March 16Banks are so strange. You might think that the way a bank would work would be that, the more money it has, the more money it can use to make loans and trade securities. If people flock to a bank to give it deposits, then it will have more money to lend. Or if a bank makes savvy interest-rate derivatives bets, and those bets pay off, then it will have more money that it can use to buy stocks and bonds. That is how businesses tend to work: If they do stuff that brings in money then they can use the money to do more stuff. This March, financial markets seized up in a whole variety of ways, and when financial markets seize up there is usually some component of "banks couldn't do stuff" involved. At the Wall Street Journal, Justin Baer has a terrific account of "The Day Coronavirus Nearly Broke the Financial Markets," examining some of the ways that financial markets seized up on March 16. Here is a mind-boggling one featuring Vikram Rao, head of debt trading at Capital Group Cos.: Mr. Rao, who was working remotely that Monday, walked down the 20 steps to his home office at 4:30 a.m. to discover the debt markets were already in disarray. He started calling the senior Wall Street executives he knew at many of the big banks. Executives told him that Sunday's emergency Fed rate cut had swung a swath of interest-rate swap contracts in banks' favor. Companies had locked in superlow interest rates on future debt sales over the past year. But when rates fell even further, the companies suddenly owed additional collateral. On that Monday, banks had to account for all that new collateral as assets on their books. So when Mr. Rao called senior executives for an explanation on why they wouldn't trade, they had the same refrain: There was no room to buy bonds and other assets and still remain in compliance with tougher guidelines imposed by regulators after the previous financial crisis. In other words, capital rules intended to make the financial system safer were, at least in this instance, draining liquidity from the markets. One senior bank executive leveled with him: "We can't bid on anything that adds to the balance sheet right now."
Yes, hrm hrm hrm, capital rules, collateral, but look at what is actually going on there. The banks had some interest-rate trades. Those trades did well. They made money. The counterparties of those trades had to hand money (really Treasury bonds) over to the banks. Money (Treasuries) came in the door; the banks got richer because their correct trades had worked out successfully. And then Rao went to the banks and was like "hey I have some bonds, want to buy them?" And the banks were like "oh no we can't, not today, today is bad for us because we got all this money, which makes it impossible for us to buy any bonds. Our trades worked out so well that we can't do more trades." That is not how other businesses work! Usually when you do a good thing that works out and brings in money, you try to do more things. Banks are like "sorry there is a fixed number of things we can do, we did a good thing today so now we're done with things." That is not a technical description or anything, and there are nuances here. For instance I assume that banks were not taking huge directional positions and had losses on the other side; this was less "a good trade made money and increased our assets and equity" and more "a market-making trade got grossed up on both sides, increasing our assets but not so much our equity." Still this is a real thing! We talked about a related story back in April: As markets seized up, people did flock to banks to give them deposits (because cash in the bank seemed safer than the alternatives), and the banks took the deposits, and then the banks might have had to cut back on activity because, paradoxically, they had too much money. The Federal Reserve went and changed its regulations so that the inflow of deposits wouldn't choke off lending. It announced: Financial institutions are receiving significant inflows of customer deposits along with increased reserve levels. The regulatory restrictions that accompany this balance sheet growth may constrain the firms' ability to continue to serve as financial intermediaries and to provide credit to households and businesses.
Really, that says: Banks have gotten so many new deposits that they can't lend money anymore, so we are going to change the rules to let them lend money even though they have lots of deposits. That's not how people normally think about banking! George Bailey didn't have to call in loans because people kept putting money in his bank! Quite the opposite! But the way modern banking works is, sometimes, that the more deposits banks get, the less they can lend. There are technical explanations for this but it seems more satisfying to leave it as a mystery. (Oh, fine. The simplest quasi-technical explanation is that people don't really give banks money, they lend banks money. Deposits are liabilities of the bank, and even interest-rate-derivatives collateral is in a sense only temporarily in the banks' custody. Banks can borrow a lot of money without meaning to, as it were; people can show up at a bank and deposit money in a way that they couldn't show up at your house and lend you money. The principal constraint on modern banks is the leverage ratio: Basically, regulators do not want banks to borrow too much, because there is a history of borrowing too much being bad for banks, so there is a rule limiting how much borrowed money they can have for every dollar of shareholders' equity. So when banks unexpectedly borrow a whole lot—because people rush to the banks to lend them money—they run up against that limit and have to cut back on their activities, or at least not expand them, to stay within the rules. The change that the Fed announced in April—there was an additional adjustment last week—was essentially that if people give a bank a ton of money and the bank parks it in Treasury bonds and Fed reserve accounts, the Fed won't count that against the bank's borrowing limits, so it doesn't have to restrict the rest of its business.) This was my favorite part of Baer's story but I recommend the whole thing; it gives a great sense of the texture of financial markets and how chaotic March 16 was. Here the finance officer of a school district who had planned to sell bonds the next day: When Citi advised putting the deal on hold for a while, Mr. Hammel huddled with his chief facilities officer and the two men did the math. Without the cash infusion the district had expected from the bonds, construction would halt in July. "It's one of those days where you just go home and say 'It's either going to be beer or wine,' " Mr. Hammel said.
Or here's the story of Allianz Global Investors' Structured Alpha funds: Allianz's Structured Alpha funds had been a big seller of insurance against a market selloff in the short term and a buyer over the longer term. The strategy produced a steady income, as the fund collected premiums from investors hedging against a downturn. The funds might lose money for a month during a selloff as they restructured those short-term trades, Greg Tournant, the funds' portfolio manager, said during a May 2016 marketing video, but over time they'd make money. "We are acting like an insurance company, collecting premiums," Mr. Tournant said. "When there is a catastrophic event, we might have to pay—very much like an insurance company. The positions we buy to protect ourselves from those catastrophic shocks—you could label those as reinsurance." When the big storm arrived in March, though, the strategy didn't work.
Well it worked. It did what it was supposed to do. On March 25, Allianz announced that two of those funds would be liquidated, and that one was down 97%, which is bad, obviously, but their strategy was selling insurance against catastrophe, the catastrophe arrived and the insurance paid out. The reinsurance didn't so much, but you cannot get too hung up in the fine details here. A strategy of "we will collect insurance premiums from paranoid people who think the market will crash" was quite lucrative for a long time, because we have had a decade of rising markets and people worrying a lot about the next financial crisis. If you let them pay you to assuage their worries, you could make a nice living. But you knew the business you were in. Bond tradingI don't know, man: It turns out the corporate bond market still needs traders. The algorithms that dealers use to buy and sell bonds with their customers failed in March at the height of extreme volatility from the coronavirus pandemic, according to investors and price data. The nimble analysis of flesh-and-blood traders was suddenly needed to price bonds, edging out machines that normally can trade large portions of the market without any human input. … Even as digital platforms set record volumes in the first quarter, market watchers said the bots failed a test when Treasuries and credit spreads were so disorderly. "Good old-fashioned blocking and tackling is still very much a part of the business,"said Chris Coccoluto, head of investment-grade bond trading at Manulife Investment Management. "It was almost nice to see in a way you had to rely on your relationships."
I mean the thing about bond trading in mid-March is that it was … bad? The algorithms backed away, the humans stepped up, and markets were volatile and scary and bid-ask spreads were wide. I suppose that is a sort of triumph for the humans—"we were there to trade bonds badly when the computers were not"—but it's not a great triumph. A clean test would be, like, on an ordinary day someone spilled a lot of Coke on a lot of computer servers and the humans had to step in to trade bonds, which they did in larger size and at tighter bid-ask spreads than the computers could have. This is not that. This is like, the bond market went haywire, the computers were like "this seems terrible, we'll be at our vacation houses," the humans stepped up and, yep, the computers were right and the market was terrible. (To be fair a lot of the human traders made fortunes trading those markets, but that too is not a pure win: It means that they provided expensive liquidity.) One argument that I have made a lot since March is that when there is an economic crisis, you should expect volatility to increase and bid-ask spreads to widen, not because of some complicated story involving algorithms but because that is what happens when things are bad and uncertain. If no one knows how bad things are or when the economy will reopen, prices will move around a lot, and people who want to sell will demand high prices while people who are willing to buy will offer low prices. In a market intermediated by computers this will look like "the computers have backed away from providing liquidity because computers are bad," but the same exact thing happened in totally human-driven markets like startup funding and oil M&A. Still let's concede two things to the humans. Here's one: JPMorgan Chase & Co. found that algorithms can in fact learn from humans in tempestuous markets. An algorithm it trained to recommend trades based on human market commentary significantly outperformed those based on only market observable features in March, strategists led by Joshua Younger said in a report dated April 23.
So maybe the algorithms can provide the high-level executive function while the humans do the grunt analytical work to inform the algos' decisions, why not. But here's the other: While humans were able to spot the new patterns quickly, the bots couldn't adapt because algorithms are built on historical data, said Chris White, founder of advisory firm ViableMkts LLC.
This is a theory I have heard a lot for a long time: Algorithms are trained on historical data, while humans take a broader view of financial markets and are better at responding to changed conditions. I admit that I have been a skeptic. Actual humans in trading seats have also been trained on historical data; it's just that usually they have less historical data than the algorithms. If you've been trading bonds for five years, a good long Wall Street career really, then you've seen about five years of market data. You can just feed 50 years of data into a computer algorithm, so the algo really may have a better historical sense than any living human trader. Lots of humans on trading desks today have never seen high inflation or a rising-interest-rate environment or, until recently, a financial crisis; every computer, in a sense, has. No computer had seen a coronavirus-related economic shutdown until a few months ago, and no human had either, but the humans might genuinely have had better intuitions for dealing with that novelty. "Hmm, what will my family do if we can't leave the house, and how can I extrapolate that behavior to the broader economy?" is a thought process that a human could have when faced with a novel situation; it's harder for a computer. (Computers never leave the house.) A lot of the work in financial markets is essentially pattern matching, figuring out if this thing will turn out like that thing or more like the other thing, and computers can see and remember more patterns than most humans. But when the pattern is totally new, the humans may have an advantage. Everything is securities fraudOne point that I often make about the U.S. legal system is that lying about financial matters is quite broadly and severely criminalized—basically any sort of dishonesty having anything to do with money is arguably wire fraud—but lying about politics is common, legal, and frankly encouraged. If you sell knickknacks on the internet and you misrepresent their dimensions, that is probably a federal felony, but if you run for president on a platform of wild lies and dangerous conspiracy theories then you are just participating in a long and glorious tradition and you'll probably win. Another point that I often make is that people dislike this situation and try to fix it by pretending that political lies are business lies. Political lies are protected by the First Amendment, business lies are unprotected fraud, so if you can shoehorn a political lie into a business-lie theory you can prosecute it. So people got mad at Exxon Mobil Corp. for participating in public debate about climate change in a way that was arguably dishonest, and fixed it by suing Exxon Mobil for securities fraud: If a public company that makes oil is lying about climate change, the theory goes, surely it is deceiving its shareholders about the future prospects for oil. (This theory did not win in court.) Or Elizabeth Warren is an expert in this approach, and she has asked the U.S. Securities and Exchange Commission to investigate companies for securities fraud for lobbying against regulations, and for that matter to investigate Donald Trump for trying to start a war. (Is trying to start a war securities fraud? Everything is securities fraud!) Still you could do better. Here's a law review article by Tyler Yeargain called "Fake Polls, Real Consequences: The Rise of Fake Polls and the Case for Criminal Liability." A lot of the argument is that people publish fake political polls, these polls are bad for basic political reasons (deceive voters, undermine democracy, hurt confidence in our government, etc.), and something should be done about it. Fine, yes, but First Amendment etc. But the other part of the argument is that people publish fake political polls to influence political prediction markets, which means that they might be wire fraud, or commodities fraud, since prediction-market contracts are arguably commodity futures. If someone publishes a fake poll to manipulate prediction markets, then that's a business lie—wire fraud, commodity fraud, securities fraud, lying to make money—even though it's about a political topic. So you can sweep political lies into the wire-fraud framework, and prosecute it. "And," writes Yeargain, "to the extent that the elements of either fraud charge is too difficult to prove, prosecutors could always fall back on conspiracy charges." Why stop at polls? What if you bet on a candidate on a political prediction market and then tweet that he is a really good candidate with great ideas for fixing America's problems? What if he is in fact only a mediocre candidate with vague and stupid ideas for fixing the problems? What if you say that his opponent has been a disaster in office, when in fact she has done a roughly average job? Can prosecutors go after you for wire fraud for lying about a political candidate? That'll be … new. It is easier to prove that polling claims are factually false, but, you know, politicians make lots of factually false claims and you can try your luck asking a jury if they're lies. I suppose these prosecutions will still require the element that the lies were designed to make money in predictions markets, but "prosecutors could always fall back on conspiracy charges." What I like (?) about this theory is how big the asserted harms are and how trivial the connection to finance is. The harms of fake polling are about undermining democracy and trust in institutions, and about dishonestly putting people in positions of immense power, but legal political betting in the U.S. is a tiny business. "Given the CFTC-mandated restrictions on the markets," writes Yeargain, "the fraudsters likely made a collective profit of no greater than a few thousand dollars." In other words the real reason to manipulate polls is not to make a few thousand bucks in PredictIt, and the real reason to prosecute fake polls is not that a few people will lose money on PredictIt. None of this is the point, it's all a pretext; the actual point is that people lie about politics to influence politics, and that other people want to prosecute those liars to prevent them from influencing politics. But the legal formula that you have to recite involves betting markets. "There is no place for dishonesty in American politics," the theory roughly goes, "because dishonesty in politics could undermine the integrity of betting markets." Okay! I think I'm with the VP hereLook it's not great or anything: A workaholic Wall Street banker blasted his bleary-eyed underlings last week for not promptly responding to an email at 3 a.m., according to an early-morning email exchange that has since gone viral on social media. An unidentified junior banker at PJT Partners — a boutique mergers-and-acquisitions adviser headed by its go-getter founder Paul J. Taubman — apologized after being prodded by his insomniac boss in the wee hours of May 13, according to a screenshot of the email chain. "We're working on these now (and have been), but we didn't receive them until 3 am our time and were already in bed," the sleepy junior banker wrote to the boss at 6:38 a.m. last Wednesday. The email got an unsympathetic reply, according to the screenshot, which was first posted with the names redacted by Instagram user @arbitrageandy. "To be clear I am working ET time and I expect my junior team to notify me when they log off," the boss responded. "Not an excuse 3 am… I sleep [an] average of 5 hours or less. Expect the same or more from my junior team especially on live deals. Please turn around my comments and work as fast as you can."
But six months ago if you were working on a bid for a live deal, you'd be doing it in the office, and it would be all-hands-on-deck, and if you wandered off just because it was 3 a.m. someone would notice and, probably, yell at you. Now you're doing it on your couch, but it's still a live deal, and if you fall asleep someone will yell at you over email. Pleasingly PJT's response to this story is along the lines of "yeah whaddaya want investment banking is hard": "While we do not comment on personnel matters, maintaining a culture of respect, collaboration and a commitment to excellence is our highest priority at PJT Partners," said a spokesperson. "We remain committed to supporting each other and serving our clients during this time."
There are a lot of stories recently about how banks are adapting to working from home forever, and the basic theme of a lot of them is that it's fine, it all works great, there's no reason to come to the office. Here's one from the Financial Times today: "We used to think of working from home as spending time with the dog," said the head of currency trading at an investment bank in New York. "It had a bad reputation but it turns out that was completely wrong — I can't believe the hours I've been working since my bedroom has become my office." The abrupt upheaval has forced traders to challenge their own assumptions. It is also prompting senior executives in London and on Wall Street to rethink not just the working patterns for staff, but what the trading floor of the future will look like. "If you'd asked me in January, I'd have wondered if we could run all our businesses [working from home], but now I can say the answer is definitely yes we can," said Robert Karofsky, co-president and head of markets at UBS's investment bank.
And that is true, apparently you can put a big monitor and a turret phone anywhere. But one big element of working in the office has always been the transmission of culture, both good and bad, and that's much harder to accomplish over Zoom and email. If you've been trading currencies at a big bank for a decade, the culture has been transmitted; you know how to do your job, both in its technical aspects and also the softer cultural ones, and it's probably easier now to do it from home. But if you're a junior analyst or a new intern, you are still being formed as an investment banker, and if you don't absorb the culture at first hand you might never become a "real" investment banker. Although that PJT banker is doing what he can I guess. Things happenBig Banks Plan Staffing Limits, Shift to Suburbs After Lockdown. In Fink We Trust: BlackRock Is Now 'Fourth Branch of Government.' BlackRock Softens Stance in Argentina Restructuring Talks. The Relentless Ambition of BlackRock's Aladdin. Natixis' Equity Derivatives Losses Soar to 250 Million Euros. Venezuela sues Bank of England over refusal to release gold. SBA Under Spotlight as Frustrations With Aid Programs Grow. A $150 Billion Pile of Frozen Loans Starts to Worry U.S. Banks. Biggest US shopping complex shows threat to mortgage-backed bonds. A Weakness in Giant CLO Market Exposed in Tiny Distressed Deal. Hertz Lenders Brace for Losses From Formerly Triple-A Bonds. Crypto hedge funds struggle to recover from 'bloodbath.' The King of Germany Will Accept Your Bank Deposits Now. Kendall Jenner Settles Fraud Suit Over Fyre Fest Instagram Promo. 'Weasels have eaten our phone system' — Citi customers trying to call customer service were met with the same odd prompt. "People just want to believe in drunken elephants." Asness v. Taleb. NASA scientists detect evidence of parallel universe where time runs backward. 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! | 
Post a Comment