This picture is a subtle clue to the quiz at the top of this post.

Photographer: Emmanuel Dunand/AFP/Getty Images

Hedge Fund Still Wants Its Tax-Avoidance Profits

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

Here's a passage from a lawsuit:

Mr Patel and Mr Barenys confirmed that MLI would be able to generate a profit for Lycalopex from the DAS, but asked whether a 75% return rather than the 100% return originally proposed (i.e. a profit over the first year of $15 million rather than $20 million) would be acceptable. Mr Patel said words to the effect that: "I can't do $20 million, but I'll definitely get you to $15 million". Mr Devonald and Mr Williams said they would consider this request and revert to Mr Patel and Mr Barenys. The same day, they sought and obtained approval for this reduced profit commitment. They relayed that approval to MLI the following day.

Here's a link to the lawsuit, but don't click it yet. First, I want you to guess: What kind of lawsuit is it? Raj Patel and Angel Barenys of MLI came to Andrew Devonald and Grant Williams of Lycalopex with good news and bad news. The bad news was that MLI couldn't guarantee Lycalopex a one-year profit of $20 million, or 100 percent of its original investment (of $20 million). The good news was that MLI could guarantee Lycalopex a $15 million profit, or 75 percent. So: Who are these people, and what were they up to?

If you guessed that the case is an enforcement action, that MLI is a Ponzi scheme, and that Lycalopex are the poor naive marks who were duped into thinking that a guaranteed 100 percent (or 75 percent) one-year return is a thing that exists in the world, then, well, that is kind of what it looks like. But that's not what it is! In reality, MLI is Merrill Lynch International, a real investment bank (and the U.K. subsidiary of Bank of America). Lycalopex is an equity-trading joint venture between Tudor Investment Corporation -- "one of the world’s oldest hedge funds," and a pretty sophisticated one -- and Vulpes Investment Management.  The case is a contract dispute. And MLI didn't come to Lycalopex with the idea of a guaranteed 100 percent return. That was all Lycalopex's plan. And, says Lycalopex, it was totally reasonable:

Vulpes and Tudor would each commit $10 million of capital to the Fund and were seeking a 100% return on their investment over the 2012 dividend season (i.e. a total profit of $20 million for the Fund), in order to fund the next phase of the Fund's operations.

This was a wholly achievable target, based upon estimates carried out by Tudor and Vulpes.

Vulpes, Tudor and, once incorporated, the Fund, needed a firm commitment from BAML that it would, in return for the Fund agreeing to trade with BAML as its prime broker, achieve this profit target for the Fund.

It is all incredibly blasé: Of course they were going to double their money in one "dividend season," so they could move on to the next, more interesting phase of their plan. No problem. Happens all the time.

Lycalopex is now suing because "the strategy only yielded about $4.6 million before Merrill Lynch ended the arrangement." Lycalopex started trading in late April 2012, and stopped in mid-August 2012, with a total profit of $4,630,680 on its initial $20 million investment. That's a 23 percent return in a bit less than four months. A 23 percent return in four months is ... pretty good?

But not good enough, so the Lycalopex investors are suing Merrill Lynch for the difference. They claim that Merrill representatives promised them the $15 million profit in multiple meetings, that these promises constituted a "binding oral agreement," and that Merrill ought to be held to its agreement:

As a result of MLI's breaches, Lycalopex lost the difference between the profit it was promised from the DAS by MLI for its first trading year, namely $15 million, and the profit it actually achieved from the DAS in that year, $4,630,680. Lycalopex has therefore suffered loss of $10,369,320.

So this is an odd lawsuit. Prime brokers are not generally in the habit of making binding promises of guaranteed profits in meetings with customers. In fact, prime brokerage agreements tend to contain boilerplate disclaiming any promises outside of the written agreement. The prime brokerage agreement between Lycalopex and Merrill Lynch included such a clause, but Lycalopex argues that, basically, it doesn't count.  “This is a spurious claim based on the inaccurate premise that a significant level of profit was guaranteed,” says a Bank of America spokesman.

But that comes nowhere close to exhausting the oddity of the lawsuit. The trading that was supposed to generate the guaranteed profit was a dividend arbitrage strategy (which Lycalopex refers to as "the DAS"). Some French stocks pay dividends, and France imposes a withholding tax on those dividends. But France has a tax treaty with the United Arab Emirates providing that UAE residents are exempt from withholding. So Tudor and Vulpes set up Lycalopex in Dubai, to avoid withholding. Then the plan was that Merrill Lynch would borrow a whole mess of French stocks from its customers and sell them to Lycalopex before a dividend payment, with an agreement to buy them back (at a fixed price) after the dividend payment. Lycalopex would get the (full) dividend, and would take no risk on the stock price. Merrill would owe its customers an amount equivalent to the (net, after withholding) dividend. The tax savings were guaranteed profit, which Merrill and Lycalopex would split.  

The margins were thin, though, which is where Merrill's alleged promises as prime broker came in. For one thing, it had to borrow the right stocks to sell to Lycalopex temporarily: High dividend stocks would be more profitable than low dividend stocks. For another thing, it had to borrow a huge mess of stocks to sell to Lycalopex: The more of this trade you do, the more money you make. And, finally, Lycalopex only had $20 million, which was not enough to do this trade in the size it wanted. So Merrill allegedly agreed to lend it up to $1 billion to do these trades. Lycalopex's essential complaint is that Merrill didn't do enough of this: It didn't trade enough stocks with Lycalopex, didn't provide enough leverage, and the trades that it did were on stocks with dividends that were too low.

You may have heard of dividend arbitrage, because it is ... super controversial?  It has been more or less banned in the U.S., but it is still feasible in many other countries, and those countries often get quite mad about it. It is a pure tax arbitrage; the trade called for Merrill Lynch and Lycalopex to take millions of dollars from French taxpayers and split it among themselves. It added nothing to the efficiency of the markets; it did nothing for price discovery or capital formation.  "We're arbitrage traders, stocks to us are just tickers," said Lycalopex's Devonald. Barely even that. This wasn't a stock-trading strategy; Lycalopex took no market risk, just shuffled ownership around briefly and symbolically. The stocks were just a vehicle for tax arbitrage, just a way to pump money out of the French treasury and into Lycalopex.

Also Merrill was going to leverage this strategy 50 to 1, letting Lycalopex buy as much as $1 billion of stocks with $20 million down. Really, that's fine; remember, there was no equity risk involved, so Merrill's billion dollars were pretty safe. Still it sort of adds to the grim optics. Merrill was going to let a client do a billion dollars of tax-arbitrage trades at 50 to 1 leverage. It just sounds bad.

And it actually got worse, so bad that Merrill eventually backed out. The Lycalopex lawsuit is surprisingly frank about why Merrill ended the trade early: It got cold feet around compliance issues. France introduced a financial-transactions tax in August 2012, which would make the trade much less profitable. Merrill and Lycalopex had discussed ways to avoid that tax too, and Lycalopex thought it had a good one, but it was all too much for Merrill. "At around this time MLI refused to execute further DAS trades with Lycalopex on the purported basis that a suitable new trading structure could not be found to accommodate implementation of the FTT," says the lawsuit. After that attack of conscience on Merrill's part, Lycalopex sued.

You don't see a lot of lawsuits like this. The moral of this story, according to Lycalopex, is that Merrill made certain promises about stock selection and leverage that it didn't fulfill, and now it has to pay. Thus the lawsuit. 

But most people who hear this story will take away a very different moral. Their conclusion will be: A powerful hedge fund teamed up with a global investment bank to do a billion-dollar tax evasion trade at 50 to 1 leverage, a trade that involved no risk, that merely shuffled share ownership on paper, and yet that was so lucrative the hedge fund planned to double its money, risk-free, in the first year.

That is the stuff of conspiracy theory! Not even the wildest Occupy Wall Street fantasies imagine a financial system that is quite so nakedly rigged in favor of the rich and powerful. If you are in the sort of business where you come across risk-free multimillion-dollar opportunities to make 100 percent (or 75 percent) annual returns at the direct expense of national tax authorities, you might want to keep quiet about it. Even if it goes wrong sometimes, just eat the losses -- or the disappointing 23 percent profits -- and move on to the next trade. Airing this sort of dirty laundry in court runs the risk of ruining it for everybody.

  1. The signature page is dated April 25, but the documents were just "released by the court on Tuesday."

  2. Though, weirdly, last time we talked about Lycalopex, we learned that its "partners in Dubai no longer pursue the investment strategy and instead sell packaged, halal-certified meals to refugee-aid organizations and supermarkets." (That means that "one of its Dubai-based executives" is now doing that. Not, like, that Tudor has pivoted to selling packaged foods.)

  3. Or like 87 percent annualized, though the annualized number isn't quite fair because Merrill Lynch cut things off early. Also both profit numbers are unfair because Lycalopex's "total establishment and operational costs were approximately US$8.8 million." So I guess it had an all-in net loss, despite its 23 percent trading profit.

  4. From their claim:

    By clauses 19 and 21 of the PBA, Lycalopex purportedly gave a number of "acknowledgments" concerning its relationship with MLI and whether any representations were made by MLI prior to the execution of the PBA. For the avoidance of doubt, these acknowledgments are not inconsistent with, and do not negate, the existence of the February and/or April Profit Agreements.

    Those "Profit Agreements" are what Tudor and Vulpes call the oral statements by Merrill Lynch employees allegedly guaranteeing the $15 million profit. In particular:

    Clause 21(h) does not apply because Lycalopex des not allege to have entered into the PBA "in reliance" on a representation, warranty or other statement made by MLI.

    But ... doesn't it?

  5. I say "split" loosely; I don't know how much Merrill expected to make as prime broker. Though its tax structuring group was quite lucrative in general.

  6. E.g.:

  7. I know, like five of you are writing me e-mails saying "but avoiding unnecessary dividend taxation does increase efficiency." Sure, fine.

  8. In this version of the story, the initial 100 percent return dominates the ultimately agreed 75 percent return (never mind the achieved 23 percent return).

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 mlevine51@bloomberg.net

To contact the editor responsible for this story:
James Greiff at jgreiff@bloomberg.net