I was thinking a little about what an anti-dilution provision on a convertible might look like. And about what the short position might look like for the convertible buyer. Some thoughts (with round numbers to illustrate):
Suppose the company wanted to sell $100 million of a 5-year convertible note at a $7.50 share price, with a 3.00% coupon. The total shares they'd convert to is 13.33 million.
Further, suppose the company wanted to buy an option to buy back those shares at $7.50. They're buying a call option and the bond buyer is writing a call option. If the company pays upfront 20% of the strike price of $7.50 ($1.50) for the option, that brings the bond buyer's net investment to $80 million. Maybe a good way to think about it is like an original issue discount. It brings the bond buyer's annualized return (assuming the bond is repaid in full) to approximately 8.00% (vs 3.00% in the base assumption). This is because the $3 million coupon is more like a 3.75% yield on a net investment of $80 million, and because the buyer gets $100 million back at maturity.
The question is now what such a bond-buyer's motivation is to short. On the one hand, taking a 13.33 million share short position is effective insurance against a default. The cost of carrying the position might be viewed as premium on a credit-default swap.
But, in such a scenario, if everything goes well and the company repays the debt, converts to 13.33 million which they then buy back at $7.50 (effectively making the convertible look a lot more like a regular, higher-priced cash bond)... how does the bond buyer close the short position?? This is the part I'm stuck on.
Leaving aside the whole using shorts as hedge vs default thing for now,
What's the difference between your idea - holding a convertible bond + short call options at the same strike (presumably European options?) vs just holding a plain old bond?
It seems they simply cancel out.
In other words a convertible bond like a combination of a regular bond and a call option. write an identical call option and you're left with just the bond.
I think that's exactly right. But the convertible bond gets the embedded call "for free." So, turning around and selling those rights back are just like getting a discount on the initial bond purchase (i.e. juicing the bond yields).
Not sure why someone would prefer to do this over just taking a higher-yield bond. That's why I think the convert + short position makes sense. I'm just not sure where the call option that gets sold to GTAT fits in without leaving a net long (or short) position.