Discussion of: Dynamic Portfolio and Mortgage Choice for Homeowners by Otto van Hemert, Frank de Jon

1
Discussion of Dynamic Portfolio and Mortgage
Choice for Homeownersby Otto van Hemert, Frank
de Jong, Joost Driessen

• Discussant Choong Tze Chua Singapore Management
  University

2
Outline

• Introduction and Summary of paper
• Important Insights of the Paper
• Comments
• Suggestions

3
Introduction and Summary (1)

• Optimal portfolio choice problem How to choose a
  portfolio that maximizes utility
• The literature has expanded from just using (i)
  the mean-variance tangency portfolio to using
  both (i) and (ii) a bond portfolio (Brennan and
  Xia 2002, Campbell and Viceira 2001)

4
Introduction and Summary (2)

• Lately, several papers have started to examine
  the role of housing in an investors portfolio
  (Brueckner 1997, Flavin and Yamashita 2002)
• This paper attempts to develop a more complete
  framework to investigate the effects of housing
  on the financial portfolio i.e. given a house,
  what is the optimal financial portfolio that the
  investor should hold

5
Introduction and Summary (3)

• The paper asserts that the optimal financial
  portfolio (using stocks, 2 bonds, cash) consists
  of
• The nomimal mean-variance tangency portfolio
• A portfolio that resembles an inflation-indexed
  bond
• A portfolio that hedges the risk of the house

6
Important Insights of the Paper (1)

• Analytical solutions for the three components of
  the optimal portfolio when there are no
  short-sale constraints
• Numerical solutions for the three components of
  the optimal portfolio when there are short-sale
  constraints

7
Important Insights of the Paper (2)

• Optimality of fixed-rate mortgages (FRM) vs
  adjustable-rate mortgages (ARM)
• Less risk averse investor would prefer the ARM
• Fairly risk averse investor would prefer the FRM

8
Comments (1)

• Are the choice of bonds (5-year, 20-year)
  significant?
• Could you for instance, replace 5-year bonds with
  1-year bonds and reduce the collinearity by a
  significant amount?

9
Comments (2)

• Are the choice of cities (Atlanta, Boston,
  Chicago, San Francisco) important?
• Could they have substantially different
  characteristics?
• Is the choice of utility function central to all
  the conclusions of this paper?

10
Suggestions (1)

• Provide more description of the data and
  methodology, especially housing price data.
• Is there a larger dataset that could be used
  (since you are constructing an equal-weighted
  nationwide index using just those four cities)
  eg. housing prices in non-urban areas, different
  regions, etc

11
Suggestions (2)

• Provide some more numbers on economic
  significance in addition to utility loss/gains
  (eg for a 1 million portfolio, ignoring housing
  in the investment decision results in an X
  increase in volatility, etc)
• Provide some comment on the possible implications
  (if any) for
• Investors
• Lenders
• Borrowers
• Regulators

12
Suggestions (3)

• Possible future extensions
• Dynamic portfolios (horizons change as time
  passes)
• International portfolio (it is easier/faster to
  transact international financial securities than
  domestic real estate)
• Extend study to different countries