AdvancedMultiple choice
Scoring a forecast out of sample · Part 3 of 3
You forecast next-day index returns. Over 500 out-of-sample days, the forecast’s mean squared error is 3.96 (in per cent squared); a forecast equal to the historical mean has mean squared error 4.00.
A rolling re-estimation of the same model gives an out-of-sample R² of −0.5%. What does the negative sign mean?
- AThe forecasts get the direction wrong more than half the time
- BThere must be a computing error, because R² cannot be negative
- CIt has a larger squared error than the historical mean
- DBetting against the model would earn an R² of +0.5%
The worked solution is in Premium
The answer, the full working and the one idea to take away – for this and all 1,322 questions in the bank. Answer it in practice and your working is marked, with a known mistake named when you make one.
More time series questions
- Your labels are each asset’s return over the next 10 trading days.Foundation
- You have ten years of daily data, start with a two-year training window, expand…Applied
- Out of sample, your return forecast has mean squared error 0.98 while the…Applied
- You compute twelve-month returns every month over ten years and test whether…Advanced
- Which of these is closest to being a stationary series?Foundation
- For white noise, sample autocorrelations have standard error about 1/√T.Foundation