So, You Wanna Be An Arb?

If you thought the fun and games with JDS Uniphase (JDSU Quote) were over now that the optical-networking superstar has been inducted into the S&P 500, think again. With its proposed merger with SDL (SDLI Quote), the games are just beginning.

To play the S&P 500 addition game with JDS Uniphase last week, you had to think like an index-fund manager. Now, to play the JDS Uniphase/SDL merger game you have to think like an arbitrager . Arbs are those shadowy traders who make their living betting on whether announced corporate mergers get completed -- and not all of them do, especially in today's climate of antitrust activism at the Justice Department.

The stakes are high in the betting on the JDS Uniphase/SDL deal. If the deal goes through, the arbs stand to make over 80 points per share on SDL. Across SDL's total shares outstanding, that's $6.3 billion dollars -- just lying there on the sidewalk, waiting for someone to dare to pick it up.

Here's how it works:

If the merger is completed, JDS Uniphase will give 3.8 of its shares for each share of SDL. Based on Wednesday's closing price of 112 5/8 for JDS Uniphase, that means a price of about 427 15/16 for SDL. Wednesday, SDL closed at 345 1/8.

The difference of about 82 7/8 is the opportunity -- the "arb spread." If the arbitragers all thought that this deal would go through without a hitch, the arb spread would be near zero. But when the arb spread widens out, it's because the arbs think there's a good chance the deal won't go through.

The arb spread is wide today -- there's a lot of skepticism about this deal. But the arb spread has been even wider. Here's a chart of the arb spread since July 10, the day the deal was announced. By July 26 -- the day JDS Uniphase was added to the S&P 500 and reported earnings -- pessimism about the deal peaked with a spread of 126 1/4.

If an arb thought the deal would go through, here's how he'd capture the arb spread. He'd sell short 3.8 shares of JDS Uniphase and buy one share of SDL.

At Wednesday's prices, that trade would start him out with a credit in his account of 82 7/8. If the deal goes through, the arb would tender his SDL for 3.8 shares of JDS Uniphase. The resulting long position in JDS Uniphase would cancel out the short position, and the trade would be done -- and the arb would keep the 82 7/8 in his account.

Let's say you're not an arb, but you are long JDS Uniphase. You could do a version of the arb's trade that amounts to the same thing. Sell your JDS Uniphase and replace it with SDL; by buying one share of SDL for every 3.8 shares of JDS Uniphase that you sell. The trade will put 82 7/8 in your account for every 3.8-to-1 unit you do.

If the deal is completed, you tender your SDL and you'll get all your JDS Uniphase back -- and you keep the 82 7/8.

Or there's even a simpler play. If you're comfortable with the market overall and the optical-components sector in particular, just simply buy SDL. You know there's another big buyer out there who, at least nominally, is willing to pay 82 7/8 higher. That's not a bad edge to start a trade with.

And of course there are all kinds of fancier ways you can do these trades with options.

And Now For the Risks ...

Bad things happen to these trades if the deal doesn't go through. Obviously the spread wouldn't be so wide if the arbs didn't think there was some pretty serious risk for which they needed to get compensated.

There's no way to really know the true worst-case risk in these trades. But one clue might be to look at the relationship between the prices of JDS Uniphase and SDL before the merger was announced on Monday, July 10 -- it's not too bad a guess to think that the spread might revert to there if things comes unglued. On Friday, July 7, JDS Uniphase was at 116 3/16, and SDL was at 295 5/16 -- that makes the no-deal spread about 146 1/4.

Now, What Can Go Wrong

But there's a kicker that makes the estimate of the worst-case scenario slightly optimistic: A tidy little $1 billion break-up fee payable by SDL to JDS Uniphase, if SDL walks away from the deal and merges with someone else. That's about 13 points per share of SDL. But it's not payable if the deal busts for antitrust reasons, so let's bake only half of that into our estimate of the no-deal spread, which makes it about 152 3/4.

That means if you do the trade today at an arb spread of 82 7/8, and then the deal comes undone and the spread moves to 152 3/4, you'll be unwinding the trade at a loss of about 69 7/8.

That's the bet: Risk 69 7/8 to make 82 7/8. That's a risk/reward ratio of a little better than 1-to-1. So if you think there's a better than 50/50 chance of the deal going through, then you should do some version of the arb trade.

By the way, note that I said you risk 69 7/8, not that you invest 69 7/8. When an arb buys SDL and shorts JDS Uniphase, he's not spending any money up to do the trade. He just posts securities as a margin, and the proceeds from the short sale more than cover the cost of the long. So whatever gains or losses he makes represent a return on an investment of zero.

The same is true if you sell your pre-existing long position in JDS Uniphase and replace it with SDL -- you're not putting any new money on the table: In fact, you're taking money off the table.

Let's Talk Odds

That leaves the $6.3 billion question: Will the deal go through? The spread is telling us that the market thinks there's less than a 50/50 chance it will. Is the situation really that pessimistic?

According to Mark Langley, an analyst with Epoch Partners who covers both JDS Uniphase and SDL, the deal faces serious risks. Langley thinks the deal would be a big win for both companies, and that acquisition-savvy JDS Uniphase would make the post-merger integration a "smooth transition." But he is concerned that the Department of Justice will throw difficult antitrust barriers in the way of the deal, either aborting it outright or forcing JDS Uniphase and SDL to make painful divestitures.

A combined JDS Uniphase/SDL would enjoy 80% market share in the key 980 nanometer pump chip market; JDS Uniphase's share is more than 50% already. What's a 980 nanometer pump chip, you ask, and who cares? A pump chip is a semiconductor the size of a grain of salt, and it's a critical component for fiber-optic networks, which allows light signals in the fiber to be amplified and regenerated without being slowed down by normal electronic processes.

By weight, 980 nanometer pump chips may be the most precious material in the world. A year's production of these beauties in JDS Uniphase's Zurich factory is worth hundreds of millions of dollars, and the whole thing could be shipped inside a single coffee mug!

Langley thinks the pump chip business is a "crown jewel," and says it's unlikely that JDS Uniphase/SDL would ever divest it. So the antitrust hawks at the Department of Justice might look for other divestiture targets -- perhaps PIRI, the "arrayed waveguide" maker that SDL acquired earlier this year at bargain prices, while JDS Uniphase was distracted in negotiation with the DOJ over its E-TEK merger.

Langley won't let himself be pinned down on the odds of the deal going through. He says, "The deal is a good deal for the companies and very possibly for their customers." But he remains very cautious about the significant regulatory hurdles ahead.

My Take

In my own opinion, the probability of the deal going through is somewhat worse than the 50/50 necessary for you to want to do the arb trade. I'm concerned that this troublesome megamerger may be a bridge too far for JDS Uniphase's new CEO Jozef Straus. And I'm positively paranoid about the Department of Injustice, which in my opinion is on an irresponsible high-tech antitrust rampage.

Call me cynical, but the JDS Uniphase/SDL merger -- the biggest high-tech merger ever -- would make a lovely scalp to hang from Joel Klein's belt.