Redeeming a bond -- bankers and finance PhDs, please reply

[leftover_salmon]

I was too lazy to find a better forum and sign up, and figured there are more than enough bankers here anyways.

Let's say Company A has a $100MM bond outstanding and wants to redeem it at makewhole. The redemption premium is $10MM (i.e. they have to pay $110MM to take the bonds out). Concurrent with the redemption, Company A will issue a $110MM bond (so they are funding the redemption with a new issue) maturing in, say, 5 years.

My question is, on a PV basis, do I recognize a cost of $10MM or a cost of $10MM discounted back from the new bond's maturity date to today? The latter is obviously more favourable and seems theoretically correct, but I want to double check.

My argument for the latter is that the company only has to actually pay that $10MM down out of pocket in 5 years time - in the meanwhile, they're funding it with a new bond (with the negative effect being an incrementally higher debt load and higher interest payments). From the no-arbitrage perspective, they can effectively sock away $6MM today to repay that $10MM in 5 years' time.

Thoughts?

 

[maverick]

I think it's a cost of $10M today. You are right doesn't hit cash flow until 5 years from now, but the company is worth $10M less because they did this transaction.

Others feel free to correct me.

 

[leftover_salmon]

Originally Posted by maverick

I think it's a cost of $10M today. You are right doesn't hit cash flow until 5 years from now, but the company is worth $10M less because they did this transaction. Others feel free to correct me.
Not to seem like an ass, but I think you'll need to explain more before I buy that. I do understand that there is a definite accounting hit of $10MM, no doubt (or ~7-8MM after-tax), but as for actual PV cost in terms of saving/losing money, I haven't received a straight answer yet (and it's no secret I'd prefer being able to PV it back since that makes a redeem/re-issue scenario look more favourable and thus mean fees).

 

[RedScarf7]

Originally Posted by leftover_salmon

My argument for the latter is that the company only has to actually pay that $10MM down out of pocket in 5 years time - in the meanwhile, they're funding it with a new bond (with the negative effect being an incrementally higher debt load and higher interest payments). From the no-arbitrage perspective, they can effectively sock away $6MM today to repay that $10MM in 5 years' time. Thoughts?
I'm not sure I wholly understand your question, but I am inclined to agree that best treatment would be to defer the cost and record it in the fifth year. Since it has no observable effect on cash flow/ income until the cost is realized in 5 years, the matching principle dictates that the expense is to be realized during the period that it impacts earnings. To me, that would be in 5 years. Again, not 100% sure I'm interpreting everything correctly so take it fwiw. Edit: I agree that the second scenario you outlined seems correct.

 

[scientific]

from an economic perspective its pretty obv just write out the cashflows. from an accting perspective no clue