Chapter 3 quiz

Reshaping Statistics

0 of 10 answered

These questions review the examples in the companion slides. Choose an answer, then check the explanation.

Question 1

A later sample contains extreme feature values. Which clipping rule preserves the development boundary?

  • Fit cutoffs on all dates
  • Fit cutoffs on development dates and apply them unchanged
  • Choose cutoffs to maximise held-out IC
  • Delete every held-out extreme

Question 2

How should a bootstrap preserve short runs of dependent monthly observations?

  • Shuffle every stock-month independently
  • Sample contiguous time blocks
  • Replace pending returns with zero
  • Duplicate the best month

Question 3

You revise a model repeatedly after reading its OOS results. How should that period now be described?

  • Untouched confirmation
  • Development evidence
  • A randomised experiment
  • An independent replication

Question 4

What is the sampling unit for the mean of monthly cross-sectional IC?

  • Every trade fill
  • Every cell in the panel
  • One IC per signal date
  • Every stock price since listing

Question 5

The historical monthly ES uses two tail months. What is the appropriate interpretation?

  • A precise future loss limit
  • A sparse-sample tail summary requiring uncertainty and more evidence
  • Proof that worse losses cannot happen
  • A daily risk measure

Question 6

The feature table has 639 stock-month rows. How many independent experiments does that establish?

  • 639
  • 71
  • The row count alone cannot establish independence
  • Nine

Question 7

Must a block-bootstrap interval always be wider than an iid interval?

  • Yes, by definition
  • Yes, for every trading sample
  • No; it depends on the dependence and sample
  • No; intervals never depend on dependence

Question 8

Which tested hypotheses belong in a multiple-testing family?

  • All candidates in the declared search, including discarded attempts
  • Only the final winner
  • Only profitable candidates
  • Only candidates with p below 0.05

Question 9

Spearman IC requires which inputs?

  • Within-date ranks of both the prediction and the target
  • A ranked target and an unranked prediction
  • A ranked prediction and a future feature
  • Only sector labels

Question 10

Which risk statistic changes when return order changes but the set of returns stays fixed?

  • Maximum drawdown
  • The empirical loss quantile
  • The mean loss
  • The unordered tail mean