Renaissance options

A pretty ordinary way for a hedge fund to invest is that it borrows money from a bank and uses that money, along with the fund’s own equity, to buy stuff. If the stuff goes up the hedge fund pays back the bank’s loan and keeps the profits. If the stuff loses value the hedge fund takes the loss, reducing its equity. If the stuff loses a lot of value then the equity will reach zero, the loan won’t get paid back, and the bank will eat the rest of the loss. If the fund puts up $20 and borrows $80 from the bank, it can buy $100 worth of stuff. If that stuff ends up being worth $120, the hedge fund doubles its money to $40 (ignoring interest on the loan); if it ends up being worth $90 the hedge fund only gets back $10 of its original $20; if it ends up being worth $50 then the hedge fund loses its $20 and the bank loses $30.[1]

This has, roughly speaking, the profile of a call option: If the stuff is worth more than the amount of the bank’s loan, the hedge fund has all the upside; if it’s worth less, the hedge fund’s downside is floored at the amount of equity it put up.[2] (You can think of the strike price as being the amount of the loan and the option premium as being the amount of the fund’s equity plus the interest on the loan.) It’s not really an option on anything; generally, the hedge fund does not buy one stock or bond and just hold it for years. But you can think of it as a call option on the hedge fund’s entire portfolio as it changes from time to time.

Lots of things have the profile of call options. That doesn’t make them call options, but it doesn’t make them not call options either. It is just a fact that there are lots of financial transactions with floored downside and unlimited upside, and it is often helpful and explanatory to say “oh that looks like a call option.”

But at some point some hedge funds and banks discovered that this thing, which looks like a call option, can be called a call option. Instead of saying “the hedge fund puts up equity, borrows from a bank, and uses that money to buy assets,” you can say “the hedge fund pays option premium to the bank, and the bank sells the hedge fund an option on a portfolio of assets picked by the hedge fund.” The documents say “call option” at the top. It’s a weird call option—the bank is selling the hedge fund an option on a changing basket of stuff, and the hedge fund gets to pick what’s in the basket and change it whenever it wants—but you can do it.

Why would you do this? Two main reasons. One is tax. If you just buy a basket of stuff, and change what’s in it constantly, you will have to pay short-term capital gains rates on the stuff you are buying and selling. But if you buy an option on that basket, and the option expires in three years, then you don’t have any realized gains on all the buying and selling during those three years. From a tax perspective, you have just bought a three-year option and waited three years until it pays off. That payoff (if it’s positive) is a long-term gain, which is taxed at lower rates.

The other reason is leverage: For regulatory and other reasons, your bank may not be willing to lend you an unlimited amount of money against your basket of stuff; you might be capped at 6 or 7 to 1 leverage. But call options have no regulatory leverage caps—there are out-of-the-money call options on Alphabet Inc. stock trading for less than a dollar, while the stock itself is above $1,300—so you can conceivably buy an “option” on a “basket” of $100 worth of stuff for a “premium” of only $5, if you can convince a bank to take on the risk.

We have talked about all of this before, because the most famous hedge fund that did this is Renaissance Technologies Inc., which went to banks and “bought” “basket options” on portfolios of securities that it selected and that changed on a second-by-second basis. This is highly controversial, the Senate has yelled at Renaissance about it, and the fund is mired in disputes with the Internal Revenue Service about whether it needs to pay back more taxes on the basket options. Renaissance’s IRS auditors think that calling the trades “options” is just a trick and that they really owe short-term capital gains taxes; Renaissance thinks that it bought real options and had long-term gains on them. The value of the disputed taxes could be about $6.8 billion.

Here’s some news from Bloomberg's Zachary Mider:

A little-noticed decision by the Internal Revenue Service’s appeals unit may spell trouble for legendary investor Jim Simons, who’s embroiled in a multibillion-dollar tax dispute with the agency.

Reviewing the audit of an investment manager in Connecticut, the IRS Office of Appeals rejected a tax-avoidance maneuver involving so-called basket options. That’s the type of transaction at the heart of a separate, larger case involving Simons’s Renaissance Technologies hedge fund. The decision, made public in a court filing in May, could offer a preview of the tax agency’s reasoning in the Renaissance case. ...

In a confidential decision in December, the IRS held that GWA, a Hartford, Connecticut-based company run by hedge fund pioneer George A. Weiss, had under-reported ordinary income in 2009 and 2010 by a total of $527 million through its use of basket options. The decision wasn’t made public until Weiss challenged the action in U.S. Tax Court this year. That case is pending.

I can’t say I’m particularly surprised. I don’t know whether these trades are “really” options, not only because I don’t know all the facts but also because that does not strike me as a particularly meaningful question. But if I were the IRS, I would surely think that they are not “really” options for tax purposes, because that is just too cute, man. They have the payoff profile of call options, but they also have the payoff profile of regular old margin loans, and the difference is that they say “call option” at the top to avoid taxes. If I were the IRS, I would not want to let people avoid taxes by just writing different words at the tops of their documents.

My favorite part is that the investors’ counterargument is that they’re not just trying to avoid taxes, they’re also trying to avoid margin requirements:

In the Tax Court case, Weiss argues that the options had business purposes beyond tax savings: They offered more leverage than was available in a typical brokerage arrangement, while at the same time limiting the firm’s risk of loss. … Renaissance has made similar arguments in defense of its transactions.

A weird feature of U.S. tax law[3] is that, if you do a thing purely to get around tax rules, then that is bad and a sham and the IRS can look through it and make you pay your taxes. But if you do the thing not only to get around tax rules but also to get around other rules (like margin requirements), then from the IRS’s perspective you have a valid business purpose and you might be able to keep your good tax treatment. “We’re not just gaming your rules, we’re gaming other regulators’ rules too” is, surprisingly, an argument that might persuade the IRS.

All the tech companies are banks now

Yesterday Facebook Inc. announced a thing called “Facebook Pay,” which will let you send money over Facebook, Instagram and WhatsApp. Today the Wall Street Journal reported that “Google will soon offer checking accounts to consumers, becoming the latest Silicon Valley heavyweight to push into finance.” Back in March, Apple Inc. announced its own credit card. 

All of these stories are … very … boring? Like from first principles it has long seemed inevitable that the tech giants would get into financial services in some way. They have way more name recognition and contact with consumers than any bank does, and at least until recently they had much better reputations than most banks. People use their products constantly and enthusiastically. Certainly they are better at advertising than banks are; Facebook and Google are basically pure dominant advertising companies. So if they want to sell you a credit card or a checking account, or have you pay your friends through their apps, they should be able to manage that.

But it’s not just a marketing advantage. It’s a tech advantage. Tech companies have fast smart computers and fast smart computer programmers. Computerized decision-making is central to modern finance, and intuitively you might expect tech companies to be better at it than finance companies. They have more data than anyone else, and they have more expertise in processing it. They know everything about everyone, so presumably they should be able to make smarter credit and investing decisions than, you know, some bank relying on FICO scores or whatever.

There are big regulatory barriers to getting into a lot of financial-services businesses, but it does sort of seem like the giant tech companies should have the money and expertise and political clout to manage those barriers and comply with, or lobby to change, the regulations.

But in actual fact “Apple launches a credit card” means that Goldman Sachs Group Inc. launched a credit card and Apple, like, designed the packaging and marketing for it. Google’s checking accounts will be checking accounts at Citigroup Inc. (or Stanford Federal Credit Union), only Google will wave a magic wand over them (and, presumably, advertise them). Facebook Pay links to your credit card (or PayPal). The tech companies are front ends for regular old banks. 

There is an ongoing scandal about the Apple card, after entrepreneur David Heinemeier Hansson tweeted that he got a far higher credit limit than his wife for no apparent reason; Goldman is being investigated for possible gender discrimination. (Disclosure, I used to work at Goldman, in the days before it got into the credit card business.) There is something a tiny bit odd about this scandal. Goldman is being investigated. Goldman is tweeting defensively about its credit decisions. Not Apple. “Goldman and Apple are delegating credit assessment to a black box,” says Hansson, but what is actually happening is that Apple is delegating credit assessment to Goldman, which is then delegating it to a black box. It seems that Goldman built the black box, and Apple is just renting it from Goldman. Fine, sure, yes, Goldman is in the credit-assessment business and Apple isn’t.

But … couldn’t Apple be? Like, Apple is quite good at building boxes that contain computer software, no? Couldn’t a foray into financial services involve building an algorithm to assess credit? Intuitively, couldn’t Apple think that's the sort of thing it might be good at? The advertising for the Apple card calls it “A new kind of credit card. Created by Apple, not a bank.” That appears to be true of the appearance of the physical card. But the credit algorithms were created by a bank, to Apple’s eventual embarrassment. It is just a little odd that Apple seems to have been so incurious about the algorithms. It's a tech company! 

It is not like these companies are generally unambitious. Google, a giant advertising company, is trying to build self-driving cars. (So is Apple.) It is trying to defeat death. Arguably these are more important problems than the fair allocation of credit, but they are also harder problems. Surely credit assessment, as a computer-science problem, is closer to advertising than driving is. You could easily imagine the tech companies actually trying to get into the substantive business of financial services, trying to do finance better than the financial companies, trying to leverage their expertise with data and artificial intelligence to solve problems that still vex banks. Mostly though they seem to be getting into the business of marketing financial services, and leaving the finance to the banks.

Oh hahahaha except Facebook, which in addition to Facebook Pay is also working on Libra, a different and much stranger way to pay for things on Facebook. Libra, Facebook’s stablecoin, has the potential to be genuinely new and disruptive, to take finance out of the hands of banks. It’s not going great! From a certain perspective, Facebook Pay—“ehhh just link your credit card to our app”—looks like a retreat from Libra. Actually doing finance stuff is difficult for the tech giants. But they sure are good at marketing. 

Silvergate

Last week a bank called Silvergate Capital Corp. went public on the New York Stock Exchange.[4] You can read the prospectus here. It was a small initial public offering, raising about $40 million for Silvergate and its investors, and it’s a small bank (about $250 million market capitalization and $2.2 billion of assets), but it is amazing.

Silvergate used to be a regular boring bank, taking deposits and using them to make commercial and real-estate loans. But in 2013 it decided to change its focus and instead become a bank for cryptocurrency customers. Now its “customers include some of the largest U.S. exchanges and global investors in the digital currency industry”; the prospectus mentions Coinbase, Genesis, Bitstamp, Gemini, Sofi, Circle and Paxos. When it was a small boring bank it had three branches, but earlier this year it got rid of two of them to focus more on cryptocurrencies.

Why would you want to be a bank for cryptocurrency exchanges and investors? Well lots of reasons I suppose, but Silvergate has a big simple one: You apparently don’t have to pay interest on crypto-related deposits. “The success of our digital currency initiative has enabled Silvergate Bank to rapidly grow and maintain noninterest bearing deposits from digital currency customers,” it says, increasing noninterest-bearing deposits from 12.4% of deposits in 2013 to 79.9% in 2019. In 2018, Silvergate had interest expense of $3.1 million on liabilities of about $1.8 billion, for a total interest expense of about 0.17% of liabilities.[5]  

This in turn lets Silvergate invest safely and boringly: Rather than lending money to businesses or homeowners, it can just take the money that its customers put in the bank and … put it in other banks. “This emphasis on noninterest bearing deposits and noninterest income,” it says, “will likely result in a continued shift in our asset composition with a greater percentage consisting of liquid assets such as interest earning deposits in other banks and investment securities.” (Actually the money is largely in mortgage-backed securities, but a good chunk of it really is just deposits at other banks.)

Schematically you can think of Silvergate as providing a simple arbitrage:

1. Lots of banks want nothing to do with cryptocurrency customers.
2. But they’ll happily take deposits from other banks.
3. Silvergate can pay cryptocurrency customers zero interest on deposits (no competition!), and then put the money in ordinary banks and earn interest. 

That’s overstated, but it is kind of the intuition: Silvergate can provide the, uh, gateway to the banking system, and take its cut by not paying interest. 

But there’s more. A big feature of the prospectus is “the Silvergate Exchange Network, or SEN, our proprietary, virtually instantaneous payment network for participants in the digital currency industry.” Here’s a description of the SEN:

The core function of the SEN is to allow participants to make transfers of U.S. dollars from their SEN account at the Bank to the Bank account of another SEN participant with which a counterparty relationship has been established, and to view funds transfers received from their SEN counterparties. Counterparty relationships between parties effecting digital currency transactions are established on the SEN to facilitate U.S. dollar transfers associated with those transactions. The Bank provides digital currency investors that are prospective Silvergate clients with the identity of select participating digital currency exchanges and mutually agreed counterparties are identified as such during the Bank’s SEN enrollment process.

SEN transfers occur on a virtually instantaneous basis as compared to electronic funds transfers being sent outside of the Bank, such as wire transfers and automated clearing house, or ACH, transactions, which can take from several hours to several days to complete. Our proprietary, cloud-based application programming interface, or API, combined with our online banking tools, allows customers to efficiently control their fiat currency, transact through the SEN and automate their interactions with our technology platform. Customers value this around-the-clock access to U.S. dollar transactions and further benefit from the SEN’s powerful network effects—as more users join the SEN, its value to existing digital currency exchanges and investor users increases dramatically. These compelling technology tools and the corresponding network effect have enabled us to attract many of the digital currency industry’s largest and most reputable market participants as customers. This growth has driven $12.7 billion of SEN transfers across the network in the first two quarters of 2019, an increase of 374% over the comparable period of 2018. 

What does that mean? Well, the way I like to think of how banks work is, a bank keeps a list of who has how many dollars. If I have $200 in my checking account at Chase, what that means is just that Chase has a list on its computer, and next to my account number it puts in “$200.00.” If I want to pay you $50, and you have an account at Chase, Chase subtracts $50 from the number next to my account number on its list, and adds $50 to the number next to your account number. But it can, even now, take some rigmarole to get Chase to update its lists; we might both use a payments app that is plugged into Chase’s system, but on the other hand I might write you a check and it’ll take a week for you to get the money. If you don’t have an account at Chase, there is more rigmarole: Chase subtracts $50 from my account, but then it goes to the Federal Reserve, which keeps a list of how much money banks have, and tells it to subtract $50 from Chase’s entry on the Fed's list and add $50 to your bank’s entry, and then your bank adds $50 to your entry on its list. That can take time. 

But if all the crypto exchanges and investors bank at Silvergate, then they can just send each other money by adjusting their entries on Silvergate’s list, without getting the Fed involved. SEN is an interface that lets crypto exchanges and investors send each other money instantly by modifying Silvergate’s list of who has dollars in their Silvergate bank accounts. Using the SEN, crypto investors who bank at Silvergate can transfer money to each other instantly, and can even write computer programs using the SEN API to transfer money automatically.

Which … doesn’t that sound like something you’ve heard of? Like, if you are a crypto investor or exchange, isn’t it because you believe in the power of cryptocurrency to displace old-school centralized payments infrastructure and allow people to transfer money to each other freely any time? Perhaps using smart contracts and programmable money? On, famously, the blockchain? Somehow crypto exchanges have gotten together and solved the problem of efficient instantaneous automatic money transfers! By all banking at the same bank! Oops! Next year on the blockchain, I don’t know.

Oh also stablecoins. The basic way to think of stablecoins is that they are (1) money-market funds (2) that don’t pay interest but (3) on the blockchain. A fact about stablecoins is that some banks seem not to want to get involved with stablecoin issuers because … uh … you know … all the stuff. Because some stablecoin issuers have legal troubles! Because banks may not trust some stablecoin issuers to do proper know-your-customer checks! Because of worries that stablecoins might be used for money laundering or sanctions evasion or terrorism financing or whatever else you’ve got!

On the other hand, you know, vast pools of non-interest-paying money. So here’s Silvergate:

We believe that the continuing adoption of stablecoins, as well as our deep relationships with many of the leading developers of stablecoins, presents a large opportunity for our business in several ways. We focus on stablecoins that are backed by U.S. dollars in a one-to-one ratio to their digital representations and that are offered by regulated institutions who agree to third-party audits. At scale, many believe fiat-backed stablecoins will be the catalyst for widespread adoption of digital currencies as a medium of exchange, allowing consumers to purchase goods and services using digital currency without the friction that exists today within the global banking environment and the volatility that exists in the digital currency markets. Additionally, institutional investors are looking for more efficient ways to move capital between global exchanges that are not currently part of the SEN. The Company does not have any direct involvement with stablecoin pricing, transactions or exchanges.

We participate in this market segment in several ways. We work with all of the U.S. regulated U.S. dollar backed stablecoin issuers, each of which uses the SEN in the issuance and redemption of their respective stablecoin tokens. In addition, we hold U.S. dollar deposits that back multiple stablecoins. In many cases, investors are moving fiat currency onto exchanges in order to buy stablecoins using the SEN, increasing the utility of the network and ultimately expanding the opportunity to earn fees commensurate with the value of our service. In the aggregate, we believe our stablecoin related activities represent growth opportunities that could further increase our deposits and revenue.

If a stablecoin is just a deposit in a bank that doesn’t pay interest and lives on the blockchain, then it is obviously desirable to be the bank that holds the deposit and doesn’t pay interest on it. Someone else can put it on the blockchain.

Should index funds be illegal?

No, say index fund managers:

The biggest U.S. money managers are working to persuade Washington that they don’t harm competition among corporations as a debate about their sway has caught regulators’ attention.

BlackRock Inc. and Vanguard Group have been getting in front of officials and disseminating research in an effort to quash concerns that large institutions’ holdings of multiple companies in a single sector hurts competition. ...

Many in the asset-management industry, government and academic circles say there isn’t clear evidence of whether common ownership causes anticompetitive effects.

Well, they would say that, wouldn’t they. Not only because it is in their economic interests not to be regulated as giant anticompetitive trusts, but also because they really don’t experience themselves that way. Nobody at BlackRock or Vanguard spends their time telling the companies they own to raise prices and stop competing. If the theory that index funds are bad for competition—which we often discuss around here, and which is getting some traction with regulators—is true, it is true as an unintended emergent consequence of the index-fund industry, not as anything that they intentionally set out to do. They’re as surprised by all of this stuff as anyone.

Things happen

Biggest U.S. Milk Company Dean Foods Files for Chapter 11 Bankruptcy. “Blockchain is dead.” SEC Wins Jury Trial in Layering, Manipulative Trading Case. Chrysler Prodded Dealers Over 40,000 Cars They Didn’t Want. Shares in students: nifty finance or indentured servitude? Alibaba’s Secondary Hong Kong Listing Gets the Nod. Regulator Considers Adjusting Rules Guiding Audit Firms on Quality Control. Steel Royalty No More, Thyssenkrupp Sells Itself Off to Survive. Hippo Bathtub Fetches $4.3 Million, Reaps Whopping 2,500% Return. How to Buy Drugs. 

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!

[1] Don’t worry about whether the loan is recourse, etc.; just imagine that the fund itself is levered and can’t go below zero. You can lose everything you invest, but no more.

[2] Realistically there is probably some sort of knock-out barrier where, if the stuff falls too close to the call “strike,” the bank can terminate or restructure the trade—what would be called, in non-options terminology, a “margin call.” 

[3] This is very schematic and wildly oversimplified and definitely not tax advice!

[4] Well it's bank holding company; the bank it owns is called Silvergate Bank. Patrick McKenzie noticed Silvergate’s prospectus and discussed it on Twitter, and some of what I say in this column is informed by his discussion.

[5] For comparison, JPMorgan Chase & Co. paid an average of about 0.95% on its liabilities, which is still pretty good, but that’s mostly because rates are low. It paid an average rate of 1.25% on its interest-bearing liabilities, which were about 80% of liabilities.