I often joke that Finance 4335 is not an accounting class and that it is better to be vaguely correct rather than precisely wrong. However, now that you have turned in a couple of problem sets in Finance 4335, I would like to offer some suggestions about rounding.

As a general rule, rounding to the nearest hundredth (i.e., two places to the right of the decimal point) when working on class problems, problem sets, and exams should be sufficient. I am mostly concerned that you learn the intuition and logic behind the models we study. When correctly applied, this level of “precision” will generally produce numeric results that are reasonably close to the solutions which my TA (Chuck Hooker) and I provide (based on Excel spreadsheets, which by default round out to 16 significant digits ). However, on Problem 1A in Problem Set 2, we noticed several students rounding to the nearest whole number. Believe it or not, in finance, a 9% or 5% return is materially (in a real, economic sense) quite a bit (40 basis points) higher than an 8.6% or 4.6% return (8.6% and 4.6% were the answers provided in the solutions for Problem Set 2. Furthermore, many problems in Finance 4335 involve numerical calculations that go through a number of different stages, which implies that a coarse level of rounding (e.g., nearest integer) compounds the difference between your results and ours. This all makes grading much more challenging and potentially error-prone.