AI summary by TwinMind
11.350 Lecture 9: Quantifying Climate Risks’ Impacts on Real Estate Values
MIT OpenCourseWare
1h 26m 46s
Four Quadrant Model & Climate Risk Framework
- Previous week covered four quadrant model for understanding market equilibrium impacts
- Key question: What triggers first response to climate change in real estate?
- Tenants vs investors/owners/financers as initial drivers
- Different responses lead to different policy implications
- Valuation implications vary by trigger source:
- Numerator adjustment: discounted cash flow down due to climate exposure (rents decrease)
- Denominator adjustment: risk/cap rate changes due to financing access issues
Empirical Evidence on Climate Risk Pricing
- Studies show climate risk incorporated via cap rates rather than rental cash flows
- Properties with salient climate risk information show ~3% lower prices
- No change in rents observed, indicating cap rate adjustment mechanism
- Makes conceptual sense: climate risk operates on multi-decade horizon (50-70 years)
- Hurricane Sandy New York study: 9% price reduction for undamaged properties
- Indicates belief updating about future risk rather than actual damage impact
Discount Rates & Investment Decisions
- Climate resiliency investments require upfront costs to avoid future risks
- Even more sensitive to discount rates than energy efficiency investments
- Climate risks operate on 50-100 year timeframes vs 10-20 years for transition risks
- Minor discount rate changes create major valuation differences due to compounding
Boston Seaport Case Study
- Sea level rise predictions: 9" by 2030, 21" by 2050, 40" by 2070s
- Expected damages without intervention: $140M currently to $8B future
- Protection infrastructure cost: $500M-$1B upfront, $7-15M annual maintenance
- NPV analysis shows investment viability depends heavily on discount rate:
- Below 4%: positive NPV
- Above 4%: negative NPV
- Literature suggests 2.6% discount rate for long-term climate investments
Climate Belief Measurement & Geographic Variation
- Yale Climate Opinion Maps survey data across US counties
- National US average: 72% believe global warming happening, 14% don't believe
- Strong geographic correlation with political patterns
- International comparison: Greece 90%, South Korea high, US relatively low
- MIT classroom poll showed much higher belief rate than national average
Climate Belief Dynamics Over Time
- Public priority for climate varies with economic conditions
- During recessions (2009): economic growth/jobs prioritized over climate
- Climate concern increases when economy strong
- Partisan gap has widened significantly over time
- Recent 2022 data shows strengthening economy as top priority
Empirical Research: Florida Flood Zone Study
- Analyzed 1997-2017 housing transactions in Florida flood zones
- Used Yale climate belief map data for community-level belief measurement
- Key finding: 7% larger price decrease in "believer" communities vs "denier" communities
- Same objective flood risk, different pricing based on community beliefs
- Implications for development patterns and social costs
Information Disclosure & Market Response
- Home seller disclosure requirements increasingly include climate risk
- State-by-state implementation correlates with political/belief patterns
- Research shows disclosure leads to population outmigration and housing vacancy
- Information provision allows market-based responses without mandates
Research Projects on Climate Sentiment
- Global Sentiment Project using natural language processing on social media
- Measures emotional toll of extreme temperatures across 100+ countries
- Hurricane information networks study: social learning through Twitter connections
- Shows how climate events spread awareness through social networks beyond affected areas
Methodology: Difference-in-Differences (DiD)
- Addresses "apples to oranges" comparison problem in climate impact studies
- Cannot directly compare flood zone vs non-flood zone properties (different amenities)
- DiD uses before/after treatment with control group
- Key assumption: parallel trends between treatment and control groups absent intervention
- Formula: (Post-treatment - Pre-treatment)treatment - (Post-control - Pre-control)control
Hurricane Sandy NYC Housing Study
- 8% persistent price penalty for flood zone properties post-Sandy
- Used FEMA damage data with multiple damage levels
- Heavy damage properties: 15% discount
- Different from commercial real estate studies showing full recovery after 2 years
- Specification: Log(Price) = α + β₁(Flood Zone) + β₂(Post-Sandy) + β₃(Flood Zone × Post-Sandy)
- β₃ coefficient captures DiD treatment effect
Market Efficiency & Information Processing
- Debate over whether persistent discounts reflect:
- Market learning about previously unknown risks
- Temporary effects from insurance premium increases
- Rational vs irrational buyer behavior regarding low-probability events
- Different studies show varying recovery patterns (temporary vs persistent effects)
- Depends on local factors: insurance availability, repair capacity, government assistance
Cross-Country Institutional Differences
- US vs China: different responsibility structures for climate damage repairs
- Moral hazard issues when government bears repair costs
- Affects private decision-making about climate risk exposure
