Level 3's anti-dilution hedge

I have been asked to comment on the anti-dilution hedge I mentioned in my Virtuous Circle post last week. I’ll do my best, but your mileage may vary.

In December of 2004, when Level 3 issued $345M of 5.25% converts due 2011, they also bought from Merrill Lynch what they called an ‘anti-dilution hedge’ tied to those bonds. What exactly this meant has always been a bit fuzzy, the only real details to be had were deep in SEC filings and written entirely in lawyer-speak. Over the years, for some reason that still escapes me, I have made several attempts to work my way through it, with limited success. However, with the expiration date on the transaction approaching this December, I think I finally understand it. So I will explain, and wait for someone more in touch with lawyer-speak to correct me.

The important bits distill to the following

• Level 3 can acquire up to 86.68M shares of its own stock from Merrill Lynch at a price of 3.98.
• Merrill Lynch can acquire the same amount from Level 3 at a price of $6.
• Prior to December 15, 2008 nothing happens unless the bonds convert first, and the bonds aren’t callable until then.
• On December 15, 2008, the hedge gets settled whether the bonds convert or not
• Level 3 can choose to satisfy both legs of the with either cash or stock or a combination, but both legs with the same choice. Prior to December 15, 2008, this choice is tied to that made for the conversion of the debt, at expiration however it does not appear to be tied to a specific conversion event.

To simplify and to increase our understanding, let’s assume that the transaction happens all at once in December rather than piecemeal. So there are three price ranges, and we can dispense with two of them quickly:

LVLT < $3.98

This is the easy one. Both legs of the hedge expire worthless, the convertible bonds are unaffected and not called and converted until perhaps a later date when the stock is higher.

LVLT > $6.00

It turns out that it doesn’t matter whether the cash or stock option is chosen here because the stock cancels out, the result in both cases is that LVLT gets $175M in cash from Merrill and walks away. The bonds are likely called and paid with shares – LVLT has the option to pay cash but that would be highly unlikely. So dilution happens, but the company comes out richer.

$3.98 < LVLT < $6.00

This is a bit tricky, but understandable if we break it down. Level 3 can choose a cash or settlement option, let’s discuss each separately:

Cash settlement

Merrill Lynch pays LVLT the difference between the stock price and 3.98 on 86.6M shares, which pays on a linear scale from $0M to $175M. The convertible bonds likely get called, almost certainly converted to stock. The end result is similar to the >$6 case, dilution is unmitigated but LVLT gets paid extra with the amount depending how far above $3.98 the stock price is.

Stock settlement

LVLT pays Merrill Lynch $345M in cash, Merrill Lynch has to acquire 86.6M shares in the market and give them to LVLT. The entire idea behind this was to allow LVLT to pay off the convertible bond holders with those shares. Obviously, with cash levels what they are LVLT isn’t going to pay $345M cash for these shares unless they raise the money from the debt markets. So effectively, this option represents LVLT’s ability to refinance the convert for $345M in cash in lieu of dilution.

That is where the ‘anti-dilution’ part of the description came from, it is the only case where dilution forestalled. But this option is increasingly unlikely this year, given the chaos in the debt markets and the need to conserve dry powder for refinancing other debt in 2009 and 2010. Therefore I believe we should see the hedge simply as a possible cash windfall: if they can get the stock above 3.98 and hopefully as high as 6, then they get free money. If they don’t, then they don’t get a penny.

Obviously there is money at stake here! LVLT would really want the cash, they have a powerful incentive to get the stock to $6. Likewise, I don’t think Merrill wants to pay LVLT a pile of cash, especially in the current financial markets, and they have an incentive to keep it down below $3.98 if possible.

Let me say however that I think this whole thing was a weak idea on LVLT’s part. It cost them some $62M at the time, and it just wasn’t worth the hassle – they should have saved the $62M and kept it simple.