Comments on

1
Comments on When Does Idiosyncratic Risk Really
Matter

• Yuanchen Chang
• National Chengchi University

2
Summary of this paper

• Propose a simple noise reduction method that
  includes two different noisy measures (value
  weighted and equal weighted indices) in the same
  predictive regression to reexamine the relation
  between idiosyncratic risk and the expected
  future market return.
• Results show that the noise effect tends to
  cancel out and produce predictive power.

3
Contribution of this paper

• Resolving the empirical debate on the pricing
  role of idiosyncratic risk and unify the
  time-series evidence with the cross-sectional
  evidence.

4
Some observations

• 1. The claim that a stocks proportional investor
  base is equal to its market value weight may not
  be true.
• Control for institutional holdings.

5
Some observations

• 2. Alternative proxies for idiosyncratic risk
• Mutual fund or hedge fund volatility.
• VIX
• 3. Alternative weighting scheme for market
  weight.
• Negatively related to market weight(Arnold and
  Hsu, 2005).

6
Some observations

• 4. How to handle variables with multicollinearity
  concerns?
• Form a principal component by grouping together
  the two variables into a composite index.
• Use a joint F test on the dual predictors to see
  if they are not significantly different from
  zero.

7
Some more observations

• 5. Why stop at dual predictors?
• Trio might be better.

8
Final comment

• I enjoy reading this paper and
• I think it has potentials to make a
  contribution to the current literature.