Money Stuff: Some Stocks Always Go Up 5%

Let’s say that you are a market maker and you trade a bunch of small, dumb, thinly traded stocks. The last trade in Amalgamated Cannabis & AI Inc. was 20 minutes ago; someone bought 100 shares for $4 each. There has been no news since then. You have to quote a price at which you would sell 100 more shares of Amalgamated. What price should you quote?

There is, weirdly, an answer. The answer is $4.20. The way you compute the answer is (1) you take the price of the last trade ($4) and (2) you add 5%. Why is this the answer? Well:

• Robinhood Markets Inc. is a retail broker that, uh, sends you a disproportionate number of your orders to buy small, dumb, thinly traded stocks.
• It is fairly common behavior for a retail customer to go to the app, go to the screen for Amalgamated Cannabis & AI Inc., and put in a market order to buy 100 shares. (A market order is an order to buy 100 shares at the best price Robinhood can get.)
• Because market orders are dangerous — what if the price spikes up to $69 for glitch or fat-finger reasons? — Robinhood, like some other retail brokers, automatically converts these market orders into limit orders.
• Specifically, the limit is up 5% from the last trade.
• So when someone goes on the app and says “buy me 100 shares of Amalgamated, I don’t care about the price,” Robinhood translates that into “buy 100 shares of Amalgamated at a price of no more than $4.20.”
• So you sell at $4.20.

If you offered $4.21, the Robinhood customers wouldn’t be able to trade with you. If you offered $4.19, you’d be leaving money on the table.

This analysis is not really correct — if other market makers are competing to trade Amalgamated, you might have to offer it at a lower price — and it tells you nothing at all about the price you should offer, like, Nvidia at. [1] For a big actively traded stock, there will be enough competition that Robinhood’s up-5% anchor doesn’t matter at all. But for after-hours trading of a small retail stock, this actually turns out to be a surprisingly useful heuristic:

Order collaring, which automatically converts default market orders into limit orders with a 5% spread over prior prices, has been used at Robinhood to protect retail investors from trading at unfavorable prices. However, in this paper, we provide empirical evidence that this policy actually harms retail traders by increasing their trading costs. Through quasi-experiments involving Robinhood's trading hours, their extension, the discontinuity around the 5% spread, and the asymmetric implementation of order collaring on retail buy and sell orders, we find that retail traders on Robinhood face higher spreads. Specifically, during regular trading hours, their collared orders execute at an additional spread of 4.2 basis points, and in extended hours, a significant number of their orders execute precisely at the 5% spread over closing prices. Furthermore, our analysis of the GameStop short-squeeze episode reveals that the order collaring policy is associated with extreme price movements in stocks. 

That’s the abstract to “Democratizing or Demoralizing: The Impact of Robinhood’s Order Types on Retail Trading Costs,” by Preston Mantel and Mehmet Saglam. From the paper:

In addition, we examine how prices exhibit extreme volatility due to the presence of the order collaring policy at 5%. Since retail traders are subject to excessive spreads up to 5% on each trade, correlated retail trading on Robinhood (e.g., successive retail buys) may cause sudden price movements as strategic market makers know that their default orders will be converted to limit orders priced at a 5% differential over the prior transaction prices. Consequently, the order collaring policy can cause extreme price volatility as observed on meme stocks that are traded heavily by retail investors.

The current explanation of the unprecedented price volatility observed in January 2021 is primarily based on a short-squeeze theory. In their report, the SEC found that short covering, however, did not make up the majority of buying pressure for the duration of the GameStop (GME) price runup. We investigate whether the order collaring policy could have also played a substantial role in these extreme price movements. Computing maximum and minimum transaction prices around the closure of Robinhood, we find consistent evidence suggesting that the order collaring policy is potentially introducing extreme volatility to the observed prices.

The intuition is something like this:

1. Imagine that the only buyers of a stock are price-insensitive Robinhood customers putting in market orders, which convert into orders to buy for no more than 5% above the last trade.
2. Imagine that the only sellers of the stock are gleeful ruthless electronic trading firms who know Fact 1. [2]
3. If the first retail buyer buys the stock at $100, the next retail buyer is going to buy it at $104.99, the one after that is going to pay $110.23, etc. 

This is not really right — a real market will include price-sensitive buyers, fundamental sellers, and competition among market makers to win business, so prices can’t really go up 5% on each trade — but it does seem like it might be part of what is going on.

In possibly related news, here is a fun paper by Cliff Asness on “The Less-Efficient Market Hypothesis,” whose basic points are:

1. It’s harder to be a value investor: “The ratio of the valuation (using whatever measure you favor) of expensive stocks to cheap stocks” has been higher, in recent years, than its historical average, so if you are in the business of buying the cheap stocks you have not had a good run.
2. This is an indication of market inefficiency: If the expensive stocks are really expensive, that’s probably because they are in a bubble, not because they have become much more valuable relative to the cheap stocks.
3. Why is this happening? Possibly some combination of low interest rates, increased indexing and more meme-driven retail investors.

The first point seems relatively straightforward. The second seems more debatable, [3] and you could tell stories about how actually today’s high-priced stocks should be worth a lot more than the low-priced stocks, though Asness has answers to some of the big ones. [4]

But my favorite is the third point, possible explanations for why the highest-price stocks might be irrationally high in modern markets. Asness points to indexing:

Imagine some fraction of the market passively hold the index and the rest are active traders/investors trying to outperform. Now divide this active group in two (this is the obnoxious part). One group are sharks, the other minnows. Minnows make bad decisions based on emotion, story, tastes that are not relevant for risk and return, and behavioral biases. Sharks outperform by taking the other side of the minnows’ misguided positions. Well, if indexing has grown, whether this has made markets more or less efficient comes down to whether more sharks or more minnows moved to indexing. If more sharks have moved, the remaining sharks should have an easier time making money over the long term as there is less competition in betting against the minnows, but the minnows have more influence than they used to at the short to medium term. Prices are a dollar-weighted average of opinions, and if a larger fraction of this is misguided, so will be prices. Given that indexing is by definition smart for the average investor to do, and given the celebrated “death of value investing” and concomitant shrinking of value investors’ assets and influence, I find it very easy to believe more sharks have moved to indexing than minnows, making the market more subject to the wild valuation spreads we’ve documented here. But it ain’t close to a proof (sorry, there won’t be any of those).

My intuition is the opposite: I would think that clever professionals would continue to trade actively, while uninformed retail investors would react to the rise of indexing by putting all their money into index funds. But his story is plausible; it would go something like this:

1. It used to be that retail investors mostly handed their money to mutual funds, where professional investors tried to pick the best stocks. Or they handed their money to wise financial advisers, who also tried to buy them the best stocks. And then some retail investors traded their own accounts badly.
2. Now, the first group instead hand their money to index funds, which don’t try to pick the best stocks. And the second group hand their money to wise financial advisers, who do the wise thing and put them in index funds. And then the third group continue to trade their own accounts badly.

That is, the model is that the smart retail investors used to give their money to smart professionals, and now they index instead. And the dumb retail investors are still dumb retail investors.

But dumber, because of memes:

Has there ever been a better vehicle for turning a wise, independent crowd into a coordinated clueless even dangerous mob than social media? Instantaneous, gamified, cheap, 24-hour trading now including “one-day funds” (Greifeld 2024) [5]  on your smartphone after getting all your biases reinforced by exhortations on social media from randos and grifters with vaguely not-safe-for-work (NSFW) pseudonyms filtered and delivered to you by those companies’ algorithms which famously push people to further and further extremes. What could possibly go wrong?

I don’t think we are quite there yet — I write a lot about sophisticated multistrategy hedge funds that make market-neutral stock bets to push prices in more efficient directions and rigorously measure their production of alpha! — but a market that consisted only of index funds and meme-stock investors would be pretty wild.

Sometimes a big company wants to buy a little company with a promising but unproven technology. If the technology works out, the big company will make a lot of money with it; if not, not. Often these deals will be structured with an earnout: The acquirer will pay some cash for the target at closing, but will also promise to pay it more cash, later, if the technology works out and makes a lot of money. The target’s shareholders get some certainty at closing, but they also share in the upside of their technology; the acquirer gets some hedge against the risk that the technology doesn’t work out.

There is an obvious problem with this: Before the acquisition, the target was a scrappy startup working as hard as possible to develop its promising technology. The odds were long, but it had the right team in place and everyone was hungry and motivated. After the acquisition, the target is a division of a giant corporation with a lot of bureaucracy, and all of its key employees are rich. Is it possible that their chances of building the technology will go down?

Also, if you are a shareholder of the target company, you might worry that the acquirer will hold up progress on the technology, in order to avoid paying the earnout. “If we commercialize this technology, we will have to pay the target’s shareholders another $1 billion, so let’s not,” the acquirer might think. In general you’d hope this wouldn’t happen — the point of the deal is for the acquirer to make a ton of money from the technology, and pay a portion of it to the target shareholders in the earnout, so incentives are aligned — but at the margin I suppose the earnout could hurt. 

One other worry: What if the acquirer doesn’t want the target’s product, but just wants to raid it for parts? What if the acquirer is already building some competing technology, and wants to buy the target just to help with that effort? The acquirer gets some patents, some employees, some ideas, and uses them to build its own product, shutting down the target’s competing product. The acquirer makes a lot of money from the deal, but doesn’t have to share any of it in the earnout, because the target itself goes nowhere under the acquirer’s control.

This is all stuff you can think about and try to write into the contract, but it’s all risky, and it can be hard to be very specific in advance. At Axios, Dan Primack reports:

Johnson & Johnson broke its promises to investors in Auris Health, a surgical robotics startup it bought five years ago for $3.4 billion. Now it's got to shell out another $1 billion, based on a ruling yesterday from Delaware Chancery Court. ...

This appears to be the largest legal reward ever granted in an investor earnout dispute, and could change the way that such provisions are written.

Here is the opinion, by Delaware Vice Chancellor Lori Will. The story is that Auris had two surgical robots that J&J wanted (Monarch and iPlatform), though J&J was working on its own surgical robot (named Verb). But, the judge writes, “Verb was falling increasingly behind the schedule J&J had announced to the market, despite J&J’s colossal investments,” and J&J worried that a competitor might buy Auris. So J&J decided to buy Auris instead, and they agreed to a deal with $3.4 billion up front and up to $2.35 billion in earnouts dependent on (1) getting various regulatory approvals for Monarch and iPlatform and (2) reaching net sales milestones.

And then they didn’t hit those targets. In part this was because the US Food and Drug Administration changed its approval processes for these robots, which made it harder to hit the regulatory milestones. But in part it was because “instead of providing efforts and resources to achieve the regulatory milestones, J&J thrust iPlatform into a head-to-head faceoff against Verb.” And although iPlatform won the faceoff, “progress toward iPlatform’s regulatory milestones ceased while technical debt from shortcuts in its development amassed.” And then everything got sort of mashed together:

On December 5, 2019, J&J management recommended to the J&J Board that the company proceed with “a combined platform where Auris’ iPlatform is augmented by Verb assets including the open surgeon console, intra-procedure data capabilities and the surgeon portal.” The combination robot, called “iPlatform+,” was described as a “[n]ext generation robotic platform designed with more flexibility, more control, and more information to elevate [the] surgeon experience [and] improve patient care.” …

Afterward, J&J worked to integrate Verb’s resources into Auris. The Auris leadership team was largely sidelined. A “[f]ull [s]peed [m]igration” of more than “200 [Verb] employees” to the iPlatform team commenced. A calamity of excess and redundancy resulted. Hostility abounded between the two factions, which had just faced off in Project Manhattan for the survival of their respective projects. J&J soon announced layoffs on both teams.

Within a year of the integration, every engineer from legacy Auris’s iPlatform clinical engineering team left the company—a “devastating” loss for the program. Meanwhile, Verb software engineers insisted on re-writing iPlatform’s code. Significant attrition of legacy Auris software engineers followed.

“By the end of 2021, iPlatform was shelved,” and the earnout milestones were written down to zero. The former Auris shareholders sued, arguing that the merger agreement required J&J to use its “commercially reasonable efforts” to develop Auris’s technologies consistent with its “usual practice” for a “priority medical advice,” and that it didn’t do that. The judge agreed:

Instead of being prioritized, J&J subjected iPlatform to efforts that impaired its development and ability to secure planned clearances. J&J’s efforts benefitted another device— Verb—at iPlatform’s expense. It is obvious from the record that J&J’s efforts toward the iPlatform regulatory milestones were not commercially reasonable, as defined in the Merger Agreement. J&J’s breaches are, instead, reasonably certain to have caused iPlatform to miss its regulatory milestones.

It is hard to write merger agreements. They are sort of like prenups: When you are negotiating the merger agreement, everyone loves one another and is excited about their union, and you don’t want to think too hard about what might go wrong. Here, Auris was going to get a pile of money, and J&J was going to get some robots that it thought would make it an even bigger pile of money. Everyone’s interests were aligned, and it would have been rude to make very specific demands about how J&J developed iPlatform. “We’re all in this together, and it’s going to make us rich,” is the basic thinking, so the contract is not particularly clear about what would happen if that changed.

Maybe my favorite company in the world these days is Venture Global LNG Inc., which is an innovator in liquefied natural gas production and contract interpretation and general smugness. We have talked about Venture Global a few times before; the basic story is:

1. Venture Global was a startup that wanted to build a liquefied natural gas plant.
2. To raise the money to do this, they signed long-term fixed-price supply contracts with big energy companies, committing them to deliver LNG to those companies once the plant was up and running.
3. They got the money, built the plant and started producing LNG in early 2022.
4. Also in early 2022, though, Russia invaded Ukraine and the price of LNG shot up.
5. Rather than sell the LNG to the big energy companies at the contracted prices, Venture Global would prefer to sell it on the spot market at much higher prices.
6. So it has been doing that, for two and a half years.

The rationale for this is that, while the plant absolutely is producing and selling lots of LNG, it “has yet to enter commercial operations,” because, I don’t know, there’s some problem with the power supply or something. It is an extremely technical quibble, but the contract says that it has to start fulfilling the contracts when the plant is in commercial operation, and that weirdly keeps not happening. I once joked that the problem is that they haven’t painted one last rivet, “and then they say ‘oops, still not finished, see, unpainted rivet,’ and keep selling LNG at high prices to spot buyers.”

Anyway I love them in part because this is a hilarious trade but mostly because, every time there is a news story about it, (1) they still haven’t painted that rivet and (2) they give a comment that is essentially “nyah nyah nyah nyah nyah.” Here’s the Financial Times with the latest. The story is always the same — the contracted customers keep complaining, going to arbitration, and not getting their gas — but look at this quote:

“Due to our ability to produce first LNG during construction we have been uniquely positioned to bring more incremental molecules into the market which lowers prices, not raises them,” the spokeswoman said. “Venture Global is honouring its contractual obligations to its long-term customers in strict conformity with its long-term contracts.”

“Honouring its contractual obligations to its long-term customers in strict conformity with its long-term contracts” means “you will absolutely never get us to paint that rivet.”

Sure I mean of course:

Apollo Global Management Inc. is planning to build out a secondary trading desk for private debt provided by a handful of lenders, according to people with knowledge of the matter who asked not to be identified as the details are private.

Plans are preliminary and Apollo may decide to not pursue them, the people added. An Apollo representative declined to comment. 

The big asset manager would follow some others in the $1.7 trillion private credit industry that have set up operations to buy and sell the illiquid loans. Golub Capital has been trading direct loans, having bought and sold about $1 billion the first half of this year. 

“My basic theory of private credit,” I once wrote, “is that (1) it’s kind of a new thing but (2) it will quickly be absorbed into the old things.” Part of the initial idea of private credit was that lenders could get a premium in exchange for taking illiquidity risk. But wouldn’t it be nicer to also have liquidity?

I mentioned yesterday that one way to characterize the basis trade is that “pension funds want to get their duration via Treasury futures, and Treasury wants to sell bonds, so hedge funds intermediate that trade by buying the bonds, using them to manufacture futures, and collecting a spread.” That is not quite right. It’s pretty close, but actually the big users of the basis trade seem to be mutual funds. Here’s a June Federal Reserve discussion paper on “Reaching for Duration and Leverage in the Treasury Market,” by Daniel Barth, R. Jay Kahn, Phillip Monin and Oleg Sokolinskiy:

We show substantial variation in mutual funds’ use of Treasury futures, both over time and across funds. This variation from mutual funds drives much of the time series variation in aggregate Treasury futures open interest, including over 60% of the recent rise in Treasury futures positions. We provide evidence these Treasury futures positions are largely attributable to mutual funds “reaching for duration” in order to track the duration of a benchmark index with high cash Treasury exposure. Specifically, we show mutual funds use futures to fill the gap between their portfolio and the index that results when they tilt their cash positions toward higher return but lower duration assets, such as mortgage-backed securities and equities, and away from cash Treasuries. Treasury futures positions are more common in mutual funds which indicate a focus on dual objectives of duration management and total return whose style has a higher allocation to Treasuries. Reaching for duration allows funds to track their index better at lower cost, but increases leverage in the Treasury market both through mutual funds long Treasury futures positions and through the leverage of hedge funds who take the corresponding short positions in Treasury futures.

The broader point remains that there are fundamental investors who want to own Treasury futures, the US government wants to sell Treasury bonds, and the service of buying the bonds and transforming them into futures is largely done by basis-trading hedge funds.

A running theme of this column is that it is good, for your financial career, to lose a billion dollars: It proves that you are a confident risk-taker, that people liked you enough to trust you with their money, and that you have learned some lessons about not losing the next billion dollars. I mean, it suggests that last part. It’s not proof. Another possibility is that you are just the kind of person who is very good at charming people into trusting you with a billion dollars and then immediately losing it. This is not a perfect signal. It is a fun signal, though. The interpretation is something like “people are mean reverting, and nobody who had that sort of disaster would do it again.” Some people are not like that, though, and having a disaster only makes them hungrier for another disaster.

Anyway:

Billy McFarland is a little over a year into planning Fyre Festival II, and he still doesn’t know where it’s going to be. Or when. Or who’s performing. But he says it’s definitely happening. One hundred percent, absolutely for-sure happening.

After all, he’s already sold tickets.

“I shouldn’t talk in absolutes, but I will here,” McFarland says between sips of a matcha latte on a sweltering June day. “Fyre II has to work.”

Fyre I was so bad that McFarland spent four years in prison for it: He sold tickets to a luxury concert experience, and then didn’t deliver the performers or food or venue or anything else that he promised. This time, though, he has learned important lessons and won’t sell tickets until he has made sure he can get the … oh wait sorry never mind he’s already sold tickets, I’m sure it’s still fine though.

Biden Prepares to Block $14 Billion Steel Deal. Verizon Strikes $9.6 Billion Deal for Frontier. BNY seeks growth in investment services with managed account deal. New Mountain Merger to Create $3 Billion Medical Payments Firm. Ford, Coors Light and Other Brands Retreat From a Gay-Rights Index. Robby Starbuck: the activist pushing US companies to ditch their DEI vows. Ghana’s Bondholders Back $13 Billion Restructure Exchange Offer. Ex-Discover Star Sues Over Pay Clawback Before $35 Billion Deal. Three former Wirecard executives found personally liable for €140mn. How Local Governments Got Hooked on One Company’s Janky Software. I’m at the combination casino/water park.

If you'd like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Or you can subscribe to Money Stuff and other great Bloomberg newsletters here. Thanks!