Uh??

I didn't understand one thing so I sent an email to the professor... and this is his reply.

Good observation. The choice between one and the other is not always definite. In most high leverage cases the cost of debt is very close to the unlevered cost of capital and the difference is small. When the changing debt schedule is predetermined the natural choice is the cost of debt. I computing wacc every year one is forced to assume that the debt ratio stays constant during that year, which means that debt moves up and down, however little, during the year with the value of the firm in order to keep the debt ratio constant as assumed for the year. That imparts additional risk to the tax shield: the risk of the free cash flow and makes it appropriate to discount the tax shield at the cost of the unlevered firm. The choice in practice varies. Whenever it is felt that the amount of debt is highly dependent on the cash flow, the cost of unlevered firm is used. Otherwise the cost of debt is used. The example of p. 93 is continued in the following pages and the 2nd para of p. 99 makes reference to the results attained with one or the other assumption. As you can see the difference is relatively small. Finally, one important point to consider is that by computing wacc every year one is assuming that the tax shield is discounted at the unlevered cost of capital. Hope this helps.

Now I don't undestand naffing! (nothing)