|Good and fun read|
What Was Bill Hwang Thinking?
Archegos’s stocks went up and then they went down.
Matt Levine28 April 2022, 03:00 GMT+8
ArchegosHere is a simplified version of the Archegos story. Archegos Capital Management was a family office run by Bill Hwang, a former Tiger Cub hedge fund manager, that invested his personal fortune. Starting in about 2020, Archegos’s investment strategy consisted of buying a whole ton of shares of like 10 stocks, using mostly money borrowed from about a dozen banks. (Technically it did this buying using total return swaps rather than actually buying the stocks on margin, but that is a minor point.)
As Archegos kept buying more of these stocks, they went up, because generally if you buy a lot of a stock the price will go up. As the prices went up, Archegos had mark-to-market profits: The shares it bought earlier at lower prices were worth more, so it had made money. Archegos used these profits, leveraged with more money borrowed from its banks, to buy more of its favorite stocks. This made the prices go up more, which created more profits, which gave it more money to buy more stocks, etc., in what I guess you could call a virtuous cycle.
I don’t know how to write an ending for this story? I mean, I know how the story ended in real life, and it’s the obvious ending. The obvious ending is that, if you keep doing this, you end up owning enormous quantities of your favorite 10 stocks, owing enormous amounts of money to your banks, and having a very slim margin for error. If a slight breeze knocks down the price of one of your stocks, your banks will demand more money in a margin call, and you won’t have any cash because you have invested every cent in buying stocks with borrowed money. Your banks will be forced to sell some of your stocks, which will drive their prices down, which will lead to more margin calls and more forced sales and more price drops, etc., in what you would certainly call a vicious cycle.
And that is in fact exactly what happened. There was a small hiccup with one of its stocks: ViacomCBS Inc., a huge Archegos holding, saw its stock price shoot up and decided to raise some money by selling stock, which pushed down the price a bit. The result was that Archegos was absolutely vaporized almost immediately: It got margin calls that it couldn’t meet, its stocks were liquidated, their prices crashed, it lost all its money, and some of its less nimble banks lost billions of dollars when they were too slow to liquidate. It is the obvious outcome, and it happened.
But I don’t really understand it, because it is so obvious. Bill Hwang is not some dope; he is a long-time hedge fund manager with a Tiger Management pedigree. The story I laid out above is stupid, and it inevitably ends the way I described. Bill Hwang presumably did not want to lose all his money. What was his exit plan? How did he think the story would end?
The simple story is so unsatisfying that I have occasionally speculated that he might have had a better plan. One better plan is to do essentially what I laid out above — buy stocks with borrowed money, push them up, reinvest the profits to push them up some more — but then stop, call up your banks, withdraw your latest (enormous) round of mark-to-market profits, and instead of using them to buy more leveraged stock positions, use them to buy some gold bars and bury them in your backyard. Then a slight breeze blows over your stocks, you get margin calls, you say “sorry I have no money,” your stocks crash, your fund’s equity goes to zero, your banks lose billions on their loans, you get a lot of negative news articles written about you, you sigh and look chagrined, eventually everyone moves on and you dig up the gold bars. 1
I can’t say there was much evidence for this possibility but I liked it a lot. There were tantalizing hints: Archegos took $2.4 billion of cash out of its account at Credit Suisse Group AG just a couple of weeks before it blew up; where did that money go? And some former Archegos employees went out to start a new fundwith $50 million of their personal capital, suggesting that they at least got out in time. A version of the Archegos story in which Archegos was vaporized and its banks lost money, but Hwang emerged smug and victorious, is just more fun than the one in which he briefly got very rich and then mechanically lost it all.
There are other possible endings, other possible explanations for what Hwang might have been thinking. The most plausible one is: He liked the stocks. This theory is, like, your favorite stock is trading at $20 and you think it is worth $100. So you keep buying, in a more and more leveraged way, until you own like 30% of the shares and the stock is at $50. Then you hope that the company announces great earnings, everyone comes to see the story your way, the stock trades to $100, and you sell at a profit. You are making a huge, risky, but fundamental bet; your theory is not just “if I buy this stock it will go up” but also “and it should go up because the market will come to understand its fundamental value.” “If I buy this stock it will go up” has the unfortunate corollary “but then I will sell it and it will go down,” but if you have a lot of conviction on a fundamental view then maybe you think you can avoid that fate. Hwang’s reputation was as a risk-taking, high-conviction fundamental stock analyst, who once got in trouble for using fundamental information that was a little too good. (His old hedge fund shut down after he was charged with insider trading, whoops.) Perhaps he really liked his favorite 10 stocks and figured they’d grow into the very high valuations he created for them. Perhaps he wanted to bet as much as he could on that outcome.
Today the U.S. Securities and Exchange Commission and federal prosecutors charged Hwang and some of his employees with market manipulation and fraud. Ah, well. (“Bill Hwang is entirely innocent of any wrongdoing,” said his lawyer. “There is no evidence whatsoever that he committed any kind of crime, let alone the overblown allegations that pervade this indictment.”)
The market manipulation case is just what I laid out above: If you buy a whole ton of like 10 stocks, their prices will go up. Hwang bought a whole ton of his favorite stocks! From the criminal indictment:
The majority of Archegos’s positions were “long,” meaning the value of Archegos’s portfolio would increase if the prices of the stock underlying its swaps increased. To drive up the prices, HWANG amassed extraordinary exposure to those stocks through billons of dollars of purchases made with money borrowed from Counterparties based on lies and misrepresentations. The increased demand for these stocks, and the dwindling supply of freely trading shares, led to significant artificial appreciation in the price of each stock. This manipulative strategy harnessed the margin frameworks of Archegos’s Counterparties by recycling “excess” margin into further buying. In turn, the strategy generated additional price inflation and, thus, still more “excess” margin.
And from the SEC complaint:
The size of Archegos’s exposures to its Top 10 Holdings allowed Archegos to assert a dominant market position over the securities of these issuers.
For example, by late March 2021, Archegos’s cumulative cash equity and derivative SBS exposures to the following issuers equated to the following percentages of outstanding shares, based on Archegos’s estimate of those issuers’ floats: GSX Techedu Inc. (“GSX”) – Over 70% of outstanding shares; Discovery Class A – Over 60% of outstanding shares; IQIYI Inc. – Over 50% of outstanding shares; ViacomCBS – Over 50% of outstanding shares; Tencent Music Entertainment Group (“Tencent”) – Over 45% of outstanding shares; and Discovery Class C – Over 30% of outstanding shares.
Hwang knew that Archegos could impact markets through the exercise of its sheer buying power. For example, in June 2020, when asked in a text message by an Archegos analyst whether ViacomCBS’s stock price improvement that day was “a sign of strength,” Hwang responded, “No. It is a sign of me buying,” followed by a “tears of joy” or laughing emoji.
These are comically large numbers; Hwang owned, economically speaking, a majority of the shares of a number of big public companies. (“SBS” here stands for “security-based swap,” i.e. the total return swaps by which he did the bulk of his investing: His banks would write a derivative giving him economic exposure to the shares, and the banks would often, though not always, buy the underlying shares to hedge that derivative.) His buying was also a big portion of the trading activity in those stocks many days:
During the Relevant Period, and particularly from January through March 2021, Archegos’s trading of the equities of and SBSs referencing its Top 10 Holdings frequently exceeded 20%, often reached 30%, and even surpassed 40% of certain issuers’ daily trading volume.
Is this market manipulation? Man, I don’t know. Anyone who buys a lot of stock will make the price of the stock go up, but surely just buying a lot of stock can’t be a crime? The SEC and prosecutors hint that it might be illegal to buy more than 10% to 15% of a stock’s daily volume; from the SEC complaint:
Hwang and [Archegos head trader William] Tomita knew that trading in large volumes on a given day – at percentages of more than 10% to 15% of the daily trading volume of a specific issuer – would create upward pressure on the share price and often result in the share price increasing.
And from the indictment:
HWANG also traded the relevant securities in amounts far exceeding volumes known within the securities industry and to HWANG to affect market prices. Specifically, HWANG, Tomita, and others working at Archegos understood that buying or selling more than approximately 10-15% of a day’s total trading volume of a given stock would likely affect the market price in the stock. HWANG routinely directed trading in excess of these volumes.
People buy 15% of daily volume all the time. Elon Musk bought more than 15% of the volume of Twitter Inc. stock on several different days this year, 2 and the stock did in fact go up as he was buying. He obviously did not do this to manipulate the stock; he did it because he wanted to own Twitter (and now he sort of does). If you buy stock aggressively, it will go up, but people do not always buy stock aggressively to make it go up. Sometimes they buy stock aggressively because they want to own a lot of it, fast.
Ordinarily, my rough but useful theory of what “market manipulation” is goes like this:
You buy a lot of stock.The stock goes up as you buy it.You send bad chat messages to your buddies saying “lol I am manipulating this stock.”
Step 3 is important! This is not a complete legal definition or anything, but the basic point is that market manipulation is a crime of intent. If you buy stock because you think it is undervalued, and it goes up because of your buying, then that’s fine, that’s just an inevitable fact of stock trading. If you buy stock because you want to push up the price, that’s bad. If you send chat messages saying “I am buying this stock to push up the price because I am a bad criminal,” it is easy to prove intent. If you don’t, then prosecutors have to rely on fuzzier circumstantial evidence, and it will be relatively easy for you to say “no I bought that stock because I thought it was a good investment.”
Here it does not seem like Hwang sent the bad chat messages. I have already quoted his text saying “‘No. It is a sign of me buying,’ followed by a ‘tears of joy’ or laughing emoji,” but that one is obviously fine! If you asked Elon Musk, as he was accumulating Twitter, “why is Twitter’s stock going up,” I think he would have reasonably replied “because I am buying a lot of it,” and he would have added an emoji. That was why it was going up, but that was not why he was buying it. “The stock is going up because I am buying it” is straightforwardly true and neutral. “I am buying it to make it go up” is the bad one, and this message simply does not say that.
That said, is there circumstantial evidence that Hwang was doing weird stuff? Sure, a little. For instance, from the SEC:
None of this trading was based on a principled view of the true value of a particular issuer and instead was intended to artificially inflate share prices.
Indeed, Hwang essentially sidelined his research operation, ignoring their stock price targets in favor of his own outsized stock price targets, which were often orders of magnitude larger than what his research operation had determined on its own and, unlike his research operation, based on little or no analytical support.
Again I do not think it can be a crime for a hedge fund manager to overrule his analysts and say “no I really like this stock, let’s buy tons of it,” but it is not ahelpful fact here. If his analysts sent him a report saying “all these stocks are gonna quadruple so let’s buy as much as we can,” that would be helpful for him, but they did not. “HWANG all but stopped holding investment strategy meetings,” says the indictment, and he “frequently spent almost all of his workday with the traders” rather than the investment analysts.
Or there is this:
Archegos also engaged in substantial trading during the last 30 minutes of the trading day – that is, “marking the close” – to again push the stock prices of certain issuers, in which Archegos held long exposures, upward. The goal was that the upward movement in stock prices would lead to an increase in margin on Archegos’s SBSs, which was based on end of day valuations, thereby providing Archegos with even more leverage to purchase more exposure to the same issuers the next day, among other reasons.
Again it cannot be a crime to do most of your trading in the last 30 minutes of the day. U.S. stock market trading is increasingly concentrated in the last 30 minutes, because (1) that’s when closing prices are set, which matters for a lot of index-y investors who are measured on closing prices, and (2) that’s when the market is liquid, because that’s when the big funds are trading. If trading a lot at the close is a crime, then everyone is a criminal. And yet sure, yes, if you are trying to manipulate a stock to keep its closing price high to support your heavily leveraged positions, you absolutely would trade a lot at the close. It could be manipulation; you just would need more evidence of intent. (The indictment says that sometimes Archegos would sell in the morning, to “‘make room’ for later trades, when they would be most likely to create the desired impact on the securities price,” which does sound bad.)
Another example of Archegos’s non-economic trading occurred in March 2021. An Archegos trader messaged among others, Hwang, that after Discovery Class C shares opened at $61.68, Archegos should enter orders of “60.00-61.00 and be aggressive below 60.00 because some short term people see whether the stock can keep 60.00 floor as psychological level.” Hwang responded within seconds, writing “I LIKE THAT PLEASE GO AHEAD WITH $50 MIL.” Over the course of the trading, Hwang periodically sought status updates on trading to gain greater exposure to Discovery Class C shares, often then increasing the size of the exposure to be added and incrementally walking up the limits orders, with a message late in the day writing “FIRM LIMIT TO 63” and including instructions to “LET’S BE AGGRESSIVE.” Discovery Class C shares closed at $62.99 that day.
I don’t know. That says: Discovery opened at $61.68, and Hwang’s trader said “we should try to buy it if it goes lower.” They entered limit orders to buy it below where it was trading. That is not traditionally how you manipulate a stock up! They bought some, the stock did go up, and they raised their limits so they could buy more. I am not saying that that is not manipulation, just that it is not particularly compelling evidence of manipulation.
The worst fact, for Hwang, is that he allegedly lied to his banks about his positions. This is bad for several reasons. For one thing, it is an independent sort of fraud: Whether or not you are also doing market manipulation, it is illegal to lie to your banks to get them to lend you money. For another thing, it is a bad look: Lying to your banks to get money to buy stocks is not proof that you were buying those stocks to manipulate the market — those concepts are barely related — but it will look bad to a jury. Also, lying to your banks about your positions is a much more black-and-white thing than market manipulation; if you say “we own 1 million shares of XYZ with other banks” and actually you own 20 million, prosecutors can just point to the factual discrepancy rather than inquiring into your intentions. 3
From the SEC complaint:
As Hwang understood, the only way for Archegos’s manipulative trading strategy to succeed during the Relevant Period was for it to continue adding to its already concentrated positions. By the second half of 2020, it was becoming more and more difficult to add to positions. Archegos’s growing position sizes began to bump up against the Counterparties’ trading capacity limits – limits on the amount of SBSs that Counterparties would allow Archegos to hold in particular companies.
So, allegedly, Archegos lied to its banks to get them to increase its limits. From the indictment:
In or about January 2021, at the direction of [Archegos Chief Financial Officer Patrick] HALLIGAN, [Archegos Chief Risk Officer Scott] Becker claimed to representatives of Credit Suisse, in substance, that as of November 30, 2020, the largest gross position in Archegos’s portfolio … totaled 35% of Archegos’s capital. That was false. In truth, as of November 30, 2020, Archegos’s largest long position (VIAC) constituted approximately 96% of Archegos’s capital. Indeed, as of that date, Archegos had six positions that were greater than 35% of capital, including VIAC at approximately 96%, BIDU at approximately 67%, GSX at approximately 55%, TME at approximately 48%, IQ at approximately 45%, and VIPS at approximately 43%.
If you do most of your trading with borrowed money, you can have half your fund in each of six stocks, no problem. Also:
In or about February 2021, Archegos sought to raise its existing trading limits at UBS. In discussions about an increase, Becker told UBS, in substance, that Archegos could liquidate its entire portfolio in approximately two weeks, or ten trading days, selling at approximately 15% of average daily trading volume (“ADTV”), measuring from the prior 20 days. … In fact, to liquidate at 15% of ATV at that time, it would have taken, for example, approximately 134 trading days to liquidate Archegos’s position in VIPS and approximately 78 trading days to liquidate its position in VIAC. Even had Archegos sold its positions at 100% of ATV, which would have severely impacted the prices of the stocks, it would have taken approximately 20 trading days to liquidate VIPS, approximately 16 days to liquidate IQ, and approximately 12 days to liquidate VIAC.
Yeah that’s a lot of stock. 4 The basic problem is that each bank only knew about Archegos's positions with that bank. If you were a bank lending money to Archegos against some stock, and you knew it owned $1 billion of that stock through total return swaps with you, then that might be fine; $1 billion might be a reasonable position size, easy to liquidate quickly, etc. But if it also owned $10 billion more with other banks, that would not be fine: If Archegos defaulted, you’d all be selling at once and the stock would crater. Banks were very aware of this risk, and asked Archegos about it, and Archegos allegedly lied:
Becker told UBS, in substance, that the portion of Archegos’s portfolio at UBS was not representative of its portfolio with other counterparties, and that Archegos’s largest positions overall were in large, liquid technology socks, such as Amazon, Google, Microsoft, and Apple. In truth, none of Amazon, Google, Apple, or Microsoft ranked in the top ten of Archegos’s positions.
The thing is, I still have no idea what Hwang was thinking? The SEC’s and Justice Department’s basic story is that he bought a lot of stock to make the stock go up so he’d have paper profits to plow into buying more of the stock. Fine, but, then what? He is a smart guy; if he was doing this to manipulate the price, what was his plan? What was the end game? This strategy, as laid out by the SEC and DOJ, inevitably ends in ruin, and it did. How was he planning to make money?
My fun conspiracy theory — he took out a bunch of profits just before the end and buried them in his backyard — seems ruled out. The indictment is explicit that he plowed every penny into buying more of his stocks, even as the strategy was cracking:
The Archegos Conspirators’ manipulative and fraudulent schemes left Archegos’s portfolio highly vulnerable to price fluctuations in a handful of stocks. In late March 2021, that risk materialized: a decline in the prices of certain stocks in Archegos’s portfolio prompted margin calls; that is, Counterparties required Archegos to provide more cash to support its trading. Because selling Archegos’s positions to raise cash could further deflate the artificial prices of those securities, leading to a downward spiral, BILL HWANG, the defendant, instead directed Archegos’s traders to engage in a desperate buying spree in an attempt to reverse the price decliens of stocks underlying Archegos’s core positions. Acting at HWANG’s direction, the traders used Archegos’s remaining cash and credit, including funds borrowed based on lies and misrepresentations to the Counterparties, to pay for billions of dollars in trades over just a few days.
If I was running a manipulative strategy to artificially inflate the prices of some stocks, and that strategy stopped working and I faced margin calls and a downward spiral in the stocks’ prices, and also I had billions of dollars in cash, I would not plow all of that cash into a desperate buying spree to prop up my artificially inflated stocks! I would, and I cannot emphasize this enough, bury it in my backyard! First I’d fly to a non-extradition country and buy a backyard there! With my billions of dollars! This makes no sense!
That doesn’t mean that it’s wrong, but it is very unsatisfying. The theory here is that Bill Hwang tried to manipulate these stocks up as high as they could go, basically to achieve a high score, knowing that the inevitable end was losing all his money and going to prison. He would not be the first person in finance to do something like that, but it is a weird theory and I don't like it. I don’t have a great alternative for you. “He really liked these stocks and wanted to own a lot of them” is not really satisfying either. But I don’t think it’s ruled out yet.
To contact the author of this story:
Matt Levine at firstname.lastname@example.org
To contact the editor responsible for this story:
Brooke Sample at email@example.com