The sketchy world of Block Trades

[ValueArb]

Matt Levine hit it out of the park today. I've already posted separately his thoughts on Grayscales post ETF decline, the game theory behind Elon's demand for more Tesla shares. This is more inside baseball but I found his discussion of how block trades work on wall street fascinating, and why the head of block trading at Morgan Stanley got a light slap on the wrist from the SEC for inside trading (sortof?) clients while Morgan Stanley got hit for a quarter billion in fines/restitution.

 

Quote

Here’s a simple model of a block trade. There’s a public company, its stock trades on the exchange, its current price is $100. A big shareholder wants to sell a block of a million shares. This will drive down the price: supply and demand, more sellers than buyers, etc. Let’s say that fully selling all of the shares will drive the price down by $3.

 

You’re a bank, and the seller comes to you and says: “I want to do a block trade. I want you to buy all this stock at a firm price and resell it at your own risk. You have to buy it after the market closes in two hours; I’ll call you for your bid then.” How much should you bid? Well, on my numbers in the previous paragraph, the stock is at $100, selling all the shares will take $3 off the price, and you’ll be able to resell it at $97. So you should bid, like, $96, to add a little cushion (in case that $3 estimate of the price impact is wrong) and some profit for you.

 

But in fact the stock will move around over the next two hours. Maybe in two hours it’ll be $99, or $101. And reselling the stock will drive it down by $3. So really your bid should be (1) whatever the closing price is in two hours minus (2) say, $4 (the $3 effect of the sale, plus $1 for your profits). And in fact the seller will probably ask you for a bid that is not a dollar price but a discount (percentage or dollars) to the last sale price. If the stock closes at $99.73 and your bid is “down $4” then your bid is $95.73.

 

You have two hours to come up with your bid. You have two goals:

1. Accuracy: You want your bid to accurately reflect the price at which you can resell the stock (plus a little profit for you); you don’t want to pay $98 if you’ll have to resell the stock at $95.
2. Competitiveness: The seller is probably calling a bunch of banks, and will take the highest price (the lowest discount to the last sale). So you want to offer the highest price.

 

Here is a good trick: You spend those two hours selling all the stock. You pre-sell the million shares that the seller is offering; you sell them short. As you do this, naturally, the price declines, from $100 to $97, to reflect all of that selling. (Also maybe other stuff happens — maybe there’s economic news or other big buyers or sellers or whatever — so the stock ends up at $96 or $98 or whatever, but the $3 effect of all the selling is fully incorporated into the price.) And then at the end of that process you go to the seller and bid “down $1”: That is, you’ll pay $1 less than the last price; if the last price was $97 you’ll pay $96.

 

Now, two things are true:

1. Your bid is more accurate. You already know the price you can get for all that stock, because you’ve already sold it. If you sold it at $97 and bid $96, you have $1 of profit that is totally safe.[1]
2. Your bid is tighter, more competitive. Instead of bidding “down $4,” as you would have in the naive case, you bid “down $1,” because the $3 impact of selling the stock is already reflected into the price.

 

Notice that your bid is not actually better: In the naive case, the stock is at $100 and you bid down $4 ($96) to reflect the likely impact on the price of reselling all the stock. In the pre-selling case, the stock is at $97 and you bid down $1 ($96) because you’ve already resold all the stock. Either way the seller gets $96.

 

But if you are in competition, your competitors don’t know that you’ve pre-sold the stock. They see the closing price of $97 and think that that reflects the market’s valuation of the stock; they think that reselling the million-share block will require another $3 price drop. So they bid “down $4” — the naive bid — which works out to $93. And you bid $96, you win, you get the stock, and you make an easy $1 profit.

 

There are some problems with this trick. One is, if you do it and don’t win, you’ve pre-sold all this stock and don’t get to buy it at a discount; you have to go and cover your short in the market and might lose money. What if another bank does the same trick? Then, between you, you’ve sold two million shares, the price has gone down too far, and whoever loses the auction will have to buy back the million shares at higher prices. What if five banks do this trick?

There are ways to mitigate this risk. A simple one is, instead of actually pre-selling the stock yourself, you just call up the likely buyers — the big hedge funds that tend to buy block trades — and say “hey, got one coming for Amalgamated Widgets, you in?” And then the hedge funds give you an order, and they short the stock, and it goes down to $97, and you proceed as above, bidding $96 and winning the auction, and you sell them the stock at $96.50 and they make money and you make money. This moves the risk from you to the hedge funds.[2]

 

Another problem with this trick is that it is pretty bad customer service. It looks like good customer service: You can give the seller a tight bid on the block, and then resell it quickly and efficiently. But it is actually bad customer service, because you drive down the price ahead of the block trade and make the auction for the seller’s shares less competitive.[3] Ordinarily sellers will swear you to secrecy; they will call you up and say “I want you to bid on this block, but don’t tell anyone that it’s coming.” If the seller finds out that you did pre-sell the shares, or that you leaked the news to hedge funds to pre-sell the shares, they will be mad at you.[4]

 

A third problem with this trick is that it is … I was about to say “very illegal,” but apparently it is only, like, the regular amount of illegal. Possibly less. It is mildly illegal? (Not legal advice!) It is a form of insider trading, or rather front-running: Your client (the big shareholder) called you and said “I want to sell stock after the close today,” and probably swore you to secrecy, and then you went and traded on it, or told hedge funds to trade on it, so you could make more money for yourself with less risk. You had nonpublic information about a big shareholder’s plans to sell, that information was material, you had a duty to keep it confidential, and instead you used it to tip off some hedge funds (or trade yourself) to make money for yourself.

 

A certain amount of illegal. On Friday, the US Securities and Exchange Commission and US federal prosecutors in Manhattan settled a case against Morgan Stanley and Pawan Passi, its former head of US equity syndicate. “Equity syndicate” is the desk that does this stuff; specifically:

1. When big shareholders ask for a price on a block trade, equity syndicate is the desk that gives them the price.
2. When Morgan Stanley wins a block trade, equity syndicate is the desk that coordinates selling the shares out to investors.
3. When big shareholders ask for a price on a block trade, equity syndicate is the desk that calls potential investors and say “hey what price would you pay for this block?” to formulate its bid.

 

Oh, I mean, not really; No. 3 is illegal. But kind of. Certainly Pawan Passi did that, though now he is the former head of equity syndicate. The SEC says:

Selling shareholders or their agents then typically send a bid-wanted-in-competition (“BWIC”) email to the investment banking firms that had expressed an interest on the initial outreach calls. BWICs are typically sent to firms two to three hours before the market close. They reiterate the information provided on the outreach call (the seller’s identity, the stock, and size of the block), request bids by 4:05 pm ET that day, and set forth the confidential auction process. BWICs typically pre-condition a firm’s participation in the auction process on the firm agreeing to keep information about the potential block trade confidential. Selling shareholders or their agents require the select group of auction participants to keep such information confidential because if the market becomes aware of an imminent block trade, as stated above, the price of the stock might decline, and the seller might receive a worse price from the auction. …

During the Relevant Period, Passi disclosed non-public, potentially market-moving information received from selling shareholders or their agents about block trades to certain potential purchasers of the block, including a portfolio manager (“Portfolio Manager A”) in the London, England office of a Hong-Kong-based hedge fund (“Hedge Fund A”) while the auction process was ongoing. ...

Passi provided this information to Portfolio Manager A with the understanding that he would take large short positions in the stock in anticipation, and prior to the execution, of the block trade, and that, if Morgan Stanley won the auction, Portfolio Manager A would request and receive allocations from the block trade to cover those short positions. 

Passi knew that providing information about block trades to buy-side investors, such as Portfolio Manager A, could cause the stock price to decline if those investors sold significant amounts of stock in advance of the block trade.

Portfolio Manager A’s pre-positioning activities benefitted Morgan Stanley as they ensured that there would be a large buyer for at least a portion of the block trade, thereby lowering Morgan Stanley’s risk on the transaction, and giving the firm comfort to offer a tighter and more competitive bid.

It seems that Passi called several investors to place a bunch of the stock (and thus largely de-risk Morgan Stanley’s bid), but it’s not clear how methodical he was; it’s possible that he just called a few of his favorites. For instance this guy, from the criminal statement of facts:

Hedge fund investors who received confidential information from the Head of the Desk [Passi] and Employee-1 about upcoming blocks recognized that this information allowed them to profit in ways they otherwise would not have. For example, an investor working at a Nevada- based hedge fund (“Investor-3”) told the Head of the Desk in an August 2021 call, “I know who my daddy is,” referring to the assistance that the Head of the Desk had provided Investor-3 in profiting from block trades. Investor-3 stated in the same call, again referring to block trades, that the Head of the Desk had “put [Investor-3] in the ... game,” and that Investor-3 “would be at the kiddie table if it wasn’t for” the Head of the Desk.

Because, yes, if you do this right, it is nearly risk-free profit for the hedge fund (who shorts the stock during the day and then gets it back at the close at a discount to the trading price) and for Morgan Stanley (who locks in buyers a higher price than it agrees to pay the seller). You don’t need to be a big or good hedge fund to get the call for this trade; you just need to be a loyal friend of Passi’s.

 

Anyway, two related points. One is that this is illegal, but it’s on a continuum. As a general matter, when you want to sell a large block of stock, you have to balance your desires to (1) tell everyone, to drum up buyers and (2) tell no one, to avoid anyone front-running you. In different circumstances, you might strike that balance in different ways. Sometimes, it is good customer service for Morgan Stanley to “socialize” a block trade before bidding, to pre-market it before actually doing the trade. Ordinarily to do that you’d wall-cross the investors first: You’d call them up to ask “hey would you buy this block,” but only if they first promise not to short the stock ahead of the trade. In fact Morgan Stanley marketed that service; from the statement of facts:

When Morgan Stanley’s marketing materials referenced investor outreach, the materials suggested that such outreach would be done only to investors that had been wall-crossed. Sellers were told that while block trades were one option for selling their shares, another option was an “overnight with wall-cross and backstop,” which was described as a registered sale that would “launch[] publicly after wall-crossing several investors capable of acting as anchor investors.” The stated advantage of this “overnight with a wall-cross” was that “anchor investor feedback helps provide early pricing insight ahead of public launch.” In other slides shown to potential sellers, the possibility of wall-crossing a select number of investors “ahead of launch to maximize demand visibility/pricing” was offered as a possible way to execute a registered block trade. 

It’s just that, when sellers instead chose to do block trades without pre-marketing, Passi pre-marketed the shares anyway — not to benefit the client but to benefit Morgan Stanley. (Also Passi “did not wall cross any investors before sharing confidential information about the Relevant Blocks with them.”)

Even without wall-crossing, though, there are some gray areas about what sort of information-gathering a syndicate desk can do before putting in a bid. Like:

• Calling up some hedge funds at 2 p.m. and saying “hey we got a block of Amalgamated Widgets coming after the close today, are you in? If so you should short the stock now”: pretty illegal.
• Calling up some hedge funds at 2 p.m. and saying “just hypothetically, how are you feeling about the widget sector these days,” to get a sense of investor demand: maybe okay?

 

You can get some deniability on these calls. Though some sellers know to look out for this; the SEC notes that one selling shareholder’s agent sent Morgan Stanley a bid-wanted-in-competition request that said:

Prior to the purchaser being selected, please do not engage in any discussion with potential investors or purchasers, even hypothetical ones, including discussions in which several names are mentioned in order to solicit general interest.

The other point is that this is illegal, but apparently not that illegal. Morgan Stanley paid a total of $249 million of penalties to the SEC and $153 million to the Justice Department, though there is overlap in those numbers. (And some of the payment is for restitution to the block sellers who got ripped off.) Meanwhile Passi will pay $250,000 to the SEC and agreed to a deferred prosecution agreement in the criminal case: He has to “demonstrate good behavior and fulfill the terms of the DPA, in which case PASSI will not be further prosecuted criminally.” So it’s not quite real insider trading: If you do this stuff they don’t send you to jail; you get a warning first.

 

Edited January 16 by ValueArb

[Gregmal]

Wow I had not seen this. But again, it just shows what kind of scumbags the run of the mill, non professional investor has to deal with. 

[Gmthebeau]

On 1/16/2024 at 4:24 PM, ValueArb said:

Matt Levine hit it out of the park today. I've already posted separately his thoughts on Grayscales post ETF decline, the game theory behind Elon's demand for more Tesla shares. This is more inside baseball but I found his discussion of how block trades work on wall street fascinating, and why the head of block trading at Morgan Stanley got a light slap on the wrist from the SEC for inside trading (sortof?) clients while Morgan Stanley got hit for a quarter billion in fines/restitution.

 

 

 

yea great write-up.  There is a ton of manipulation that goes on which is why it is so hard for most people to beat an index fund.  You can see it in stocks and ETFs that go sideways to down for months and months and months and months....then suddenly they ramp 30% in 2-3 weeks - for no reason.  It wasn't that buyers suddenly poured in the doors.  It was they got everyone who they thought would sell at lower prices to finally sell, and boom up it went.   

Edited January 19 by Gmthebeau

[gfp]

10 minutes ago, Gmthebeau said:

 

yea great write-up.  There is a ton of manipulation that goes on which is why it is so hard for most people to beat an index fund.  You can see it in stocks and ETFs that go sideways to down for months and months and months and months....then suddenly they ramp 30% in 2-3 weeks - for no reason.  It wasn't that buyers suddenly poured in the doors.  It was they got everyone who they thought would sell at lower prices to finally sell, and boom up it went.   

 

I mean, those kind of shenanigans sound like they would make it easier to beat an index fund.  But I guess you said "most people" and most people aren't going to beat the S&P regardless of the opportunities these type of shenanigans regularly create.

[Gmthebeau]

2 minutes ago, gfp said:

 

I mean, those kind of shenanigans sound like they would make it easier to beat an index fund.  But I guess you said "most people" and most people aren't going to beat the S&P regardless of the opportunities these type of shenanigans regularly create.

 

Right because most people will get shaken out, then start chasing what is working right when the trend is about to change.   They can't control their emotions.