Libor Fines and Robot Analysts

Matt Levine is a Bloomberg View columnist. He was an editor of Dealbreaker, an investment banker at Goldman Sachs, a mergers and acquisitions lawyer at Wachtell, Lipton, Rosen & Katz and a clerk for the U.S. Court of Appeals for the Third Circuit.
Read More.
a | A

Libor fines?

Libor fines! Barclays, which in 2012 paid $160 million to the U.S. Department of Justice, $200 million to the Commodity Futures Trading Commission and 59.5 million pounds to the U.K. Financial Services Authority for manipulating Libor, will pay another $100 million to a collection of U.S. state attorneys general for manipulating Libor. There are 44 states involved in this settlement, which I guess technically means that Barclays has now been fined 47 times for the same Libor manipulation.

In a sense it got off easy. It could have been 48 times: When the European Commission fined a bunch of banks 1.5 billion euros for manipulating Libor, "Barclays received full immunity for revealing the existence of the cartel and thereby avoided a fine of around € 690 million." But also, I mean, there are 50 states. There are at least 193 countries. Interest-rate manipulation isn't like shoplifting; it doesn't just happen in one place. You might do it from your office in London, but the interest rate filters out into the world. Surely someone in Cameroon was somehow affected by the manipulation of dollar Libor. There are probably some county ordinances in Iowa that were violated by Libor manipulation; where are the counties' settlements? 

What is particularly weird about the overlapping governments in the Libor case is that some of them were in on it. Most famously, there has been some suggestion that the Bank of England, let's say, strongly hinted to Barclays that it should get its Libor submissions down a bit, though that claim is controversial. But even just from yesterday's settlement:

For example, a September 26, 2007 email from a Barclays Euribor submitter to a large distribution list, which included an employee at the Federal Reserve Bank of New York, a representative at the European Central Bank, and a World Bank employee, states: “USD: Same old boring story. Day to day money is trading very cheap, the quarter end turn is looking relatively well bid at 5.20-5.10. There is liquidity in one to six months but our feeling is that libors are again becoming rather unrealistic and do not reflect the true cost of borrowing.”

There's nothing new here -- the states' settlement is based on previous settlements, and we've seen that e-mail before -- but it is still sort of amazing. Barclays wasn't hiding its Libor manipulation from the regulators. Barclays was e-mailing the Fed and the ECB about Libor manipulation in 2007. The Fed and the ECB knew all about it. But of course they haven't fined Barclays. The attorney general of New York didn't know about it, never mind the attorney general of Idaho, which makes it a bit less awkward for them to be outraged now.


In a statement on Monday, Barclays said it was pleased to have resolved that matter. “We believe this settlement is in the best interests of our shareholders and clients, and allows us to continue to focus on the future.”

I guess that is what you say after you settle an investigation. It seems unconvincing these days. Banks never put anything behind them; there's always some other regulator that can fine you for the same thing.


This seems fun:

A machine intelligence system, dubbed Emma AI, is starting a fund that hopes to outsmart the humans and computers that make a living trading stocks.


Shaunak Khire, Emma’s creator, claims his system differs from current finance computing — high-frequency trading and “quant” data science — because its system of neural nets takes into account a more complex set of factors affecting stocks, like management changes or monetary policy in Europe, that other programs miss.

“This is not algorithmic trading,” he said. “This is literally replication of an analyst.”

I hope they put a fleece on the AI. For some reason "Emma will start trading stocks from pharma giant GSK and Tesla along with U.S. Treasury bonds," which seems like sort of a limited coverage universe for this robot analyst, but everyone's got to start somewhere. It's more of a robot intern, maybe. But Khire's basic point seems right. There is some low-hanging fruit in computerizing trading; high-speed market-making and statistical arbitrage are relatively straightforward to computerize, and so they were computerized relatively early. But there's no reason to think that computers would stop there, or that fundamental analysis of companies would be impossible for them. Fundamental analysis is essentially a process of discovering and evaluating patterns in data, just like market-making is; it's just that the patterns and data are more complex and varied. Good luck, Emma!

Elsewhere, here is a machine-learning algorithm to trade the SPDR S&P 500 exchange-traded fund, if that's of any use to you.

How's Tidjane Thiam doing?

Francine Lacqua of Bloomberg Markets interviewed Tidjane Thiam, the former McKinsey consultant who now runs Credit Suisse, and it reads a little like an interview with a consultant? 

FL When will you feel as if you’ve cut enough costs?

TT When we don’t have to cut them anymore. My real philosophy about cost is productivity gain. I am, for all my faults and sins, an engineer. I come from a culture of automation technology. I believe in productivity improvements. An organization should improve productivity by 2 or 3 percent every year. And I always prefer to talk about productivity more than cost. It’s a much better word, because that’s what’s relevant. We’re promoting a culture of continuous improvement, and the restructuring and reengineering is very deep. It’s about changing the structure of processes and redesigning them in a more efficient way. Our global markets division has done a pilot of that, and their solution was a front-to-back, but also back-to-front, reengineering in eight weeks. They’re really excited about the opportunities, where we can serve our clients better, cheaper, and faster. But for that you need to break the silos, you need to get people around the table and say, “How can we work better?”

I don't know, I guess I am being unfair. You don't need to be a consultant to talk about breaking the silos; bankers talk about it all the time, which is a little weird. If everyone is so anti-silo, why are there still so many silos to break? Anyway Thiam rejects the "consultant" characterization:

I’ve been working for 30 years. I was a consultant for 10 of them. I left McKinsey in 1994. That’s a long time ago. Kids who were born then are in college now, like my son.

Fair enough. This might be my favorite part:

FL If you could change one thing about Credit Suisse straightaway, what would you do?

TT You mean other than the share price? [Laughs.]

FL Is that what you would change?

TT No. See the thing with change is that the process is often as important as the outcome. You grow, and you develop, and you learn by achieving difficult things. That has enormous value. A magic change would almost not be worth having, because you wouldn’t learn anything and you might end up repeating the same mistake. I’m tempted to say nothing, because it’s good to learn things. There’s a lot that an organization and its people can learn from that process.

Okay but wait this is actually a good corporate-governance hypothetical? Like: If you were the chief executive officer of a bank, and a genie gave you one wish (for the bank), what would you wish for? Would it be some sort of good sustainable underlying operating performance, regardless of what the market thinks? Or is your obligation just to wish for a higher share price, even if the genie gets you there through less-than-sustainable methods? In any case, surely "nothing, because it's good to learn things" is not the right answer. Shareholders care about the destination, not the journey. Is it a violation of your fiduciary duties to turn down a magical improvement in your company, just because it doesn't feel earned and authentic?

Hoot fine.

Look, we all know that back at the height of the dot-com boom, sell-side equity research was a den of iniquity, giving dishonest favorable coverage to terrible companies just to win investment banking business. But the Great Research Settlement of '03 cleaned all of that up, and now sell-side equity research is just an epistemological abyss, in which analysts are required to believe what they say but not allowed to say what they think. Here is a story about Deutsche Bank:

In one instance, a registered representative contemporaneously relayed to at least one customer information from a Research Hoot which indicated that the impact of a positive news announcement had not been factored into the price of the company's security. The Firm subsequently issued a research update substantially increasing its price target on the company based on the impact of the news event.

So I think what happened there is something like this:

  1. Company X announced good news.
  2. The Deutsche Bank analyst covering that company was like "hey, this good news is good news."
  3. He said that over the "Research Hoot" line that went out Deutsche Bank's squawk box system to any employees who might be interested, including brokers (registered representatives) in its Private Client Services division.
  4. One PCS broker heard it, called his client, and said something to the effect of "hey, did you see that good news that Company X announced? Turns out our research analyst thinks it's good news."
  5. Later, the analyst put out a formal research note increasing his price target on Company X stock.

For this -- well, for having "failed to establish adequate supervision over registered representatives' access to hoots or their communications with customers regarding hoots" -- the Financial Industry Regulatory Authority fined Deutsche Bank $12.5 million. This is not a case of an analyst telling some clients something that conflicts with his published reports, or (really) giving them early access to a change in his recommendation. (Though it's arguably that, because he did later publish.) This is a case of a company announcing good news and the analyst (indirectly) telling a client, yeah, that news was good. But the nervousness around research analysts is such that even that innocuous reaction can only be conveyed in a published research note sent at the same time to every client. 

Fintech and competition.

Thomas Philippon is a financial economist at the Stern School at New York University who is perhaps most famous for trying to figure out why the financial industry makes so much money. (It is not, in his view, because the industry is so good at its job.) He has a new NBER Working Paper out (free version here) about "The FinTech Opportunity," the possibility that financial technology firms might undercut incumbents and make financial services less expensive. Here's one recommendation for fintech regulation:

There are many regions of the financial system where incumbents are entrenched and entry is difficult. This is precisely where regulators should actively encourage entry. An example of a highly concentrated market is custody and securities settlement. In theory, the blockchain technology could improve the efficiency of the market, but if there is no entry, this would simply increase the rents of incumbents. A restricted blockchain could in fact be used by incumbents to deter entry and stifle innovation.

This then brings the thorny issue of biases in the competition between entrants and incumbents. Ensuring a level playing field is a traditional goal of regulation. Darolles (2016) discusses this idea in the context of FinTech and argues, from a microeconomic perspective, that regulators should indeed ensure a level playing field.

This line of argument, however, does not readily apply to many of the distortions that plague the finance industry. For instance, what does a level playing field mean when incumbents are too-big-to-fail? Or when they rely excessively on short term leverage? The level playing field argument applies when entrants are supposed to do the same things as incumbents, only better and/or cheaper. But, as we have discussed, the goal of financial reforms in the wake of the Great Recession should be to change the industry. 

It is useful to recognize how culturally unlikely this is. Take the blockchain example. One way to use blockchains to improve securities settlement is to have a consortium of banks get together, along with their securities and banking regulators, and set up their own limited-access blockchain that is basically just a shared database of securities ownership. Another way is for some tech startup to build a settlement system on the open-access bitcoin blockchain and try to sign up trading partners. The latter approach has a lot of potential advantages, in terms of promoting innovation and fostering competition and lowering cost and not entrenching too-big-to-fail incumbents. But the bitcoin blockchain is scary and anonymous and open to everyone, and the tech startup probably doesn't have the deep regulatory relationships and army of compliance officers that the bank consortium has. Which approach do you think the regulators will prefer? 

Elsewhere: "U.S. to auction $1.6 million of bitcoin from various cases."

Good rich guy.

Here's a guy who won 100,000 shares of Inc. in a silly contest, which are now probably worth something north of $10 million:

“It feels good; it feels really good,” Martin, a York, Pennsylvania, resident, said in a phone interview during his lunch break. “The way I think to describe it is ‘occasional hysterical laughing.”’

He doesn't even know how much the shares will be worth in Jet's acquisition by Wal-Mart, "but he's hoping it's in the 'multi-millions.'" (Isn't private-company governance weird?) We talk sometimes around here about hedge-fund billionaires who work 18-hour days and fight to get back into the industry even after they are banned from it by regulators, and their psychology is totally alien to me. Why would anyone keep working if they had a billion dollars, I wonder, and then shrug and tell myself that that's why I'm not a billionaire. But this guy? "Occasional hysterical laughter"? I'd be into that. Of course he's not a billionaire either.

People are worried about duration.

No, today is the opposite worry; it's call risk again.

Companies and government agencies are “calling” bonds at the fastest pace in four years, taking advantage of provisions that let them redeem securities under certain circumstances and save money by reissuing at lower rates.

Then of course you reissue for 30 years at near-zero rates, and people can worry about duration risk.

People are worried about unicorns.

Times are lean in the Enchanted Forest, and it is getting harder for unicorns to earn their horns:

By this point last year, at least 53 startups raised capital at a valuation of $1 billion or more for the first time, according to VentureSource. This year, nine have reached that milestone. Also so far this year, just nine U.S.-based tech companies have staged an IPO, according to Dealogic. That is down from an average of 24 through the first week of August the prior four years.

And so "flush corporations are providing lucrative paydays for some startups that once had grander ambitions" of exiting the Enchanted Forest by the front door of an initial public offering. Meanwhile, JPMorgan "has more than 20 global IPOs lined up and ready to go in September alone," though "the highest-value tech deals are likely to wait until 2017 and 2018."

People are worried about bond market liquidity.

The traditional bond market liquidity worry in the U.S. is that, if prices drop, there won't be enough buyers for bonds. But in Russia the problem is not enough sellers:

Russia is at the epicenter of a liquidity crunch in emerging-market debt caused by a sudden surge in demand for higher-yielding assets as Britain’s vote to leave the European Union in June prompted a fresh wave of stimulus measures from central banks worldwide. Since sanctions barred many Russian companies from issuing debt abroad, some have bought back bonds, further exacerbating the supply squeeze.

It still counts as a liquidity worry though.

Me yesterday.

I wrote about the Fannie Mae and Freddie Mac stress tests.

Things happen.

The Republic of Congo defaulted on its debt for about nine days. Brexit Bulletin: What the Rest of the EU Wants. Goldman Sachs’s Fight to Avoid Paying Employees’ Legal Fees. Chinese Traders Roil Commodity Markets. TIAA to Buy EverBank for $2.5 Billion. Berkshire Hathaway has a lot of cash. But no more credit derivatives. Italy offers "the highest currency-hedged yield among the five biggest sovereign debt markets." New Rules and Fresh Headaches for Short-Term Borrowers. Are Negative Rates Backfiring? Bridgewater making waves with dual-party valuation. Moore Capital founder wins Bahamian privacy case. States Vie to Shield the Wealth of the 1 Percent. Jury Trials Vanish, and Justice Is Served Behind Closed Doors. Chevron Wins Ruling Blocking Enforcement of $9.5 Billion Ecuador Judgment. Roddy Boyd on Prospect Capital. Trump Jars Hamptons Anew With Economic Picks Including Paulson. How to Stay Motivated When Everyone Else Is on Vacation. Rich people are bragging about their luxe panic rooms. Thermostat ransomware. Toasteroid. Martha Stewart, Snoop Dogg Partner for VH1 Dinner-Party Series. "I may not be magic, but I know that Taylor Swift is a Slytherin." 

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

To contact the editor responsible for this story:
Brooke Sample at