Schweser notes give an intuitive explanation (which is very much in line with the explanation above by cpk123 ):
The formula for the cost of trade credit is very similar to converting HPY to an effective annual return, where the term:
disc.rate/(1-disc.rate)
is the holding period return to the firm of taking advantage of a discount. Say you are supposed to pay $100 and get a 2% discount, then you end up paying $98 and “gain” $2. Then we compute:
0.02/(1-0.02)=2/(100-2)=0.02041
Which is our return on an “investment” of $98 (from the perspective of the buying firm, who has to pay). Of course this only holds, IF we actually use the discount. Obviously for the cost of trade credit, we then want to know what the cost is of not taking the discount, and thus we do the whole annualizing afterwards.