned by Citadel Securities). The agreements, initially negotiated by BofAML's then-head of Global Equity Trading and then-head of electronic trading, provided that BofAML would route certain "direct strategy access" or "DSA" orders to the ELPs before routing the orders to public stock exchanges. Direct strategy access orders are orders that are submitted by institutional clients to BofAML's electronic trading algorithms for execution. Pursuant to the agreements, the ELPs could choose whether to fill the DSA orders themselves. In return, the ELPs did not charge BofAML for the executions. The prices at which those trades executed with the ELPs were, in industry parlance, at the "far side" of the spread between the bid and the ask.

BofAML entered into agreements with the ELPs because it wanted to increase the volume of trades it sent to the ELPs, which BofAML sometimes described in internal communications as "trading partners." In discussing one ELP agreement in an internal BofAML email, the then-head of electronic trading explained that BofAML needed to increase the volume of trades it sent to that ELP "based on other revenue opportunities currently being discussed with them."

That is, instead of going to the stock exchange, or selling you the stock out of its own inventory, BAML would let Citadel or D.E. Shaw or Two Sigma or whoever sell you the stock out of their inventory. (But only if they wanted to: If they passed, the order would go to the stock exchange.) Again, if the stock was $10.00 at $10.01, Citadel or whoever would sell you the 100 shares at the offer (i.e. for $1,001), and BAML would charge you a commission. Your execution would look pretty much the same as it would if you went to the exchange, or if BAML filled your order directly.

The only thing that would be different is the little code on the electronic report -- called a FIX Protocol message -- that you get back after you trade. If BAML went to the exchange the report would say what exchange it went to. If BAML sold you the stock itself the report would say "MLCO" (the code for Merrill Lynch). If BAML went to Citadel or D.E. Shaw or whomever, the report would say CDRG or SHAW or whatever their code was.

Unless BAML did this:

In March 2008, after entering into the agreements with the ELPs, described above, BofAML's then-head of Global Equity Trading and then-head of electronic trading directed BofAML employees to alter FIX messages for BofAML client trades executed by ELPs, by replacing the codes which identified ELPs with a new code referencing BofAML. Pursuant to this direction, BofAML employees re-programmed BofAML's internal trading systems to automatically remove the original codes which accurately reflected the ELP venues in Tag 76 (e.g., "CDRG" for Citadel," "SHAW" for D.E. Shaw, and "MADF" for Madoff) and insert a new identifying code ("MLCO") which inaccurately reflected the trading venue as BofAML.

That ... would be ... bad?

Those quotes above are from BAML's settlement agreement with New York State Attorney General Eric Schneiderman, who fined BAML $42 million for deceiving clients about how their orders were filled. "During that period, BofAML altered the information in FIX messages on more than sixteen million DSA executions, accounting for over four billion shares."

How bad is this? You can sort of imagine a justification. You sent the order to BAML, to either fill out of inventory or take to the stock exchange. BAML essentially outsourced its inventory-management function: Sometimes it would fill your order out of inventory, except it wouldn't be its inventory, it would be Citadel's or D.E. Shaw's or whoever's inventory. Either way you got filled quickly at the offer. "At all times we met our obligation to deliver the best prices to clients," says a Bank of America spokesman. So what is the problem?

Well, for one thing, the behavior is terrible; you don't want to find yourself altering documents to deceive your clients about how you came by their stock. For another thing, customers do seem to care a lot about who sees their orders and when. If you send an order to BAML, you expect BAML to either fill the order or pass it on to the stock exchanges; if it passes it on, you assume some benign reason for that: It doesn't have any inventory or is at its risk limits or can get a better price on the exchange or whatever.

But if some high-frequency trader gets your order and passes on it, then you might be a little more annoyed. High-frequency traders, in the public perception, make their living by looking at investors' orders before anyone else does, and then using that information for their own profit. (What if they are ... front-running your order?) Institutional investors put a lot of effort into algorithms and dark pools and other mechanisms precisely to keep high-frequency traders from seeing their orders. And if BAML was secretly showing their orders to those high-frequency traders, you can see how the investors would be mad.

Seating charts.

We talk from time to time around here (and elsewhere) about the symbolic and practical importance of seating charts, but I feel like I have never mentioned a pretty basic rule of thumb of corporate desk location, which is: People who work for you should have desks at your company, and people who work elsewhere should have desks somewhere else. This is not a hard-and-fast rule, in these days of telecommuting and co-working, but it does seem like a useful symbolic guide. If every day your office fills up with another company's employees, or if your employees go to work somewhere else every day, then there is something a bit odd about your business.

Here is a story about a very odd business indeed, bank supervision, where there are two schools of thought about desk location. One school of thought, which dominated both the Federal Reserve Bank of New York and the Office of the Comptroller of the Currency until recently, put bank examiners' desks inside the banks they examined, on the sensible theory that they'd have an easier time examining the banks if they sat where they could see them. But now the New York Fed, at least, is relocating its examiners to its headquarters, on the also sensible theory that, if you go to work at a bank every day, you might feel a little too much like you work for that bank: 

If a bank examiner sees “the same thing day in and day out, you get too trusting that that routine is OK. It’s easy to get caught in that,” said Douglas Roeder, a managing director at PWC, a consultancy, and a former OCC senior deputy comptroller in charge of overseeing large banks.

A New York Fed spokeswoman said bringing supervisors back to headquarters allows them to “more readily interact with colleagues from other teams and compare and contrast firm practices, processes and risks.” She added that the regulator will retain some workspace inside banks, where examiners can “continue to have regular and consistent interactions” with bankers.

The popular phrase here is "regulatory capture," but it doesn't have to be nefarious. If you are around Bank X bankers every day, and if they seem nice and well-intentioned and serious about their jobs and also they invite you to join their softball team, then you are just going to get used to thinking of them as the good guys. And if you see them doing something, openly and casually and repeatedly and in front of you, you will naturally assume that that is a thing that banks are supposed to do. Even if no other bank does it; even if your fellow bank examiners would be scandalized by it.

On the other hand just going back to headquarters is not a cure-all either. If you supervise a bank remotely, and only show up at the bank for guided tours, then you are going to miss a lot of stuff that the bankers would do openly and casually and repeatedly and in front of you if you were there every day. There is a real trade-off; what you want is almost an undercover-cop approach where the supervisor hangs out at Bank X all day and then goes back to the station at night to debrief with her real colleagues.

Anyway the OCC was going to move its examiners back to headquarters but now it apparently won't, and while I can certainly accept that there are some plausible justifications for that choice, this isn't one of them:

Mr. Otting, the OCC’s new leader, is sympathetic to a different school of thought, which views the location of an examiner’s desk as a minor concern compared to other factors, such as training, priorities and resources.

Honestly! What could be more important than desk location? 

Elsewhere in New York Fed desk-location news, San Francisco Fed President John Williams might be moving his desk across the country, because he "is the leading candidate to become the next president of the Federal Reserve Bank of New York." But "the Federal Reserve is facing a backlash over its lack of diversity in key positions" because of the news that Williams is the front-runner, and here at Bloomberg View, Senator Cory Booker argues that "The New York Fed Needs a New Perspective."

Toys.

Toys "R" Us is planning to liquidate after failing to restructure in bankruptcy, weighed down by too much debt that it took on in a private-equity leveraged buyout in 2005. But this guy has a plan:

The head of the company that created Bratz dolls and Little Tikes is putting $100 million toward a long-shot bid to save Toys “R” Us.

MGA Entertainment Inc. Chief Executive Officer Isaac Larian is aiming to buy the toy-store company’s assets as part of an investment group that includes a crowd-funding campaign. He said it’s his own money on the line, and MGA isn’t part of the bid. If he’s successful, the executive expects that 200 to 400 U.S. stores can be saved.

It is a literal GoFundMe campaign. Here is the website. At Barron's, Mary Childs explains the rewards:

If you give up to $999, you get MGA's Little Tikes® Cozy Coupe®, and up to $4,999, a T-shirt saying “I’ll ALWAYS be a Toys ‘R’ Us Kid” and a "Little Tikes® Build-a-House." Invitations to the "reopening block party" start at $10,000, and $25,000 gets you a tour of the Ohio Little Tikes® factory—"the oldest toy factory currently operating in the U.S.—including travel and accommodations." ...

The #SaveToysRUs GoFundMe campaign states clearly for potential funders that this is neither a tax-deductible charitable donation nor an actual investment where you retain economic interest: You do not get any equity or stake in any acquisition. You are simply overpaying for an inflatable bouncy free-throw-lane-with-wrestling-ring-sides Little Tikes® Super Slam ‘n Dunk, a product sold by Larian's MGA, to help revive a childhood institution. 

Ah.

There seems to be a widespread consensus that financial engineering killed Toys "R" Us: A private-equity buyout fueled by easy access to junk bonds loaded the company down with too much debt, which made it impossible for it to invest in improving its business. But Larian's core idea is that a different kind of financial engineering could save Toys "R" Us: He can do a new buyout funded by new investors, but instead of giving them a high junk-bond coupon, or even equity, he will give them factory tours and T-shirts.

I ... look, it's just a website. It doesn't seem particularly serious. As of 9:30 a.m. today it had $200,047,609 of pledges, but that number is misleading; the first $200 million comes from Larian and his investors, who will all presumably be actual investors with an economic interest in the bid, while the last $47,609 comes from GoFundMe contributors who will get stickers or parties or whatever. 

Still it has tapped into a certain cultural moment. GoFundMe is a thing. Kickstarter is a thing. Initial coin offerings -- where you invest in an enterprise owned by others in exchange, not for ownership or repayment, but for "tokens" with no contractual rights -- are so intensely a thing. Novelty-flamethrower-based funding is, more or less, a thing.

The private-equity buyout of Toys "R" Us is a late consequence of an intellectual revolution that began in the 1970s and 1980s, when Michael Milken pioneered high-yield bonds and leveraged buyouts by understanding -- and by convincing investors of -- the counterintuitive fact that lending money to companies that might not be able to pay it back could be even more lucrative than lending money to companies that could definitely pay it back. 

But if there's a GoFundMe rescue of Toys "R" Us -- and, to be fair, there obviously won't be, come on -- then it will be part of the early stages of a new intellectual revolution that is happening all around us. This one is considerably more counterintuitive than the junk-bond one; I, for instance, cannot understand it. The point seems to be, though, that giving money to people in exchange for no investor rights -- no equity, no promise of repayment, just a badge indicating that you are a participant in some nebulous community -- is even better than giving money to people in exchange for some rights. Sometimes "better" seems to mean "more lucrative" (certainly the people who invest in initial coin offerings for tokens that will give them no rights to products that have not been built seem to expect those tokens to be wildly valuable), and sometimes it doesn't (what do the Toys "R" Us people expect?). In each case there's some expectation that what you get back, while it has no ownership rights, has some connection to the business; in an ICO you get back tokens that can be used on whatever platform the business is building, while in the Toys "R" Us GoFundMe you get, you know, toys. So there is an element of product pre-selling, but there is also often an element of speculation, and an element of emotional community-building, and an element of piling into viral internet enthusiasms, and probably some other elements that I do not understand.

If you were trying to understand it, you might start from the same place that I start from in trying to understand tech unicorns. There is a lot of money sloshing around, as developing countries get richer and as the world gets more connected. The economy is increasingly global, and profits are increasingly driven by scalable intellectual capital rather than by local physical capital, so good ideas are increasingly valuable. This gives entrepreneurs relatively more power, and capital sources relatively less power, than they've had in the past. And so you'd expect the nature of capital-raising to change. Some of this looks like the old forms of capital-raising but with different rights allocations: Entrepreneurs issue stock, which is a normal instrument, but the stock has no voting rights, which is weird. (Just like junk bonds were like regular bonds, but for companies with worse credit.) But there's no reason for it to stop there: Why shouldn't the new paradigm involve very new forms of capital raising? Why start from the package of rights implicit in stock? Why issue stock at all? Why not issue tokens? Or T-shirts?

What's Eddie Lampert up to?

William Cohan profiled hedge-fund manager and Sears Holdings Corp. Chief Executive Officer Eddie Lampert in what is billed as Lampert's "first in-depth interview in 15 years," but he also also interviewed Renaldo Rose, who kidnapped Lampert in 2003, and I feel like interviews with hedge-fund kidnappers are considerably rarer than interviews with even reticent hedge-fund managers:

Rose says that after being abducted Lampert “freaked out and one of the guys started punching him in the head. So I had to yell at them: ‘Listen, you both calm down. Keep quiet and you’re gonna be all right.’ I made [Lampert] a promise, ‘Listen, you don’t give us no problem and we’ll let you go.’ And he did, so he never freaked out again after initially.”

It still haunts Rose today that he might not have gone to prison had he killed Lampert and the other kidnappers: “So it was either like, O.K., get rid of everybody. [But] with Shemone Gordon, [Lampert] was like family almost. He argued against all that. I still think we should have just got rid of everybody. But, I don’t know. I did have to consider that. Lampert . . . never gave any problems, so I kind of had to keep my word on that.

It's an admirable, though arguably inefficient, commitment to the sanctity of contracts. Plenty of people in Lampert's line of work are often willing to strategically default on their commitments when it is in their best interests to do so; it's interesting that the kidnapping business has different standards.

People are worried about bond market liquidity.

"'It’s becoming like a deserted village. All that’s left is for us to fade away and die,' said Jun Fukashiro, who oversees bond investments for Sumitomo Mitsui Asset Management Co.," and while I always found U.S. worries about bond market liquidity to be a little overblown, they've got nothing on Japan. To be fair, "in Japan, the daily volume of government-bond trading is often measured these days not in trillions or billions, but in millions of dollars—and sometimes just with a single digit, zero," so they really do have something to complain about.

Things happen.

Citigroup to Again Be a Nationwide Bank, but in Digital Form. Inside the Secret Plot to Reverse Brexit. A private equity star's picks shine ... until cash-out time. The Long Shadow of GE Capital Looms Over GE. SoftBank Probes Who Was Behind Smear Campaign Against Top Executives. "Elon Musk is hopping on the #DeleteFacebook digital bandwagon." Cerberus’s Remington Gunmaker Seeks Bankruptcy Protection. How Nasdaq C.E.O. Adena Friedman Beat the Odds on Wall Street. Ethical cobalt. Roast Chicken à la Dyson. Florida Men — 1 Disguised in Bull Costume – Allegedly Tried to Burn Down Ex-Boyfriend’s Home With Spaghetti Sauce. The Dying Art of Fishing for Shrimp on Horseback.

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.

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.

For more columns from Bloomberg View, visit http://www.bloomberg.com/view.

©2018 Bloomberg L.P.