Money Laundering and Quant Code
If you are a bank, and you move money for criminals, then you will get in trouble. There is something a little bit weird about this: Airlines don't generally get in trouble for flying criminals around, and supermarkets don't get in trouble for selling them food, but banks get in trouble for moving their money. It's not that weird, as a legal matter, when you consider that banks are not purely private enterprises but have some public-utility functions. The U.S. financial system is in part a tool of U.S. law enforcement, and U.S. banks (and foreign banks that use the U.S. financial system) are deputized to help out law enforcement, and punished if they don't.
What is maybe weirder is that, when banks fail to catch criminals, they don't just get in the traditional forms of trouble -- paying big fines -- but also come in for the moral condemnation that goes with the underlying crime. "In this case," Matt Taibbi once wrote about HSBC, "the bank literally got away with murder -- well, aiding and abetting it, anyway." It paid $1.9 billion in fines, so "got away" is a stretch, but "murder" is an even bigger stretch. It got away -- or didn't -- with moving money; the people for whom it moved money were murderers.
Anyway yesterday Citigroup Inc.'s Banamex USA unit agreed to pay $97.4 million to resolve a Justice Department investigation into compliance failures in its business of moving money between the U.S. and Mexico. "After years of billion-dollar settlements in anti-money-laundering and mortgage-related cases brought by prosecutors working for the Obama administration, white-collar defense lawyers and investors have been watching to see whether the new administration will ease up on the banks," notes the New York Times. Did it? Banamex's fine is less than some predecessors' fines, but Banamex's misbehavior is pretty bland, compared to previous cases about financing terrorists or genocidal governments. Banamex's violations are more like this:
From February 16, 2004, through December 17, 2007, Beneficiary C received more than 1,400 individual remittance transactions from more than 950 different senders located in more than 40 different states across the United States. Of those transactions, BUSA’s AML monitoring rule produced 24 separate alerts, comprising 1,382 transactions totaling $824,102. BUSA never filed a SAR.
That's just some numbers and letters! No moral blame attaches to that; Beneficiary C is just a guy with a wide network of friends who liked sending him money. Or something. Of course you could imagine it differently:
From February 16, 2004, through December 17, 2007, Drug Lord C received more than 1,400 payments from more than 950 members of his vast drug conspiracy located across the United States, totaling $824,102. He used this money to pay for his lavish lifestyle, and to kill people in gruesome ways described more fully in the next eight paragraphs.
That would be bad! There are other ways to imagine it though:
From February 16, 2004, through December 17, 2007, Prolific Knitter C received more than 1,400 payments from more than 950 customers of her Etsy store, totaling $824,102. She used this money to pay for yarn, and to buy gifts for her grandchildren.
Banamex's non-prosecution agreement leaves it for us to guess. Presumably the Justice Department doesn't know which one it is, and neither does Banamex. If you wanted to fine Citigroup a ton of money, and make it plead guilty to a crime, and bring moral condemnation down on it, you'd probably work hard to find a drug kingpin or two who was on the receiving end of these transfers. If you just want to fine Citigroup some money for its administrative failings, though, you write them up as administrative failings.
Elsewhere in money laundering, here is a story about "the US government’s first courtroom bid to hold a compliance officer personally responsible for not preventing financial wrongdoing":
“Compliance officers find this case very troubling,” said Todd Cipperman, an industry consultant in Wayne, Pennsylvania. “To hold him accountable and not hold other senior executives accountable seems strange and unfair.”
Compliance officers — among the corporate world’s least glamorous players — fear they are being sacrificed to the government’s desire to punish individuals for financial industry misdeeds.
On the one hand, you know, you had one job. If there's one person to hold responsible for not catching money laundering, it's the person whose job it was to catch money laundering. On the other hand, the incentives that creates are really bad! The stereotyped dynamic in any regulatorily sensitive business is:
Business unit employee: Let's do the thing.
Compliance officer: We're not supposed to do that thing.
BUE: What's the worst that could happen?
CO: You could go to jail.
BUE: Oh fine we won't do the thing.
What you don't want is this dynamic:
Business unit employee: Let's do the thing.
Compliance officer: We're not supposed to do that thing.
BUE: What's the worst that could happen?
CO: I could go to jail.
BUE: Hahaha that's what we hired you for buddy.
Compliance officers have authority at a company to the extent that people believe that their job is to keep everyone else out of jail. If it's to keep themselves out of jail, then they're just sacrificial lambs.
Quants quants quants quants.
The Wall Street Journal's series on quants continues with this story about how a pension fund wanted to count up how much money it paid in fees and used a computer to do it. I do much the same thing to track my own spending -- I use Excel -- but I am not quite sure it counts as quantitative investing. Here's another skeptic with a different complaint:
“If you’re using software to deal with the complexity in your portfolio maybe you should simplify your portfolio first,” said Marc Levine, chair of the Illinois State Board of Investment, which oversees $20 billion for state employees, judges and lawmakers.
Come on, man! "If you're using software to help operate your car, maybe you should just stay home." "If you're using software to file your taxes, maybe you should try making less money." "If you're using software to run your factory, maybe you should just pound some rocks together to build stuff." The idea that software -- "software"! -- is an indication of excessive complexity, that you should be able to hold all the details of your portfolio in your mind at all times, is quaint but wrong.
No but seriously. It is perhaps a sub-"quant" innovation, but the real idea here is that you should be able to hold all the details of your portfolio in your computer, and use that computer to ask questions of the portfolio and gain deeper insight into your investments. A question like "how much are we paying in fees" is easy if all of the information about all of your investments is available to a computer that is programmed to be able to answer sensible questions, and difficult otherwise. Or you could ask questions about risk exposures:
“You can look at your portfolio and say ‘Oh wow, I’ve got a lot more inflation risk than I should have or a lot more credit risk than I should have,” said Robert T. Bass, a BlackRock Inc. managing director, who provides software that conducts this analysis for pension funds. “Maybe I should take that down a little.’”
Maybe you are still making decisions yourself, based on your gut instinct for how much credit risk you want, but you should at least know how much credit risk you have. If you just think of yourself as owning a list of bonds, that is a hard question to answer. If you think of yourself as owning a portfolio with a bunch of factor exposures, and if your computer knows what all those exposures are, then it's much easier.
Elsewhere in the quant series, here's a breakdown of an actual trading algorithm from Quantopian, designed to buy stocks that are trading below their moving average price and sell stocks that are trading above, "based on a well-used methodology called Bollinger Bands." It ends with this eloquent passage of Python:
I feel like I am often reading warnings against over-scientism in finance, worries that quantitative techniques imported from deterministic sciences will lead to overconfidence. It is worth generally taking those warnings with a grain of salt -- it strikes me that data scientists tend to be more epistemically humble than, you know, MBAs -- but still that bit of code might give you pause. Order the optimal portfolio given that the objective is the objective, the constraints are the constraints, and the universe is today's universe! It just reads a little glib.
Also you can build your own pretend trading bot.
Controversial news, everyone: The fiduciary rule is happening!
The retirement-savings regulation known as the fiduciary rule will take effect June 9 without further delay, Labor Department Secretary Alexander Acosta said Monday.
In an opinion piece for The Wall Street Journal, the new Labor Department chief said “respect for the rule of law” precludes a further delay in next month’s applicability date, which had been postponed from April 10 after President Donald Trump directed the department to re-evaluate the Obama-era regulation meant to protect retirement savings from conflicted investment advice.
I am perhaps the only person in America without a strong view one way or the other about the advisability of the fiduciary rule so, you know, fine, whatever. But as a regulatory matter this is fascinating: During the campaign, Trump fund-raiser Anthony Scaramucci compared the fiduciary rule to slavery, and Trump's instruction to re-evaluate the rule came with a pretty strong hint of "and get rid of it." And Acosta still might get rid of it, saying that "the Fiduciary Rule as written may not align with President Trump’s deregulatory goals."
But he can't get rid of it yet. First of all he has to go through much the same process -- economic analysis, public notice and comment -- that the Obama Labor Department went through in making the rule. The way the U.S. administrative state works is that regulating is hard, but deregulating is pretty much equally hard. And so Acosta wrote a paean to the Administrative Procedure Act on the opinion page of the Wall Street Journal, which is not exactly what I expected from the Trump administration. I will quote from it generously here because, come on, how often do you read paeans to the Administrative Procedure Act?
Engraved above the doors of the Supreme Court are the words “Equal Justice Under Law.” Those four words announce that no one is above the law, that everyone is entitled to its protections, and that Washington must, first and foremost, follow its own rules. This means federal agencies can act only as the law allows: The law sets limits on their power and establishes procedures they must follow when they regulate—or deregulate.
The Administrative Procedure Act is one of these laws. Congress had good reason to adopt it: In the modern world, regulations are akin in power to statutes, but agency heads are not elected. Thus, before an agency can regulate or deregulate, it must generally provide notice and seek public comment. The process ensures that all Americans—workers, small businesses, corporations, communities—have an opportunity to express their concerns before a rule is written or changed.
So it will take a while to even consider repealing the fiduciary rule. And meanwhile it will start going into effect, and once brokerages have moved customers to unconflicted fee-based accounts, it may be hard to move them back. The president doesn't want the fiduciary rule, and his Labor Secretary doesn't want the fiduciary rule, and Congress probably doesn't want the fiduciary rule, but they're going to get it anyway, because the rule of law operates separately from their current desires.
People are worried that people aren't worried enough.
"The pain of an almost paralyzed stock market has seeped its way into the money machine of equity quants, raising anxiety levels among analysts who say they’ve seen this movie before." The particular movie that they've seen is the one where investors react to low volatility by increasing leverage:
“If you’re in a long-in-the-tooth factor that everyone is involved in, that’s going down in volatility, one way to offset the reduced returns is with leverage,” said Mark Connors, the global head of risk advisory at Credit Suisse Group AG. “For any event, whether it’s a blip or a big macro event, there’s then a greater chance of sharp deleveraging.”
I often use this heading to make fun of over-interpretation of low readings on the VIX volatility index as signs of "complacency." But this worry -- the worry that quiet markets will cause people to take more risk, and that when volatility does come back it will snap back hard -- is in some sense obviously right, Minskyan, the basic story of market cycles. There is not that much you can do about it, though, except worry. Elsewhere, here is my Bloomberg View colleague Tyler Cowen with "An Optimistic View of Markets' Relative Calm."
People are worried about covered interest parity.
Congratulations to this somewhat niche worry, which has now won a major award!
AQR Capital Management, LLC (“AQR”), a leading global investment firm, today announced the winners of its sixth annual AQR Insight Award.
This year Deviations from Covered Interest Rate Parity was awarded First Prize. The research was authored by Wenxin Du, Ph.D., Federal Reserve Board; Alexander Tepper, Ph.D., Columbia University; and Adrien Verdelhan, Ph.D., MIT Sloan School of Management. The authors will share the $100,000 annual prize.
It was also the subject of a conference delightfully titled "CIP-RIP?" But at that conference, Guy Debelle of the Reserve Bank of Australia gave a speech called "How I Learned to Stop Worrying and Love the Basis," and he, as the title suggests, isn't worried:
A positive basis is often described as indicative of a US dollar shortage. I don't find that term very helpful at all. ‘Shortage’ very much implies a quantitative constraint, rather than a pricing constraint. But in most circumstances, we are talking about the latter. That is, the supply of dollars might go down, so the price goes up. There is no shortage, there is just more demand than supply, and, as with other markets, the price adjusts to equilibrate the market. The market is functioning effectively and efficiently. The focus of analysis should be on the various factors that affect demand and supply.
Remember the basic story of this worry. Covered interest parity is the idea -- a fundamental law of finance, people used to think -- that interest rates implied by foreign-exchange swaps should equal interest rates in the domestic currency. If the domestic dollar interest rate is higher than the synthetic interest rate you get by swapping a foreign currency into dollars, then in theory someone should borrow in foreign currency, swap to dollars, invest at Libor, and arbitrage the difference away. But they don't, now; a basis remains between the two rates. The existence of an arbitrage bothers people, and it is usually explained in terms like "banks and hedge funds used to arbitrage this basis away but now they can't because of regulatory or balance-sheet constraints." Debelle's explanation is not necessarily different; he's just less worried about those constraints:
A useful taxonomy is the following: banks, hedge funds, corporates and asset managers/real money pools, including central banks. Each of these groups has varying degrees of flexibility to arbitrage the basis. They have varying constraints on how they can use their balance sheets, varying funding sources and varying credit appetites and limits which constrain the investable universe. There is a fundamental difference between a bank (or a hedge fund) that needs to fund itself to arbitrage the basis, compared to a real money investor, which may be naturally long in the funding currency. On the other hand, real money often has restrictions on the ability to take on leverage in the way that a bank doesn't. Real money can often be constrained in their ability to arbitrage the basis because their mandate doesn't allow them to.
In my experience, one of the fundamental laws of finance is to never underestimate mandate constraints.
The open-plan-office backlash is in full swing:
As employees and managers squeeze closer together, productivity and morale have suffered. In a review of more than 100 studies of work environments, British researchers found that despite improving communication in some instances, open-office spaces hurt workers’ motivation and ability to focus.
And it is being led, of course, by chief executive officers. Here's one:
Having his own space, he says, makes him feel more like a leader. After brainstorming or regrouping with some music or even watching an inspirational video, “I can get back out there with my Superman cape on,” he says.
Yes right watching videos at work is a major appeal of having your own office; it is helpful when CEOs recognize that.
Novice Nomura Trader Tells Jury How Bosses Taught Him to Lie. Ukrainian Hacker Gets 2 1/2 Years for Stealing News Releases. Greek creditors fail to reach deal in 7 hours of talks. Supreme Court Limits Where Patent Lawsuits Can Be Filed. Brussels sets rules for Brexit regulatory agencies fight. Elliott and Arconic Are Now in It Together. Closed-end funds "with hard-to-read annual reports trade at substantial discounts relative to CEFs with easy-to-read annual reports." Cybersleuths Unearth More Clues Linking WannaCry to North Korea. Citigroup to Tap Nasdaq for Blockchain Payment Technology. Auto Lender Santander Checked Income on Just 8% in Subprime ABS. Hedge Fund Manager Charged With Making Ponzi Payments, Defrauding Investors. Why Warren Buffett Took Out an Ad on the Same Page as One for a Strip Club. Kung fu master uses genitals to pull bus down street. Thousands Of Nathan's Hot Dogs Recalled Thanks To Metal Shards.
If you'd like to get Money Stuff in handy email 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:
James Greiff at firstname.lastname@example.org