Outline of my Presidential Address to AREUEA, New Orleans, Jan 5, 2008

After the luncheon, a number of people asked me for this. It will become a paper soon.

How Well Can We Expect to Understand House Prices? Six Puzzles to Consider
Richard K. Green
AREUEA Luncheon
New Orleans, LA
January 5, 2008

Five Influences Underlying this Talk
• George Akerloff’s AEA Presidential Address
• Jim Shilling’s AREUEA Presidential Address
• Robert Shiller’s recent writing on house price dynamics
• Bill Wheaton’s out-of-sample forecasts
• The current subprime crisis
•
Popular Press

Lots of Papers about House Price Dynamics
• Case-Shiller (1989) The Efficiency of the Market for Single Family Homes
• Meese-Wallace (1994) Should I leave my House in San Francisco?
• Green (2002) Can we explain the Santa Clara Housing Market?
• Common theme—parsimonious models (a la Friedman) fail to unlock completely the
mysteries of the housing market

Why do we do so badly?
• Human beings do not seem to maximize the objective function economists would
write down for housing
• Human beings seem to have mis-calibrated expectations
• There appear to be lots of adverse selection problems
• We still don’t understand supply issues well enough
•
Puzzle One: Tenure Choice and User Cost
• Two anecdotes:
– The Wharton West Story
– The Bangladesh-Barber Story
• More formal evidence
– Meese and Wallace (1994)
– Genesove and Mayer (2001)
– Green (1996)
– Capozza, Green and Hendershott (1996)
–
Puzzle Two: Expectations about House Prices
• Case-Shiller Surveys
• Implied Expectations from Observing Market Conditions
• Misunderstanding of appreciation (Harding, Rosenthal and Sirmans 2007)
• Just the whole idea that “housing is a good investment” (separate from issue that
owner-occupied housing can be utility maximizing). See Goetzmann for a view on
housing as an investment
• Norms?

Puzzle 3: Borrower Behavior
• A remarkable willingness to take on a large obligation that is not well understood
• For awhile, less than ruthless prepayment
• In the past, less than ruthless default. Is this still true?
– Can we draw inferences about prime borrower behavior from subprime borrowers?

More on that large obligation that is not well understood
• Even for the most straightforward mortgage, disclosures are not very helpful
• APR—the “sticker price”—does not convey accurate price information to borrowers
• Fees are not built into APR, and APR is sensitive to expected duration of the
mortgage
• Stanton and Wallace (1998): What’s the point?
• For Adjustable Rate Mortgages, even more confusion

Puzzle 4: Why does this state of affairs continue?
• Why don’t borrowers shop for no-cost loans? Would a default product make sense
(see recent paper by Barr, Mullainathan and Shafir 2007)?
• Alternatively, why not have the Guttentag (2007) proposition: everything needs to
get rolled up into rates and/or points
• Why is there not more demand to buy mortgage counseling?
• Is it possible that there are too many products?

Puzzle 5: Lender-Investor behavior
• Lots of adverse selection and moral hazard
• Main difference between conventional conforming market and private-label market:
the conventional conforming market is far more homogeneous in observed
characteristics—and presumably unobserved characteristics as well
• What does that mean about the sort of people who borrow in the private label—and
especially subprime—market?

Subprime Loans Are Far More Heterogeneous (pretty picture)

Puzzle 6: The compensation chain
• Brokers
– incentive to originate
– Fee driven
• Originators—little capital at risk
• Rating agencies
– Backward looking
– Inconsistent
• Why don’t investors see this? Two common explanations:
– Expectations that house prices would rise forever everywhere
– The world is swimming in cash

Brokers: Controversial Mortgage Terms

Broker channel likely to use instruments that have been the subject of controversy (picture)
– Higher share of loans with prepayment penalty
– Steering more ‘prime’ borrowers with good credit towards subprime and hybrid
loans
–
Regulated Institutions Originate Fewer Loans with Controversial Features (picture)
􀂃 Regulated vs. unregulated originator behavior
􀂃 Low/no doc and 2-year hybrid loans claim a higher share as a percent of total
number of loans securitized by unregulated institutions

Akerloff (2006)
• The five neutrality results:
– Consumption from Wealth instead of income
– Modigliani-Miller
– Natural Rate
– Rational Expectations
– Ricardian Equivalence
• In empirical analysis, they don’t work particularly well
• What do they have in common: expectations.
–
Shilling (2003)
• Compares ex ante expectations and ex post results for commercial property
• Finding: investors are always too optimistic about returns
• Implication of this and other results: we need to do a better job of understanding
behavior.
• Herbert Simon’s point: talking with people is one of the more efficient routes to
fundamentally new ideas.

Calling Daniel Kahneman
• We need to think of how to do experiments on how people make housing decisions
• Three questions I would like to ask everyone who engages in a housing transaction:
– How much do you think house prices will rise in your market?
– Did you make an explicit comparison with the rental market when you bought (and
the converse question for renters)
– What is ownership per se worth to you?