I'm a bit perplexed by what Cicero is saying here. Does he mean to compare 12% compound interest with 48% simple, or is he suggesting a straight 48% with 12% simple. The cost of simple or compound interest is directly affected by time, so such a thing is a key component of the calculation.
For instance, let's say someone borrowed 10,000 denarii payable over 10 years.
At 48% flat, the borrower would owe (of have paid) 4,800 interest plus the principal for a total of $14,800.
At 12% simple (INT - Principal x rate x time) (10 month roman years, or 100 months in total, for interest purposes) the borrower would owe 10,000 denarii in interest for a total of 20,000 due to the lender. (This transaction forms the mathematical basis of the Roman borrowing system... interest = principal over a 10 year period).
As such, the transaction is entirely dependent upon time to determine which method of calculation is better for either borrower or lender.
On the same 10 year loan, if we are using simple interest vs. compound interest the cost would be something such as...
At 48% simple (I = PxRxT) (10 month Roman year), the borrower would owe 40,000 in interest for a total due (P & I) of 50,000
At 12% compound [Actual = Principal x (1 + rate)10] the borrower would only owe 21,000 of interest or a total of 31,000. However, I do believe that such compounding had been outlawed in Cicero's time and I believe the first comparative example above is more applicable.
Clearly time is the key component in factoring the most favorable results in any of these transactions. For a rather confusing description of how the Roman calculated interest... Fenus... William Smith Dictionary
(Personally, I am far too reliant on 10-keys and computers for such calculations anymore, but the concept is fairly easy once the brain shakes off the cobwebs. For the ancients, such things were a matter of course for mathematicians.)