ideally, in addition to the carry, you want to find a bond that is theoretically cheap. then, in addition to the carry, you'll also see the convert richen relative to the stock.the risk is that your delta isn't correct. it's not very exact. also, you've got exposure to interest rate, volatility and credit spread risk.

Convert arb is not risk-free, because it is a gamma strategy where the time-decay (theta) works against you. As cometopapa already says, you could actually lose money if the company goes down (credit risk) ... Converts issued by such companies as Level-3 Communications, Double-Click and other dot-coms in 2000 are examples ...

if i am correct the convertible market was a buyers market for decades. companies that issued converts had to do that at relatively too low vola levels for the embedded option. that changed at the end of the nineties with convertible arbitrage funds adding up to 10% of the outstanding issues in converts. rough numbers in my head are about USDbn 10-15 in convert funds, leveraged between 2 and 4, and about USDbn 600 for the convert volume outstanding. i think the game has changed for the convert funds. i thought three years ago that their time has gone, but they had a very good year in 2003. so i was obviously wrong. i think these people are really great in the way that they need to know about equity, fixed income and credit and be be able to model them in conjunction. hats off. this is quant work.peace