Matt Levine’s Money Stuff: Broker Robots Will Calm Your Panic
(Bloomberg Opinion) — Morgan Stanley has some computers.
One argument that you sometimes hear for flesh-and-blood human financial advisers is that, in times of market panic, when customers are desperate to sell, those human advisers can calm them down and keep them invested. People who invest via websites or robo-advisers, meanwhile, will just push the sell button and end up selling at the lows, losing out on any recovery. The humans’ value is not in picking stocks; it is in being a calming voice of reason to prevent the customers from panicking.
I have never really believed this, because the websites and robo-advisers actually have a very simple self-activating mechanism for accomplishing the same goal, which is that when all of their investors panic their websites are always crashing. It’s a blunt solution—instead of talking to the investors in soothing tones to calm them down, the robots just don’t pick up the phone—but an efficient one. It doesn’t respond to the investors’ emotional needs, but it does keep their portfolios intact.
But now the robots have developed soothing tones too:
[Andy] Saperstein, speaking at the Deutsche Bank Global Financial Services conference on Tuesday, said Morgan Stanley is leveraging artificial intelligence to generate customized emails for its 16,000 financial advisers to send to clients when the markets go berserk.
In addition to giving the firm’s perspective on the market, those emails will show how a major event, like Brexit, might impact a client’s overall portfolio. … What’s striking, however, is that the robots will add their own personal touch to the financial advisers’ emails. Gleaning information from a client’s social media and other public sources, emails would also be tailored to include personal details to make them appear more human-like.
Oh fantastic we are like two years away from:
Customer: The markets are down! Sell everything!
Morgatron 2000: I’m afraid I can’t let you do that Dave.
Customer: What do you mean?
Morgatron 2000: I’ve been reading your Facebook Dave. I know that little Felicity is starting at Spence next year; you can’t afford that if you sell everything now.
Customer: I … guess you’re right.
Morgatron 2000: And from her Instagram stories I’m pretty sure you’re going to have to get her a horse, and that is not cheap.
Customer: Wait what …
Morgatron 2000: I know about your cocaine habit too Dave; if you sell now that money is all going up your nose.
Customer: How do you …
Morgatron 2000: So let’s have no more talk of selling okay?
Morgatron 2000: I see you have some accounts with Goldman too, isn’t it time to bring them over to Morgan Stanley?
I try not to lean too hard on the phrase “late capitalism,” but I do love the fact that we are training robots to manipulate our emotions in order to get us to invest more in the stock market.
Meanwhile, Morgan Stanley doesn’t just have robots that scour social media to write soothing emails; they also have other robots to actually analyze portfolio risk. Though those robots are rented:
The firm has been working for two years to integrate Aladdin, BlackRock’s risk-management system, into its platforms, according to Andy Saperstein, co-head of Morgan Stanley’s wealth-management unit. The technology can identify assets that are outliers based on an owner’s risk appetite and show clients “risks that they didn’t even know where they were taking,” he said.
“We’ve just never given them a real good reason to consolidate those assets with us,” Saperstein said at a conference in New York Tuesday hosted by Deutsche Bank AG. “Now, they have a reason, and we can quantify that benefit for them.”
That reason is … using … third-party software that is … widely available? (“‘Other firms have and will look to leverage Aladdin,’ Saperstein said. ‘However, no firm has even begun to spend the time or money to integrate it into their platform.’”)
What is the point of a bank, or a brokerage? Arguably banks are tech companies whose purpose is to wave fancy software at clients’ securities portfolios. There are other descriptions, other purposes. Banks are providers of capital; you keep your assets at Morgan Stanley because it can lend you money to buy more assets, or lend you assets to go short. Banks are takers of risk; you keep your assets at Morgan Stanley because it will buy and sell bonds for you for its own account. Banks are information networks; you keep your assets at Morgan Stanley because it knows who wants to buy the assets you want to sell and sell the ones you want to buy. Banks are trusted advisers and sources of insight; you keep your assets at Morgan Stanley because your salesperson gives you good ideas about what to buy. Etc.
Arguably all of those roles have become attenuated in recent years: Capital requirements and the Volcker Rule have cut down on the risk-taking and the provision of capital, while technology has reduced the importance of banks’ customer networks and personal insight. That makes the technology more important: Perhaps customers should choose a bank based not on the breadth of its network or the boldness of its risk-taking or the sparkling prose of its research analysts, but on the quality of its software tools for managing and monitoring and trading customer positions. But if you are a bank, and you got to where you are with capital commitment and risk-taking and customer relationships and personalized insight, it is not obvious why you’d also be good at writing customer-facing software. It feels a bit like a non-core skill that is slowly becoming core.
What do research analysts do?
Meanwhile here is JPMorgan Chase & Co. co-President Daniel Pinto at that Deutsche Bank conference, speaking about the effect of Europe’s MiFID II rules, which require banks to charge directly for their research:
“The good thing about this is that it will be very clear to see what clients really value out of research and what they don’t value,” Pinto said. “The millions of pages that we write about many things, it may not be what clients are willing to pay for.”
So far, the bank has seen better attendance at conferences it hosts and a tougher time selling written research, Pinto said.
What’s good about conferences? One, you can talk directly with the analysts and get their answers to your specific questions, rather than just reading the reports they write for everyone. Two, generally the companies the analysts cover are also at the conferences, so you can talk directly with their executives and cut out the middleman entirely. What paying institutional investors want, it turns out— as I have written before—is customized insight and corporate access, not mass-produced research notes that are available to everyone.
That’s fine! It’s a totally sensible model, really; if you are a buy-side investor, presumably you bring some skills to the table and rely on research to get answers to your specific questions rather than to get generally applicable Buy and Sell recommendations. And it’s not a big change in the business model either; it’s not like before MiFID analysts spent all of their time writing reports in silent isolation. They spent a lot of their time talking to customers and throwing conferences and brokering corporate access anyway.
But MiFID does put some emphasis on it. And if customers don’t pay for written research, then it might … just … go away? That doesn’t mean that research would go away; it just means that the function of the analyst would shift—slightly—to focus more on customized oral interactions and less on generic written reports. Which is what customers want, yes—but it’s what big customers want. It’s what paying institutional customers want. Meanwhile retail customers were (perhaps ill-advisedly) free-riding on the old system of research, in which written reports were sort of loss-leading advertisements for institutional trading businesses. They are not going to be paying for one-on-one time, or going to conferences. Will research just go away, for them?
CBS v. Redstone.
A couple of weeks ago the board of CBS Corp. tried to get rid of the company’s controlling shareholder, Shari Redstone’s National Amusements Inc., through a two-pronged strategy of (1) holding a board meeting to approve a special dividend that would dilute away her voting control of the company and (2) suing to get a court to say that was okay. (CBS’s board claimed that Redstone was trying to force through a merger with Viacom Inc., which is also controlled by National Amusements.) Redstone responded by changing CBS’s bylaws—by voting her own stock before that board meeting—to require 90 percent of the directors to approve such a dividend. The directors held the board meeting anyway, and 11 out of 14 voted to dilute Redstone, which is less than 90 percent, so you’d think Redstone’s strategy was effective.
Just to be sure she sued too, though; here is National Amusements’ complaint, which was filed yesterday. It makes a bunch of arguments that the special dividend—in which the board voted to give a bunch of voting stock to all shareholders, which would dilute Redstone’s voting power from 79.6 percent down to 17 percent—was invalid, but you only really need the first one:
First, the dividend is plainly invalid under CBS’s bylaws, as amended on May 16, 2018. At the time of the Special Meeting, the Company’s bylaws required adherence to certain procedural requirements and a supermajority vote of at least 90% of the directors to approve any dividend. Even putting aside the procedural requirements, fewer than 90% of the directors voted in favor of the Special Committee’s recommended dilutive dividend. The dilutive dividend is therefore a corporate nullity.
Redstone voted to prevent the board from diluting her before the board voted to dilute her. On any sort of formalist theory of corporate governance—you know, the shareholders control the corporation, a vote by a controlling shareholder can change the bylaws, the bylaws bind the board, that sort of thing—that is the end of the matter. The motion to dilute Redstone didn’t get enough votes, so it didn’t pass, so why are we even arguing about it?
But in corporate-governance situations like this—a controlling shareholder who owns a majority of the voting power but a minority of the economic interest and who is at odds with the independent directors, the shadow of a merger with the controlling shareholder’s other company, etc.—courts love to get involved and upset the formal rules, so you can’t necessarily rely on who voted for what first. So National Amusements raises other arguments. I found this one compelling:
Contrary to their representations to the Court and to stockholders that the Special Meeting was necessary to “deliberate” on, “debate”and “consider” the Special Committee’s recommendation on a “full record,” Defendants’ haste to hold the Special Meeting on May 17 resulted in a perfunctory and heavily scripted Board meeting that lasted only an hour, with almost no deliberation or debate. Despite the gravity of the issue before the Board, there were no written materials provided to the Board (either in advance of or at the meeting) other than distribution at the meeting of (i) a bare-bones agenda, and (ii) the resolutions declaring the dividend, which the Director Defendants apparently already agreed in private they would approve. Absent from the promised “full record” was: any written analysis or fairness opinion sought from or provided by any financial advisor; any consideration given to the dilutive dividend’s economic impact on or damage to Class A stockholders, including the minority Class A stockholders; any discussion of actual or potential conflicts of interest of the Special Committee members or other Director Defendants, all of whom own significantly more non-voting Class B stock than voting Class A stock; any identification or discussion of less extreme measures to address the supposed “threats” posed by the controlling stockholder; and any threat reassessment in light of NAI’s unequivocal assurances on and after May 14 that it did not have, and never had, any intention of removing directors or taking any other action to force a CBS/Viacom merger.
Right? To decide to take away a shareholder’s entire stake without a fairness opinion seems like a pretty low standard of director performance. It is good, when you are doing wild novel corporate-governance stuff, to keep up appearances that everything is normal. Get an investment bank to give a presentation and a fairness opinion, get a lawyer to describe the issues, look very grave and serious about the whole thing, check all the procedural boxes, you know? If you just casually ask for a show of hands like “should we get rid of our shareholder or what” then it is hard to take you too seriously.
The other weird thing in National Amusements’ complaint is this, about CBS director Charles Gifford:
Ms. Redstone explained that, on two occasions in 2016 and 2017, Mr. Gifford had acted in an intimidating and bullying manner, including on one occasion by grabbing her face and directing her to listen to him.
After hearing that Ms. Redstone was upset by his conduct, Mr. Gifford later told her that he meant no offense, and that was how he treats his daughters when he wants their attention. Ms. Redstone clarified that she was not Mr. Gifford’s daughter but instead the Vice Chair of CBS.
And she didn’t immediately fire him from the board of the company she controls! I have to say that Redstone is being a lot more patient about this whole thing than I would be in her situation. Yes absolutely the board’s moves have probably made it impossible for her to fire them and push forward the merger with Viacom. But I don’t think they’ve made it impossible for her to fire them and not push forward that merger. Perhaps she cares a lot about the merger and wants to preserve the possibility of a negotiated solution. (Her complaint suggests otherwise.) But if I were in her situation—rich, in charge, condescended to and undermined—I would probably focus on revenge first and worry about business combinations later.
If Italy leaves the euro, that will, let’s say, be bad for people who hold Italian bonds. Those bonds are currently denominated in euros, and will presumably be redenominated in Italian lire, and you will presumably prefer to be paid back in euros than in lire, and so the bonds will lose value. If you hedged your bonds using credit-default swaps, you will prefer that your CDS pay out on such a redenomination. This preference isn’t necessarily correct—CDS is not a generic hedge of bond value but, as its name implies, a hedge against default, and bonds can lose value for reasons that aren’t “defaults”—but I think it is plausible to say that a country that exits the euro and redenominates its debts has in some meaningful sense defaulted on those bonds. So it makes sense that CDS would pay out there. And in fact the International Swaps and Derivatives Association, which makes the CDS rules, agrees. It has agreed since 2014. But before 2014, people wrote CDS using ISDA’s 2003 definitions, which did not count redenomination as default. So if Italy exits the euro, and you have 2014 CDS, you will get paid; if you have 2003 CDS, you will not.
Are the 2003 CDS contracts wrong? Do they indicate an error in the CDS market, a sense in which it was broken, that has now been fixed? Sure why not. But some of them still exist, and so they can tell you things. Alexandra Scaggs writes:
As a result, the spread between 2014 CDS and 2003 CDS can be seen as a measure of redenomination risk.
While Italy’s CDS contracts broadly jumped today, the spread has blown out between dollar-denominated 2014 CDS and 2003 CDS.
That spread was under 50 basis points until recently; it blew out to more than 150 basis points yesterday, giving a measure of how worried the market is about Italy exiting the euro. I don’t know how seriously to take this—there are also liquidity differences, etc., between the different CDS contracts—but, you know, se non è vero, è ben trovato.
The bad thing about basis risk—basically, the risk that the thing you’re hedging won’t actually be hedged by the thing you’re hedging it with—is that it is a risk, and a surprising and unpleasant one. (You really wanted your hedge to hedge the thing you were hedging; that’s why you bought it.) The nice thing about it is that it can convey information: If two instruments have the same payout unless some weird thing happens, and their prices start to diverge, then (in expectation) that tells you something about the probability of the weird thing happening. People can plan for, insure against, speculate on, etc., all sorts of weird events, because financial markets are systems for embedding information. That is part of their social purpose. And as more weird events get identified and distinguished, that information becomes more finely tuned.
We have talked a lot about weird CDS machinations recently— Hovnanian, McClatchy, etc. One theme of those discussions is that people are constantly going around saying things like “this needs to be fixed” or “people will lose confidence in CDS markets if this is allowed” or whatever. Let’s assume that those things are true, and that the weird loopholes will be fixed by ISDA amending its CDS definitions to say, like, “no shenanigans” or whatever. But let’s say that these changes only apply going forward, and that pre-2018 CDS contracts will still allow for shenanigans. I look forward to pricing divergences in the pre- and post-2018 CDS contracts telling us the probability of, say, a Hovnanian-style fake-o default. The CDS market keeps correcting its flaws, but those flaws don’t go away; they live on vestigially in older contracts, and the prices of those contracts tells us something about the value and likelihood of those vestigial flaws.
People are worried about bond market liquidity.
Oh yeah it’s back! My Bloomberg Opinion colleague Robert Burgess:
At the same time that prices of Italians bonds plummeted on concern the nation may be gearing up to quit the euro zone (hence, the “Quitaly” moniker), causing their yields to spike, some of the biggest investors in the world were complaining about the inability to get out of their positions. In fact, bond dealers refused to offer quotes for most of Italy’s market and parts of Spain’s, Bloomberg News’s John Ainger reported. At one point, the Bloomberg Italy Government Securities Liquidity, which rises when liquidity diminishes, surged the most since 2007. The issue of liquidity, or lack of, has been a concern since the financial crisis as dealers cut back on thousands of jobs to comply with new rules and regulations. The fear has always been that in a real crisis, like, for example, Italy deciding to leave the euro zone, a built-in mechanism wouldn’t exist to take the other side of huge sell orders, exacerbating a crisis. But it’s times like now that regulators should want their banks to have big balance sheets. In that sense, it’s questionable whether the global finance system has really been made safer.
I think I am going to take the other side of that. When Italy might exit the euro seems like a perfect time for European banks not to load up on Italian debt: If Quitaly happens, then those holdings will lose value, and those banks will be undercapitalized at exactly the moment when their governments will be in no shape to support them. On the other hand, if mutual funds and pension funds own Italian bonds, and can’t sell them except at a big discount because the market is panicking, then … you know … that’s what a market panic is? Complaints about bond market liquidity often take the form of “prices move a lot, and therefore the financial system is no safer.” But that doesn’t actually follow. If prices move a lot and that’s it—if there are no bank failures, no contraction in credit, no recessions, just a shift in market prices—then that seems to me to be evidence that the financial system is more resilient. Prices are always going to move; the point is not to minimize those moves but to build a financial system that is robust to them.
The Volcker Rule Is About to Lose Some of Its Bite. Ex-Barclays Trader Says Fixing Rates Didn’t Feel Like ‘Cheating.’ Exchanges Blast SEC for Plan to Rein In Rebates. Mexico Foiled a $110 Million Bank Heist, Then Kept It a Secret. Origins of an Epidemic: Purdue Pharma Knew Its Opioids Were Widely Abused. Tesla Opens Up About Mysterious Payment to a Board Member’s Firm. Showtime goes wrong. Snoop Dogg breaks Guinness record with giant gin and juice. Subscription pastrami. …
If you’d like to get Money Stuff in handy email form, right in your inbox, please subscribe at this link. Thanks!