Sometimes it's useful to be reminded that not all financial structuring is designed to get around capital requirements or defraud customers. Some is designed to get around taxes and defraud the treasury! One group of people who like to think about that kind of thing is the Congressional Joint Committee on Taxation, who took some time out from shouting about payroll taxes yesterday to have a geeky hearing about the state of financial instrument taxation. Short version is, they're not all that happy about it.
The JCT staff generally say pretty smart stuff about tax policy, and they get that shit is fucked up and bullshit. Or as they put it more diplomatically:
The timing, character, and source rules apply differently to (and are sometimes uncertain for) equity, debt, options, forward contracts, and notional principal contracts. These five basic instruments can be combined in various ways to replicate the economic returns of any underlying asset. ... The flexibility of financial instruments also creates great difficulties in the taxation of financial instruments. This report provides examples of taxpayers’ uses of financial instruments to achieve desired timing, character, and source outcomes and describes how the tax laws have or have not addressed this tax planning.
There are two useful takeaways here. One is kind of weird: there are a bunch of fairly basic things (exchange-traded notes, CDS) where nobody - not the IRS, not the JCT, nobody - knows how they're supposed to be taxed. That ... seems like a bad thing. And, I'm going to guess, not so much the fault of evil financial innovators.
The second takeaway, which is related but more satisfying to fulminate about, is that evil financial innovators can mix and match stuff until they get any tax result they want:
“Because our system of taxation has no basis in the reality of economics, sophisticated taxpayers are free to choose a tax treatment that minimizes their taxes, and choose they do,” said David Miller, a partner at Cadwalader, Wickersham & Taft LLP in New York.
Really "any tax result they want" is too strong, but as a basic takeaway it's not a bad one to be left with. And, as I spent some time in a former life doing that mixing and matching, I was pleased to encounter some old friends in the JCT report. Here's one that I don't necessarily recommend reading all the way through:
Feline PRIDES are a complex instrument sold by an issuer to raise capital. For Federal tax purposes, the issuer seeks to effectively issue stock (without doing so currently) while generating current interest deductions. The following simplified description of feline PRIDES is based on the facts of Revenue Ruling 2003-97 in which the IRS ruled that interest accruing on a feline PRIDES-like instrument was deductible for tax purposes.
Feline PRIDES are two instruments packaged together into a single investment unit. The instrument consists of a three-year forward contract to purchase the issuer’s common stock and a five-year note paying interest. The forward contract obligates the holder to purchase, and the issuer to sell, an amount of the issuer’s common stock in three years. The amount of common stock to be purchased is determined by reference to the market price of the stock on the settlement date three years in the future. The note obligates the issuer to pay a sum certain in five years. The five-year note serves as collateral for the holder’s obligation under the forward contract.
Now, "feline PRIDES" is surely the greatest name ever given to a financial product.* It also happens to be maybe my favorite financial product ever, for reasons that are too geeky to get into even in this post, but I'm tearing up thinking about it. To tragically oversimplify, feline PRIDES are a way for companies to issue tax-deductible equity. Since dividends on equity aren't normally tax-deductible, that's a plus. As you'll note from the JCT text, not only do companies engage in this dastardly shenanigan to cut their tax bill, but the IRS specifically said they could do so in a 2003 ruling. Not - probably - because the IRS is evil and captured by bankers, but because as a piece of technical mixing and matching, it works. You take one piece of debt, which has deductible interest, you add an equity forward, which if done right has minimal tax consequences, you glue them together, you sand off the rough edges, and presto, you've created tax deductible equity.
There's no shortage of people who think this is a Bad Thing, and I guess it is kind of silly: a lot of time and money is spent on designing this kind of stuff. I used to get a small slice of that money. Good times. In any case, the Joint Committee talked to a bunch of experts who worry about this stuff, and got answers that mostly ranged along the "maybe study it some more" spectrum, plus Miller, the Cadwalader tax partner, who wants mark-to-market taxes on derivatives because ... well, let him tell you, in the form of a cosmic prose poem:
Claudius Ptolemy was the second century astronomer who created a model of the heavens that predicted the positions of the sun, moon and planets, and was used for over 1,400 years.
There was just one problem with his universe: The earth was in the middle.
How could a model that was just plain wrong provide sufficient accuracy to be used by Western civilization for over 14 centuries?
With a great deal of complexity.
To explain and predict heliocentric planetary patterns in a geocentric model, Ptolemy’s planets traveled in a series of ellipses or epicycles around the earth. But this alone was insufficient. To correct further, Ptolemy had the planets move closer and then further away from the earth, and even slow down and reverse in their orbits.
Our federal system for taxing financial instruments is truly Ptolemaic. As Ptolemy’s system was geocentric, our federal tax system is based on the equally archaic system of realization – the concept that income is not earned, and therefore not taxed, until a taxpayer actually sells property for cash or exchanges it for materially different property.
Which is probably right. For myself, I'm not sure a mark-to-market approach will avoid epicycles either. Warren Buffett seems to be with me.
If your favorite tax isn't the payroll tax or the Ptolemaic tax, maybe it's the Tobin tax or its more swashbuckling friend the Robin Hood tax, which would impose just a tiny little tax, honestly you'll hardly notice it, on any trading of financial instruments. John Carney is not a fan:
Financial transactions are very likely highly elastic to changes in taxes. Many transactions, in fact, are driven to minimize current tax implications. If you change the tax code to increase taxes, the participants in financial markets will change their behavior, but not necessarily in ways that benefit society. ...
Financial transactions will also tend to become more synthetic, more like the kind of half-imaginary collateralized debt obligations constructed just before the collapse of the mortgage market. Because the construction of a synthetic CDO does not involve actually purchasing mortgage assets — merely indexing a new financial asset to an existing group of mortgage assets — it would receive favorable tax treatment. So the wizards of Wall Street would sell synthetic everything — synthetic equities, bonds, futures, options, commodities. This would make financial markets less transparent and financial institutions more vulnerable to failure.
That seems plausible to me, but shouldn't be overstated. The JCT hearing is both evidence for his point - that structurers are often better at obtaining the tax results they want than regulators are at preventing them - and evidence that it may not matter. If your policy goal is (1) to raise money and/or (2) to stop financial risky business, then the Robin Hood tax is likely to fall short because market participants will structure around it in nontransparent and risky ways. But as experience shows in other forms of tax policy - and, for that matter, other ways of reining in financial system risk - that's true of pretty much anything else you try too.
U.S. Studies Derivatives That ’Game’ Tax Rules [Bloomberg]
The Robin Hood Tax Won’t Work [NetNet]
JCT hearing witness list [house .gov]
* It's also a Merrill Lynch trademark, and the sort of term that's only really used by tax authorities now, because it sounds so obviously shady if a banker says it. It's like, "hi, please buy my Death Star CDO, what could possibly go wrong?" It's not even used any more in deals done by the ghost of Merrill. But "feline PRIDES"! Beautiful. I hope it stands for something.