I like Klarsolo’s example, though there’s one significant difference (in my opinion) between Schweser’s example and Klarsolo’s.
In Klarsolo’s example, X has an absolute advantage in borrowing fixed (6% vs. 7%) and a comparative advantage in borrowing floating (100bp spread vs. -50bp spread). Y has a absolute advantage in borrowing fixed (LIBOR + 100bp vs. LIBOR + 150bp) and a comparative advantage in borrowing floating (50bp spread vs. -100bp spread).
I use a negative spread for analyzing comparative advantage because you have to evaluate the results relative to each other, not just the absolute spread. It’s sort of like an NPV analysis - in order to ascertain which project you would take, you calculate all the NPV’s and take the one with the biggest NPV, subject to your resources available to carry out the project, right? So, if you have projects with NPV’s of +50, +20, +2, -12 and -100, you’d almost certainly try to start with the +50 and work your way down. Same thing with comparative advantage - for two companies with two sources of borrowing, even if one company can borrow fixed and floating cheaper than the other (Schweser’s example), unless the spread is exactly the same, the company will have a comparative advantage in borrowing either fixed or floating, but not both.
So, to put it in context of Schweser’s example, Company Y has an absolute advantage in borrowing both fixed and floating (5% vs. 6.5% and LIBOR + 10bp vs. LIBOR + 100bp), but a comparative advantage in borrowing fixed only (150bp spread vs. 90bp spread). It’s better for Company Y to borrow at fixed only and let Company X borrow floating, even if it’s more expensive for Company X to borrow (assume there’s some kind of limitation on borrowing amounts). Once they trade off, the comparative advantage is realized with that 60bp savings.