Bank earningsThe two times when it's good to be a modern universal bank are: - When the economy is good, and
- When the market is weird.
There is a rough negative correlation between these things. Often when the economy is good and improving, financial assets move up in a steady and boring way. Banks make money because their loans get paid off, investors want to buy assets and companies raise money to pursue opportunities, but there is a lot of competition and prices are high, so it is hard to make giant windfalls by buying low and selling high. When markets are wild and volatile, banks can make money in trading; volatility makes their trading services — providing liquidity and hedging risk — valuable, and creates opportunities to buy assets cheap from distressed counterparties. But when things are wild and volatile, your loan book gets worse: Borrowers default, and you worry that more of them will default and write down your loans. And so in the first quarter of 2020, as the Covid-19 pandemic was crippling the global economy, banks had pretty mixed results. On the one hand: If all businesses shut down, it's a bad time to be in the business of lending to businesses and consumers. On the other hand: If everything is nuts, it's a good time to be a trader. And so we discussed the fact that banks had great quarters in their trading businesses, even as they were writing down loans and worrying about the economy. The two forces — bad economic conditions, but exciting volatile crisis-condition markets — were obviously connected, and had offsetting effects on different parts of the banks. The first quarter of 2021 is unusual in that (1) the economy is good but (2) everything is still pretty nuts. Economic data is generally good and the stock market is at a record high, but the first quarter also featured … GameStop? Archegos? Plus the boom in initial public offerings from special purpose acquisition companies. Plus non-fungible tokens, which are not actually a business of big universal banks but still feel somehow relevant here. You've got the good aspects of a boring, steadily rising market, plus the good aspects of weird bubbles everywhere, plus the good aspects of wild price swings in particular situations. U.S. bank earnings season kicks off today, with JPMorgan Chase & Co., Goldman Sachs Group Inc. and Wells Fargo & Co. all reporting this morning, and it was arguably the best quarter ever for banks. JPMorgan: JPMorgan Chase & Co.'s dealmakers just helped usher in the firm's best quarter on record, but shares fell as the bank warned that loan demand remains tepid. Investment-banking fees soared 57%, beating analysts' estimates and boosting net income to $14.3 billion, the most JPMorgan has ever earned in a single quarter. A larger-than-expected reserve release added to the windfall as the bank determined it didn't need as much socked away for future loan losses. Government stimulus programs and potentially massive infrastructure spending mean "the economy has the potential to have extremely robust, multiyear growth," Chief Executive Officer Jamie Dimon said in a statement.
JPMorgan reported $14.3 billion of net income on $32.3 billion of net revenue. Of that, $5.2 billion came from reserve releases: JPMorgan marked down loans in previous quarters because the economy was bad and it figured those loans would default, but in fact the economy turned out to be good so now it figures those loans will be fine. But another $9.05 billion came from trading: The bank's traders generated $9.05 billion of revenue in the first quarter, up 25% from a year earlier and exceeding analysts' expectations. That included a 47% increase in equities and a 15% jump in fixed income. Trading revenue remained elevated after a banner year as the coronavirus pandemic roiled markets and sent volatility soaring.
Covid roiled markets, which was good for banks' markets divisions, but also roiled business, which was bad for their loan books. Now it is un-roiling business while leaving markets pretty roiled, so life, for banks, is great. Goldman: Goldman Sachs Group Inc. cashed in on another roaring period for its traders and investment bankers, with revenue and earnings rising to a record. Trading revenue surged 47% to $7.58 billion, fueled by equities. Goldman's dealmakers also contributed to the positive quarter, with investment-banking fees surging 73%. Revenue from equity underwriting quadrupled to $1.57 billion as the red-hot market for special purpose acquisition companies and technology-company initial public offerings gave bankers a bonanza in the first three months of the year. Investors will look for signs of how long IPOs and trading will keep the machine roaring at the investment bank, which has had an extraordinary run through the Covid-19 crisis. In the year since the deadly pandemic first disrupted the global economy, Goldman Sachs has benefited from the resulting surge in market volatility and companies tapping wide-open capital markets.
The one piece of bad news in JPMorgan's results was the weak demand for loans. Things are too good for loans! Why take out a loan when you can get money from a stimulus check or a SPAC or by selling an NFT? The enormous amounts of money sloshing around are in many ways good for banks, but they do reduce demand for banks' core business of lending money. CoinbaseCryptocurrency exchange Coinbase Global Inc. goes public today, and I remain amazed at what a good business it is, compared to regular finance. That's especially noticeable on bank earnings day. Here is the prospectus for Coinbase's direct listing. "$456 Billion Lifetime Trading Volume," it boasts. In approximately eight years of existence,[1] Coinbase users have traded about $456 billion of cryptocurrencies, or a bit less than people trade on an average single day in the U.S. stock market. "$90 Billion Assets on Platform," boasts Coinbase; it also notes that that represents about 11.1% of the total value of all crypto assets. Goldman Sachs's earnings presentation notes that it has about $2.2 trillion of assets under supervision just in its asset management division; JPMorgan's earnings supplement notes that it has almost $3.7 trillion of assets on its balance sheet. Goldman reported net income of $6.8 billion on net revenue of $17.7 billion. JPMorgan reported net income of $14.3 billion on revenue of $32.3 billion. Coinbase has not yet reported its first-quarter earnings (its full-year 2020 net income was $322 million on revenue of $1.3 billion), but last week it pre-announced its estimated results; it expects net income of $730 million to $800 million on revenue of approximately $1.8 billion. It also expects some of the metrics in the last paragraph to get bigger, expecting trading volume of approximately $335 billion and "Assets on Platform of $223 billion, representing 11.3% crypto asset market share." One way to look at that is that the total value of all cryptocurrency in the world is about $2 trillion, and Coinbase took about 0.09% of that value as revenue for itself this quarter. (As of 2020, the total value of all crypto in the world was about $800 billion; that year, Coinbase took about 0.16% for itself as revenue.[2]) Almost $1 out of every $1,000 in the entire crypto market — not $1 out of $1,000 traded, but $1 out of $1,000 of all the crypto that exists — went to Coinbase. Another way to look at it is that Coinbase touched $335 billion of crypto trades in the first quarter, and took about 0.54% of the value of those trades for itself. Just for being the exchange, the platform where cryptocurrencies were traded. I think that if Goldman could collect $1 of every $1,000 of financial assets in the world, it would be incredibly pleased. As it is, the total market capitalization of the MSCI World Index — one index of one big asset class (equities) — is about $56 trillion; Goldman's quarterly revenue was 0.03% of that. But only about $3.7 billion of Goldman's revenue came from its equities division (plus another $1.6 billion from equities underwriting in investment banking); of all the money in equities in the world, Goldman collected considerably less than 0.01%. And Goldman is, you know, doing stuff for that money; it is taking on principal risk to do trades, it is writing and hedging complex derivatives, it is using its own balance sheet to finance client activities, it is underwriting and advising on initial public offerings. Coinbase is mostly just a computer platform where people can trade crypto among themselves. You can see why Coinbase is so valuable, but also why big traditional banks might want to get into the crypto business. As a financial market, it is still not that competitive; there are still huge juicy fees to be earned by people — like Coinbase, but also perhaps like legacy financial institutions — that can figure out how to make it safe and legal and appealing to normal people and big institutions. Right now, if you can do that, you can earn outsized returns. But if Coinbase can do it, that should show the way to other people who want to do it, and those outsized returns are going to attract competition. ArchegosThis keeps going huh: Credit Suisse Group AG unloaded about $2 billion of stocks tied to the Archegos Capital Management blowup in the second such block sale since the bank wrote down the bulk of its exposure in the first quarter. The stock offerings included Discovery Inc. and Iqiyi Inc., adding to some $2.3 billion worth of shares tied to the debacle that the bank sold last week, according to people familiar with the matter. The trades follow a torrent of similar transactions that had already erased about $194 billion in market value as banks from New York to Zurich and Tokyo unwound leveraged equity bets by Bill Hwang's family office. Shares of Credit Suisse fell as the sale adds to evidence that the Archegos collapse could impact the bank beyond the first quarter, when it took a 4.4 billion franc ($4.8 billion) writedown, its worst trading hit in more than a decade. … Credit Suisse's latest sale comes weeks after several rivals dumped their shares to skirt losses. While the firm was one of several global investment banks to facilitate the leveraged bets of Archegos, it was slower than others to unwind the positions and had initially tried to reach some sort of standstill agreement, people familiar with the matter have said.
Back in the 2008 financial crisis, the developer of the Cosmopolitan casino in Las Vegas defaulted on a construction loan from Deutsche Bank AG, and Deutsche Bank just sort of gamely stepped in and said "welp, guess we own a casino now, let's make the best of it." CNBC noted in 2011: Deutsche Bank spared no expense on the Cosmopolitan, whose casino floor is dominated by a three-story glass chandelier that encompasses a cocktail bar. Guests rave about the oversize luxury hotel rooms with wrap-around terraces, which are often sold out. And management has lured popular restaurateurs and retailers.
They didn't sell it until 2014. It always made me happy to think of Deutsche Bank accidentally acquiring a casino through a loan gone wrong and deciding to just brazen it out, get into the casino business, buy chandeliers, entertain clients there, whatever. If you like an asset enough to lend against it, maybe you'll like it enough to own it for years. It would be funny if Credit Suisse just decided that it loved Discovery and Iqiyi, declined to sell, started going to shareholder meetings. But, no. They're selling. Just slower than everyone else. In other Credit Suisse Archegos news: The lender, one of the biggest prime brokers among European banks, is now weighing significant cuts to its prime brokerage arm in coming months, people familiar with the plan have said. It has already been calling clients to change margin requirements in swap agreements -- the derivatives Hwang used for his bets -- so they match the more restrictive terms of other prime-brokerage contracts, people with direct knowledge of the matter said. Specifically, the bank is shifting from static margining to dynamic margining, which may force clients to post more collateral and could reduce the profitability of some trades.
Bloomberg News has explained that change: Static margining sets a fixed amount of collateral that a client has to post to maintain a certain size of position or account. With dynamic margining, a dealer can require more collateral if the underlying risk of the position or account increases due to factors such as volatility or concentration.
The basic trade of a swap is a sort of two-sided arms-length trade: One side is long, one side is short. One side happens to be a big bank and the other side is generally a big (but smaller) hedge fund or family office, so there is some imbalance of power and attention. (The bank will often be the one writing the contract, etc.) But there is a basic formal equality, and a basic symmetry to the trade; as far as the swap contract is concerned you are two counterparties making equal and offsetting bets. And so the collateral terms of a swap might be "you have to post ___% initial margin; then if the position moves against you you post cash for the difference, and as it moves in your favor we post cash for the difference." ("Static margining.") It will be contractual, reasonably fair, approximately symmetrical; the hedge fund gets money if its bet wins and loses money if it loses. It would be unsporting if the bank did otherwise. The basic trade of prime brokerage is a bit different: The bank is providing the money to make the hedge fund run, and it is clearly a service provider that wants to keep itself safe. There is not a formal two-sided bet, and there is less equality and symmetry between the parties. The hedge fund needs its prime broker more than the prime broker needs any particular hedge fund; the hedge fund gets most of the profits of its positions while the prime broker just gets some fees and interest. And so the collateral terms of a prime brokerage agreement might be more like "you have to post ___% collateral for your position for now, but if we get nervous and change our mind we can ask for more collateral and you have to post it." ("Dynamic margining.") It is understood that the prime broker gets to protect itself, that it can reassess the risks that the hedge fund poses to itself and demand that the hedge fund compensate it for that risk. If you are providing "synthetic prime brokerage" through swaps you might run into a little bit of confusion over this. You provide leverage to a big customer via swap, you learn that his positions are big and concentrated and starting to get more volatile, so you decide, hey, we need more collateral. As a prime broker, you would just call him up and demand more collateral: "Your concentration and volatility are up, wire us some money." As a swap counterparty, though, it might not work that way: You have a contract, you have a bet with him, and you don't get to change the terms of the bet in the middle of the bet just because it has become more risky for you. Credit Suisse would like to change that expectation, for next time. Citi's oopsieThis is from last week but remains mysterious: A lawyer for Citigroup Inc. told a federal judge on Friday he was aware of another big bank that had recently made an even larger payment error than its own $900 million transfer to Revlon Inc. lenders. The lawyer, Neal Katyal, dropped the bombshell at a hearing in which Citibank urged the judge to extend a freeze on more than half a billion dollars it wired asset managers for Revlon creditors last summer -- an epic back office blunder that led to a closely watched trial. Some of them returned the money, and Citibank unsuccessfully sued 10 others for the $504 million they refused to give back. The judge froze the funds during the dispute. Katyal never identified the other bank after alluding to it on Friday. The hearing centered on whether the freeze would continue. The judge didn't rule.
I have to say, I totally believe this. When it first came out that Citi had accidentally wired $900 million to some Revlon lenders, my reaction was 100% "yeah this happens all the time and is no big deal." I wrote: "If you are a big enough bank and you do enough nine-digit payments every day, eventually you're going to do a wrong one." Usually people give the money back. We live in a society, and if a big bank sends a big pile of money to big investors who don't have a legal claim to it, they will just send it back. What's weird here is that the investors had a legal claim to it: Revlon owed them money, they were in a fight with Revlon about the terms of that debt, and when Citi accidentally sent them the money they could say, with almost straight faces, "aha yes we have gotten back the money we deserve." I was skeptical of that argument, but it worked; the lenders convinced a judge that they should be able to keep the money. Now Citi is appealing, and … I still kind of think Citi should win? Here's Citi's brief for the hearing last week, arguing that the lenders shouldn't be able to spend the money yet because Citi is going to win on appeal. In particular, the lenders got to keep the money based on the legal doctrine of "discharge for value": Because the lenders really were owed money (by Revlon), and they thought that Citi (the administrative agent on Revlon's loan) was intentionally paying off that debt, they get to keep the money. But Citi argues (pages 10 and 11) that this doctrine only applies if the lenders were owed money and it was due: If someone owes you money and you get paid on the due date for that loan, you can reasonably think "ah yes, my repayment," and keep it; if someone owes you money in three years and you get paid today, you should probably assume something has gone wrong. Anyway here's one more problem for Citi: For Citibank, the stakes may be even greater than the sum it sued to recover. Some of the money managers that gave the cash back did so provided that they be treated like the others and that Citibank return the funds to them if the bank lost the case, according to people with knowledge of the matter.
Citi mistakenly sent $900 million to lenders, then asked for it back; it got about $400 million back, but is still chasing the other $500 million. But some of the $400 million it got back came with strings attached: The lenders who gave it back insisted on a condition that, if Citi wasn't entitled to get the money back, it would have to return it to them. "If we were mistaken in correcting your mistaken transfer of money, you have to correct our mistake by transferring the money back to us." Good lawyering! Those lenders want to be polite to Citi, but not at the expense of looking dumb themselves: If they were entitled to keep the money, they want it back. Elsewhere, Charles Schwab Corp. sent a woman $1.2 million by accident, and she kept it, and now she's in jail. If someone sends you money by mistake and you decide to keep it, it helps a lot to be a big hedge fund instead of a regular person. Things happenBernard Madoff, Mastermind of Giant Ponzi Scheme, Dies at 82. Credit Suisse Identifies $2.3 Billion at Risk in Greensill Funds. O'Melveny & Myers Discusses the Legal Challenges of NFTs. BNY Mellon Opens $4 Trillion Repo Niche to Holders of China Debt. Chinese Students Pay Agents $12,000 for Shot at Wall Street. "The practice of sharing the cost to get a trapped ship moving derives from Rhodian Sea Law, a body of rules that governed sea trade and navigation during the Byzantine Empire starting in the seventh century." 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! [1] Coinbase was founded in 2012, and the prospectus's numbers are as of Dec. 31, 2020. [2] This is not apples-to-apples as the value of crypto rose dramatically over the course of 2020, particularly in the fourth quarter. The total value of crypto was $782 billion as of the fourth quarter, but the average over the year was closer to half that; see page 96 of the prospectus. So Coinbase's revenues probably ran closer to 0.08% per quarter than 0.16% per year. | 
Post a Comment