Bank Strategy and an AIG Breakup
Today Deutsche Bank announced the details of "Strategy 2020," its plan to get smaller and safer and more profitable and more boring. Among the things Deutsche Bank will jettison in pursuit of that goal:
- "Argentina, Chile, Mexico, Peru, Uruguay, Denmark, Finland, Norway, Malta, and New Zealand";
- 90 billion euros of risk-weighted assets
- "market making in uncleared Credit Default Swaps, certain Legacy rates products, agency Residential Mortgage-Backed Securities trading, and higher risk-weight securitised trading";
- about half of its Global Markets and Corporate & Investment Banking clients, "especially in higher operating risk countries, given that approximately 30% of clients produce 80% of the revenues in these business divisions";
- its dividend; and
- about 35,000 employees.
It all sounds pretty decisive, though it comes with this sort of grim caveat from co-CEO John Cryan:
"Deutsche Bank does not have a strategy problem,” Cryan told reporters in Frankfurt. “We know exactly where we want to go. However, Deutsche Bank has faced a grave problem for many years in implementing this strategy. In the last two decades, many strategies and targets were announced, but rarely were they consequently realized."
So is he ... saying that they're kidding about this one too? Deutsche also announced earnings; its trading revenue was pretty good, but the net result was a loss of 6 billion euros driven mostly by goodwill impairments. If you're abandoning so much of investment banking, I suppose your investment bank will be worth less. Meanwhile, although Deutsche plans to "increase levels of individual accountability," "prioritise investments in its Know-Your-Client (KYC) and Anti-Money-Laundering infrastructure," and get out of "relationships and locations with unacceptable risks," the past still haunts it, and it is expected to pay about $200 million to partially settle U.S. sanctions charges.
Too big to succeed.
Carl Icahn's letter to AIG telling it to break up includes this argument:
“Because of AIG’s size and interconnectedness” the Financial Stability Oversight Council (“FSOC”) has deemed AIG a non-bank SIFI, subjecting the company to Federal Reserve oversight and increased capital requirements. We believe you must acknowledge that enhanced regulation is intended to be a tax on size, designed to approximate the cost that large companies impose on the financial system. The regulators have made clear that the best outcome is for SIFI’s to shrink and “reduce their systemic footprint.” If nothing is done, returns and AIG’s competitive position will continue to suffer as the SIFI regulation, including its costs and capital requirements, is fully implemented.
AIG disagrees, though my Bloomberg View colleague Brooke Sutherland thinks that "a breakup has got to be worth a serious look." One thing you can say about, for instance, calls to break up JPMorgan, is that the very real costs of SIFI (systemically important financial institution) regulation are probably offset by the funding advantages of being a SIFI. Like, if you are a too-big-to-fail bank, you'll have to deal with expensive regulation, but being too big to fail also means you can borrow cheaply. That is kind of the deal. It seems pretty true of JPMorgan. It is less obvious to me that AIG, in 2015, is actually -- politically -- too big to fail, or that investors think it is, or that it has that funding advantage.
Don't do this.
Two things that we talk about sometimes are:
- Many bond markets are illiquid and lack price transparency, so brokers can -- and sometimes do -- sometimes maybe illegally -- take advantage of customers' ignorance of market prices to add huge markups; and
- electronic front-running by evil high frequency traders, or whatever.
Here is an absolutely bizarre Securities and Exchange Commission case against two former brokers at now-defunct Rochdale Securities who kind of did the bond thing, but in stocks:
The SEC’s Enforcement Division alleges that the former co-head of equities trading at Rochdale Securities, Hal Tunick, and his subordinate, Patrick Burke, defrauded customers by using their order information to advise two longtime customers to trade ahead of these orders. Once those favored customers purchased or sold short the shares, Tunick and Burke arranged for them to unload their positions to the customers who had placed the original orders. The favored customers profited from these trades while the defrauded customers generally received worse prices than they would have if their orders had been routed directly to the market.
Tunick and Burke seem to have been in it for the commissions:
For example, on September 30, 2010, Tunick received a customer order to purchase 20,300 shares of ABC Co. (“ABC”). For no purpose other than to generate commissions, Tunick then instructed his Customer to purchase 10,000 shares of ABC through an account at a third-party broker-dealer and then to submit an order to sell these shares through his account at Rochdale. Tunick’s Customer purchased these shares at approximately $54.89 per share. Tunick’s Customer thereafter submitted an order to sell the 10,000 shares through his account at Rochdale at a price of $55.03 per share. Market data reflects that at this time (approximately 10:00 a.m.), offers (sale orders) for more than 5,000 shares of ABC were available in the open market at an average price of $54.98 per share. Despite this, Tunick instructed his Customer to increase his order to sell at $55.03 by two cents, to $55.05 per share, which Tunick crossed with the pending customer order to purchase shares of ABC.
So, one, that is just terrible, super obviously super illegal. (Tunick and Burke settled for $125,000 and $50,000 of penalties, respectively, and bars from the securities industry, "without admitting or denying the findings.") Two, it's a reminder that the human stock traders who are being supplanted by high-frequency algorithms weren't always angels either. In the example above, the front-running customer bought at $54.89, then sold to the other customer at a 16-cent markup, an order of magnitude bigger than the biggest markup an algorithm has ever dreamed of.
And, three, this really shouldn't have worked. These aren't mortgage bonds! There's a consolidated tape. These trades had to print publicly. The customer can see what the price was when he sent the order, and compare it to his execution. Rochdale's compliance, and Finra and the SEC, can also check. Maybe you can get away with charging whatever you want for illiquid bonds, but these are stocks, there is a market price, come on.
McKinsey & Co. is predicting that a lot of banks will more or less automate themselves:
“Banks taking the all-in route will end up with business models and economics that look very different from those they have today,” according to the report. “On the sales and trading front, digital channels will become the default for transacting with clients. Coverage models will shift radically as banks ruthlessly examine every front-office routine and manual activity for its automation potential.”
I feel like consultants love to put out underminey reports about how bankers are doomed, but still, it seems plausible. Investment banks are basically people businesses, but not in the way that, you know, home health care is a people business. The people at investment banks tend to be rewarded for rationality and profit maximization, not empathy or manual dexterity. Computers are pretty good at rationality and profit maximization, and they wouldn't have cooked up that Rochdale scheme. Though: "High-Speed Traders Slow to Break Into Interest-Rate Swaps Market."
Elsewhere, "CME Group Inc. and Dwolla Corp., a digital-payment network, have struck a deal that will allow the exchange’s users to send and receive collateral payments in real time throughout the trading day." The move to real-time cash settlement seems even more inevitable than the move to automation, though I might not have guessed Dwolla? Good for them. Whatever happened to everything being on the blockchain?
The story of Valeant and its network of affiliated pharmacies keeps getting stranger. Bloomberg News visited with R&O Pharmacy, the small California pharmacy whose lawsuit against Valeant set this all off, but R&O has no answers, only questions:
Less than a year ago, 64-year-old pharmacist Russell Reitz agreed to sell his business there to a Delaware-registered company for $350,000. R&O soon became the victim of widespread fraud, Reitz’s lawyer alleged in documents filed in a billing suit against him in state court in California.
Even before the sale agreement was executed, other pharmacies began using an R&O identification number to bill for prescriptions that R&O hadn’t filled, sometimes for drugs the store didn’t stock, according to the court documents. Reitz got requests to hand over the money, and personal visits -- not only from the one-man company buying R&O, but also from representatives of a firm called Philidor RX Services, whose links to the buyer Reitz struggled to understand.
The worry here is that someone -- Valeant? Philidor? persons unknown? -- might have been using R&O's identity nefariously, either for outright fraud or to get around pharmacy licensing requirements. Business Insider has more on R&O's lawsuit, including copies of the Purchase and Sale Agreement and Management Services Agreement that R&O signed with Isolani.
Elsewhere, Theranos planned "to raise upwards of $200 million in new funding" shortly before the Wall Street Journal reported that its technology may not work.
The Consumer Financial Protection Bureau and the Justice Department decided that Ally Financial made auto lending decisions that discriminated against minority borrowers. They are pretty sure about this, "even though the lender is prohibited by law from collecting data on the race or ethnicity of its borrowers," so there's no way to know if any particular borrower is actually a member of a racial minority. But, you know, statistically, you can be sure enough, and Ally settled the case in 2013, agreeing to "pay borrowers at least $80 million" without admitting wrongdoing. But now it has to send out checks. And ... it still doesn't know the race of any particular borrower. So it can't know who it discriminated against. So it has to guess:
To find minority borrowers, the U.S. Consumer Financial Protection Bureau and Justice Department opted to essentially make educated guesses using an algorithm that assigns probabilities to whether borrowers are minorities based on their last names and locations.
They have been sending letters in recent months to Ally customers they believe were overcharged based on the results. One version of the letter, sent to borrowers that the CFPB estimates have at least a 95% probability of being minorities, tells recipients that they qualify for a payment and don’t need to do anything more. It asks borrowers to write back if they aren’t minorities.
There is a good general lesson in epistemology here. You can know, with a very high degree of confidence, that something is happening in a statistical aggregate, without knowing anything about any individual case.
Donald Trump can't stop winning.
Apparently Donald Trump is America’s greatest stock investor.
Or that would appear to be the inference, at least, from the claim he made in his new book, “Crippled America.” He writes that 40 of his 45 stock purchases “rose substantially in a short period of time.” In the parlance of money management, that’d give the Republican presidential candidate a positive strike rate of 89 percent. Wow. If factual, it’d be better than the rate posted by any institutional investor in America between April and September.
Why not, why not.
People are worried about the debt ceiling.
I confess that part of why I added this as a daily section was that, unlike bond market liquidity, it had a clear end date. The government was going to run out of money in November! Something had to happen before then! Something did. I hereby declare that people are no longer worried about the debt ceiling. Meet me back here in two years, when I will once again say unto you: Debita impendiorum maximorum vendenda sunt.
People are worried about bond market liquidity.
"Shining a Spotlight on Bond-Market Liquidity," is the headline here, because I guess it hasn't gotten enough attention, despite my tireless efforts:
The International Organization of Securities Commissions surveyed about 100 banks, asset managers, trading platforms and regulators from August to October, according to people familiar with the work, asking questions about corporate-bond-market conditions before and after the crisis. Iosco also held two industry roundtables earlier this year in Frankfurt and Toronto on the subject.
I bet those were fun.
I wrote about Goldman Sachs, the New York Fed and the revolving door.
Pfizer and Allergan Begin Merger Talks. Will This Online Lender’s Risky Business Model Hold Up? Former Goldman Executive Winkelried to Join TPG. Facebook's Zuckerberg Chided by Judge Over Director Pay. Lead Defendant in U.S. Libor Trial Challenged Over Inaction. Who’s better at buying companies: PE firms or strategic buyers? Inside the Old Psych Hospital Reborn As a Home for Money Managers. Mary Jo White: "Building a Dynamic Framework for Offering Reform." Twitter should charge by the letter, exponentially. "Captain Math will get you high tonight." The Dogs of Wall Street. Freighter cruises. Exhausted owl. Loose blimp.
If you'd like to get Money Stuff in handy e-mail form, right in your inbox, please subscribe at this link. Thanks!
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
To contact the author of this story:
Matt Levine at email@example.com
To contact the editor responsible for this story:
Zara Kessler at firstname.lastname@example.org