Hi! I am still here. I was once a banker and now I write for Dealbreaker and answer your questions about banking and whatnot. You can send questions to firstname.lastname@example.org with "ask a banker" in the subject line, or ask on Twitter (@planetmoney).
Last week I waxed lyrical and long about derivatives, which are basically contracts that allow people to change their payoffs in potential future states of the world. This prompted a number of comments along the lines of what D F wrote:
Q. Or to summarize: It is a bet. It's like Vegas, but with more graphs and pie charts ... and we use your Grandma's pension.
This is a comment, not a question, but let's take it as our question for this week. (Let's also ignore the base slander about pie charts. Pie charts! I would never!)
The answer to your comment is: sure! Sure, derivatives are bets. In some simple sense anything you do to change your possible future can be called a "bet" — if you invest in my lemonade-stand startup, you are "betting on my success" — but one common and sensible use of the word "bet" involves zero-sum-ness.
If I bet you $100 that the Jets will win this weekend, then the Jets will win and I will have $100 more and you will have $100 less, or the Jets will lose and you will have $100 more and I will have $100 less, and the Jets don't care either way. We've just shifted money around between ourselves; we've done nothing to change the outside world.
There are some important clarifications; for instance, banks are much less in the business of making bets (gambling) than they are in the business of taking bets from customers (bookmaking). There are good reasons to believe that the act of betting improves welfare just by letting people tailor the world to their risk preferences.
Your Grandma's pension, to some first approximation, is probably not currently being gambled on complex credit derivatives, though I can't really promise you that and you might want to look into it.
Most importantly, banks and bankers are not really that troubled by this; lots of people believe that gambling is immoral but most of those people just don't work at investment banks. The people at investment banks tend to be enthusiastic poker and blackjack players. But here is a slightly more troubling fact: Gambling is illegal.
This is not entirely true; you can gamble in Vegas or Atlantic City or various Indian casinos; you can play the state lottery; you can probably even bet on football or poker with your friends at your house without going to jail, though I'm not your lawyer and if you are arrested I'll deny ever having written this. But more or less if you and I write a contract saying "I bet you $100 the Jets will win this weekend," and they win and you don't pay me, and I sue you, courts will throw my case out because gambling contracts are not enforceable.
Derivatives contracts mostly are.* That's important! Banks, by the magic of derivatives, transform a thing that doesn't work under the law — "betting" — into something that does work — "hedging" or "speculating" or whatever.
And that brings us to an important point about derivatives, and the financial industry generally, that you are not going to like, D F. You probably don't like it already.
What we talked about last week was using derivatives to shift economic outcomes in different future states of the world: giving me $100 more or $100 less dollars depending on whether the Jets win this weekend. But in practice derivatives are often used to accomplish something slightly different.
Derivatives let you shift economic outcomes, yes, but they're equally important for shifting outcomes of legal and regulatory regimes - like, just to start with, the gambling laws. Derivatives are tools for hacking the tax code, or the securities laws, or U.S. generally accepted accounting principles. This is called "regulatory arbitrage."
Some examples. A very simple sort of derivative is the "equity total return swap." This is a contract in which a bank agrees to give you money if a certain stock goes up, or pays dividends, and you agree to give the bank money if the stock goes down. So if IBM stock is worth $200 today, and you enter a one-year total return swap on IBM, then if IBM ends up at $220 you'll get $20. If it ends up at $170, you'll lose $30. In any case you'll get the dividends — about $3.40 per share — that IBM pays over that year.
Now, the exact same thing happens if you just buy IBM stock: You make money if it goes up, lose money if it goes down, and get the dividends. So why not just cut out the middleman and buy the stock yourself?
The answer is regulatory arbitrage.** Economically, buying a swap is about the same as buying a stock,*** but formally they are different, and those formalities matter to some people. For instance: tax lawyers.
If you buy IBM stock in your Cayman Islands hedge fund, and it pays you a $3.40 dividend, the IRS holds on to 30% of that dividend, so you only get $2.38. But if you buy a swap, the swap pays you the full $3.40 dividend. You've made an extra dollar and change just by changing the formalities.
Taxes aren't the only thing you could worry about. You might be a hedge fund that wants to influence a company to change something about its business, making its stock more valuable and making you a profit.
If you buy more than 5% of the company's stock, you need to report that publicly, which will alert other people to the fact that something is up. Those people might buy stock to piggyback on your idea, pushing up the price and making it more expensive for you to buy more. But if you buy the stock "on swap" — that is, you don't buy it, but instead buy swaps on it — then you can limit and delay that disclosure, so you can buy up more of the company before you have to report it. Activist hedge funds love this, but companies love it less, so there is much debate over the application of these rules and whether they can be avoided by using swaps.
There are vastly more complex examples. A beautiful one is a trade that SunTrust Bank did with an investment bank in 2008. SunTrust had owned a bunch of Coca-Cola shares for 100 years; the shares were worth well over a billion dollars. SunTrust was looking to raise money to increase its capital: banking regulators want banks to have a certain amount of capital,**** and SunTrust had less than the desirable amount, and one way to raise capital was to sell this stock and get cash for it. But the bank would have had to pay hundreds of millions of dollars in taxes if it sold the stock.
SunTrust was thus in a little bit of a bind: for regulatory capital, it really ought to sell the stock; for tax purposes, it really ought to hang on to the stock and not pay taxes. What if there was a way to do both: sell the stock and raise capital as far as its banking regulators were concerned, but hang on to the stock and not pay taxes yet as far as the IRS was concerned? There was.
An investment bank came to SunTrust and offered them a derivative trade — called a "variable forward purchase agreement" but that's almost beside the point — that would let SunTrust tell its banking regulators that it had sold some Coca-Cola shares at a profit, generating lots of delicious capital, while at the same time telling the IRS that it hadn't sold those shares and so didn't need to pay hundreds of millions of dollars in taxes on them yet.
This was all perfectly legal — both the IRS and the banking regulators explicitly signed off! — and also perfectly delightful. One thing to realize about investment bankers is that they are motivated not only by money but also by fun, and this stuff is just plain fun.
There is great joy to be obtained from understanding a vast and semi-logical body of rules, and then understanding a different vast and semi-logical body of rules, and then figuring out the differences between them that you can exploit, and then building a product to exploit those differences, and then giving it a catchy name. Throw in a dash of math, a heap of jargon and a modicum of graphic design, and you get something that appeals to lots of people who in a simpler and less financialized world would be working to cure cancer, is I guess the cliché, though more likely they'd be designing online role-playing games.
This comes back to your comment, D F, about gambling. The knock on the derivatives business, as I explained it last week, is that it is "just" gambling - just a zero-sum speculative shifting of outcomes that doesn't make the world any better. I don't think that's true! I had hoped to explain why real people in the real world might actually want to change around their possible future states of the world, and why that might be a good thing for everyone. You're free to disagree, though, and you're far from alone.
But the regulatory-arbitrage focus is different. It's harder to see the social benefit here: If Congress in its wisdom wants to tax dividends paid to foreigners, it's not clear why it shouldn't tax quasi-dividends paid to foreigners via swaps. If the SEC wants hedge funds to disclose 5% stakes, it's not obvious why 5% swap stakes are different. Maybe those rules are wrong, by the way — maybe dividends shouldn't be taxed! maybe hedge funds shouldn't be forced to disclose! — and, in that case, these regulatory arbitrage trades do make the world better, though not by as much as just fixing the rules. But as an area of human endeavor it might give you pause.
Does it worry you that a bunch of smart and well-paid people are running around making and taking bets? That's barely the half of it. How do you feel about the fact that a lot of them are focused less on the betting and more on finding ways around tax and accounting and securities rules?
* This is roughly the work of a federal law called the Commodity Futures Modernization Act and its predecessors, which prohibit states from treating eligible commodity and security derivatives as illegal gambling. There are some gray areas about what is a "derivative" and what is a plain old "bet"; for instance "political event contracts" - betting on the election - were recently consigned to the bad side of that line.
** Mostly. More technical discussion that you can safely ignore: swaps are a way to provide levered exposure to a stock, so instead of putting up $200 now, you just enter into a swap contract, effectively borrowing the purchase price from the bank. (Though I have to post margin for a good chunk of the purchase price.) There is some regulatory-arbitrage element here too; you could just buy the stock on margin, borrowing a portion of the purchase price, and some of the choice between doing a swaps and buying on margin does depend on which format offers better margin rules.
About that name, by the way, "regulatory arbitrage." In finance theory, an "arbitrage" is a risk-free profit: if you can buy something for $100 and sell it at the same time in the same place for $101, you've made a $1 arbitrage profit. You mostly can't do that, so the term "arbitrage" is extended to things that feel less risky, to some people, than some other things; they say "we are making arbitrage profits" until they lose all their money. There is a thing called "risk arbitrage," which, as a name, troubles many people. But "regulatory arbitrage" can be a way of making risk-free profits because the profits come from elsewhere - the IRS, for instance.
*** Important way in which they are not: if you buy a swap, you need the bank to be around to pay you. If you buy a stock, you don't need the bank to be around. You can see why there would be times when this would be a big issue. Less important point: nobody would really say "buy a swap"; the phrase is "enter into a swap." There's no "buyer" and "seller," it's a two-way transaction. I'm saying "buy" in the text because it's shorter.
**** Don't feel bad if you don't know what capital is; many bankers don't either. A simple description is that if you take all the stuff that a bank owns, add up its value, and subtract the money that the bank owes to lenders and depositors, what's left over is capital. This simple definition leads to surprising amounts of puzzlement in practice.