real estate's big data revolution: the new way to create value
TRANSCRIPT
Real Estate’s Big Data Revolution:
The New Way to Create Value
January 21, 2015
Introductions
2
Partner & Managing Director, The Boston Consulting Group
• Led Homebuilding & Real Estate
• Deep experience advising Fortune 500 &
PE firms
Prior
• Marakon Associates
• Arthur Andersen (CPA)
Education
• MBA, Wharton, University of Penn
• BS, USC– Summa Cum Laude
Jeremy Sicklick CO-FOUNDER & CEO
PhD Candidate, Applied Statistics, University of Texas at San Antonio
• Dissertation Topic – Dynamic Models of
Financial Risk
• Other Research– Bayesian Hidden Markov
Models, Bayesian Decision Theory,
Dynamic Time Series
Prior
• California Economic Forecast
Education
• MA, Economics, UC Santa Barbara
• BA, Economics, UC Santa Barbara
Chris Stroud CO-FOUNDER &
CHIEF OF RESEARCH
Regional Vice President, Strategic Marketing, PulteGroup
• Led strategic planning and acquisition
underwriting
• Emphasis on decisions through data-
driven decision-making
Prior
• Hewlett-Packard
Education
• MMC, UGA – Magna Cum Laude
• BSBA, UTulsa – Cum Laude
JP Ackerman PRESIDENT, STRATEGIC REAL
ESTATE PRODUCTS
Big Data Is Growing Around Us Exponentially
3
Where We Are Going Where We Have Been
“ If you recorded all human
communication from the dawn of time
to 2003, it takes up about five billion
gigabytes (5,000 petabytes) of storage
space. Now we’re creating that much
data every 7 hours”
Every 7 hours
Source: IDC
6.12 10.87 21.61 40.03 2014 2016 2018 2020
7x…every hour
Data in millions of petabytes
‘Big Data’ will revolutionize real estate like it has many other industries
AGENDA
4
About HouseCanary and Big Data
Case Examples Using Big Data
Where We Are Headed
5
We build products that combine
proprietary data and predictive analytics
to help people make better real estate decisions.
5
5
Residential Real Estate Data 1.0 Lacks Sophistication
There is no sophisticated, comprehensive, localized information source for professionals to use to make investment decisions about residential real estate.
6
Problem
Real estate is a volatile industry with huge investments at stake. Consistently winning requires sophisticated information and insights.
• Data aggregators charge exorbitant rates for raw data, that lacks intelligence
• Rear-view looking data is used, that lacks forward-looking market realities
• Most information is too high-level (county / MSA level) and lacks the local relevance to decisions at hand (zip code, block, property level)
Real Estate Market Data 2.0 is Analytics & Applications
Real Estate Market Data 1.0
7
Real Estate Smart Data 2.0
= Data Aggregation
= Data Aggregation
+ Predictive Analytics
+ Applications to Business Decisions
+ Data Feedback Loops
Our Products
8
NATIONAL HOME CONSTRUCTION DATABASE
Powered by
The most objective, consistent, and efficient way to appraise homes
HouseCanary will create the “New Home MLS”
Launching Q4 2015
The most advanced tools and advisory for real estate professionals
What We Do
9
Accurate big data analysis
Create an information advantage
Predict future monthly prices… All 381 US MSA’s 20,000 Major Zip Codes Using big data…. 1000’s of variables analyzed 40+ years of monthly data …and state of the art analytics Leading finance methods + Leading statistical methods
Reports & infographics that explain how price will move & why Communicate risk / opportunity clearly and simply Translate forecasts into impact Enable a culture of disciplined decision-making
Dispersion of returns is increasing
at a local level
Volatility is increasing and will
continue with rising rates
Demographic shifts creating
profound change in supply /
demand
The Need for Good Data & Insight Is Increasing
10 -‐8% -‐6% -‐4% -‐2% 0% 2% 4% 6% 8%
Raleigh, NC Houston, TX
Austin, TX Dallas-Fort Worth, TX
Seattle, WA Baltimore, MD Charlotte, NC
San Francisco Bay Area, CA Phoenix, AZ
Tampa, FL Las Vegas, NV
Riverside-San Bernardino, CA Orlando, FL Atlanta, GA
Los Angeles-Orange Cty CA Chicago, IL
Washington DC Metro Miami, FL Detroit, MI
Sacramento, CA 7.0% 6.8% 5.7% 5.6% 5.6% 5.5% 5.4% 5.4% 5.3% 5.2%
5.0% 4.9% 4.9% 4.1% 3.7% 3.5% 3.5% 3.1% 2.7% 2.7%
Difference (Max – Min)
Dispersion of Returns (10 year CAGR – unlevered)
Top quartile return for sub-markets Bottom quartile return for sub-markets
11
Local and Macro Fundamentals
Housing Data
Capital & Credit Markets
Consumer Data
Household-Level
Local Market Data
Mortgage volume & mix
Mortgage health & delinquency
Homebuilder capital growth
Residential & Mortgage REIT indices
Construction materials futures
Mortgage yields & spreads
Mortgage Debt, ARM, RMBS growth
120M parcel level details geo-coded
Property details and valuation
• 900 MLS
• 3,000 County Assessors
Land supply available
Permits
Local jobs / employment
Construction jobs
Consumer flows in-out areas
Consumer equity vs. debt
Affordability components
Income driven
20,000 home price indices
Sales volume
Foreclosures
Starts / permits
Months supply
REO
Single vs. Multi-family
Rent versus own economics
Career & Income
Commute times
Migration patterns
Potential demand
Family makeup
Education level
Ethnography
Rental vs owner makeup
Predictive Analytics: Integrating Disparate Data into Insight Thousands of data elements tracked monthly that align with the following areas
Household balance sheet
Sub-market Ratings
Market risk scores
Local economy (GMP)
Recession probability forecasts
Inflation measures
Schools
Crime
Housing Makeup
Livability / amenities
Our Approach to Forecasting
12
How we achieve strong results
• Use Bayesian model averaging to scan the space of 4 Billion potential models for each area, each month
• Average across the best ~1,000 models
• Requires mass computing power
Percent Change in Home Prices - Los Angeles
Time
perce
nt ch
ange
2004 2005 2006 2007 2008 2009
-0.03
-0.02
-0.01
0.00
0.01
0.02
0.03
n Potential Models
Forecast Average
Actual
How We Forecast & Backtest
How We Apply HouseCanary to Builders/Developers
13
Market location: Where to invest
Market timing: When to invest
Consumer focus: Who to focus on
Product design: What to build
Pricing: How to price
How should we value new communities? How should we change price monthly to get margin & absorptions?
Which sub-markets should we focus our projects and land buying? How should we deploy capital across our portfolio?
Can we call the top of the market? Can we define the recovery? When should we deploy / pull-back capital?
Supply: Where is the land
Where is there land available? How much supply is coming on line? How are new homes absorbing locally?
Which consumer segments live there / are moving there? Who should we focus on to maximize growth? What supply of product exists in the market today? What lot size and product design is the highest and best use?
Where to Invest Case Study 1: Market Location
14
• Company often outbid by the herd of competitors in A submarkets.
• Focus was on defining a strategy to maximize returns by using a forward-looking perspective.
Forecast Models Used to Define Potential Growth
15
Focus is then to de-average markets down to the zip code level
Phoenix−Mesa−Scottsdale, AZ
Time
Hom
e Pr
ice
Inde
x
2000 2005 2010 2015
100
120
140
160
180
200
r2 of average model = 0.95
Begin by Derive Sub-market Clusters and Inherent Risk
16
A–F Submarket Clustering for Riverside Area
Across 20,000 zip codes, we clustered zip codes Several factors together help us cluster zip codes
• Median Home price • Income & wealth • Education level • School scores • Crime level • Demographic / International mix • Owner / rental mix • Commutability
These zip code clusters have very different behavior and price volatility within a market
F D C B A
0
50
100
150
200
250
300
350
400
Submarkets Within an MSA Behave Very Differently
17
A
B
C
There is value from differentially investing in A, B, C… markets based on where we are in the cycle We can derive A, B, C…. markets and forecast them accurately in all MSA areas
Home Price Index
D
F
Market Examples Beta
A 92210 – Indian Wells 1.3
B 92590 - Temecula 1.6
C 92882 - Corona 2.0
D 92503 - Riverside 2.1
F 92543 - Hemet 2.7
Where to Invest? Leverage Zip Forecasts to Counter Market Norms: 2013-15
18
Strong Sales Volume
• Established, extreme competitive
set
• Easy commute
Rapid Appreciation
• Prices accelerating at unsustainable
levels
Extreme Competition
• Finished lot values nearly doubled
Conventional Wisdom Corona-Eastvale: Zip 92880
Where to Invest? Leverage Zip Forecasts to Counter Market Norms: 2013-15
19
Limited Volume
• Virtually no competition
• Available lot supply
• Development required
Rapid Appreciation Ahead
• Record low affordability
• Similar price increases ahead
Early Cycle Opportunity
• 2011-13 appreciation < Corona
• 3-Year forecast outpacing Corona
HouseCanary Riverside: Zip 92505
Where to Invest? Leverage Zip Forecasts to Counter Market Norms: 2013-15
20
Limited Volume
• Virtually no competition
• Available lot supply
• Development required
Rapid Appreciation Ahead
• Record low affordability
• Similar price increases ahead
Early Cycle Opportunity
• 2011-13 appreciation < Corona
• 3-Year forecast outpacing Corona
HouseCanary Riverside: Zip 92505
21
Strong Sales Volume
• Wide availability of land
• Strong, established competition
• Draw from master plans
Rapid Appreciation
• 15%+ in recent years
• Outpacing growth in nearer-in
markets
• “best of what’s left” perception
Perceived Affordability
• Lower price points than more job
proximate locations
Conventional Wisdom Menifee: Zip 92584
Where to Invest? Leverage Zip Forecasts to Counter Market Norms: 2015-17
Where to Invest? Leverage Zip Forecasts to Counter Market Norms: 2015-17
22
Mid-to-Late Cycle Play
• Stable, well-rated market
• Most competition going deeper to
Inland Empire for perceived
affordability
Stable Future Pricing Growth
• Local affordability well below
historical peaks
• Job proximate
Controlled Land Supply
• Available via master plans
• Infill opportunities
HouseCanary Ontario: Zip 91764
Where to Invest? Leverage Zip Forecasts to Counter Market Norms: 2015-17
23
Mid-to-Late Cycle Play
• Stable, well-rated market
• Most competition going deeper to
Inland Empire for perceived
affordability
Stable Future Pricing Growth
• Local affordability well below
historical peaks
• Job proximate
Controlled Land Supply
• Available via master plans
• Infill opportunities
HouseCanary Ontario: Zip 91764
24
Insight
Created a historical & projected perspective on the flow
of appreciation through the markets in the Inland Empire
Client was able to develop a differentiated geographic
strategy for near- and long-term deals
Identified the vast disparity in returns
31% returns ‘14 – ‘17 from contrarian investments
• Fontana Montclair, Riverside, Hemet, Perris
versus
22% returns ‘14 – ’17 from traditional A & B submarkets
• Rancho Cucamonga, Corona, Eastvale, Temecula
Impact
Able to underwrite acquisitions at higher returns and
purchase land at discounted rates
+2 years In Additional Lots with No Additional Capital
+50% Increase in Exposure to Market Appreciation
Who to Serve Case Study 2: Consumer Focus
25
• Company evaluated opportunity to increase focus on 55+ segment.
• Once 55+ strategy was defined, needed a way to operationalize 55+ at a local level.
Longer-term Household Growth Driven by the Boomers
Growth in Household
Requirements (by segment)
(5,000,000)
-‐
5,000,000
10,000,000
15,000,000
20,000,000
1970 1980 1990 2000 2010 2020 2030 2040 2050
65+
45-‐64
20-‐44
Mix of Growth
in Household
Requirements
(by segment)
Entry level Move-up Senior Age segment
Ages 1960-1970 1970-1980 1980-1990 1990-2000 2000-2010 2010-2020 2020-2030 2030-2040 2040-2050
20 - 44 37% 64% 56% 20% -11% 32% 25% 28% 40%
45 - 64 34% 12% 15% 66% 84% 10% -6% 28% 23%
65+ 29% 23% 29% 14% 27% 58% 81% 43% 38%
Total 100% 100% 100% 100% 100% 100% 100% 100% 100%
Source: HouseCanary, Census
Across time, following the Boomers has been a successful strategy
$65k
Households
Current Households (Million)
Future 2030 Households (Million)
Avg. Wealth of segments ($k)
Homeownership rate of segments
22
Boomers: Growing, Wealthier, and Own Homes
$802k
42%
62%
77%
20
21
38
$573k 24
25 - 34
35 - 44
45 - 54
55 - 75
$217k
Source: Census Bureau; HouseCanary. Note: Bubble Size is shown to scale
12 $678k n/a 75+
72%
23
26
61
10
Decline in Homeownership Except Among Oldest Cohorts
0%
10%
20%
30%
40%
50%
60%
70%
80%
90%
1980 1990 2000 2010
65+ years
15 – 24 years
25 – 34 years
35 – 44 years
45 – 54 years
55 – 64 years
Home ownership rate by age of householder
Source: HouseCanary, Federal Reserve Bank of New York, Equifax, Census
The Wealth Factor Separates the Age Cohorts
High net worth of older cohorts offsets the impact of rate hikes by enabling
larger down payments
Younger cohorts already hitting limits
of DTI given lower wages and increasing
debt load
Younger cohorts will be further
devalued in an increasing interest rate
environment
3.5%
10%
20%
35%
50% 50%
0%
10%
20%
30%
40%
50%
60%
0
50
100
150
200
250
Ages <35 Ages 35-44 Ages 45-54 Ages 55-64 Ages 65-74 Ages 75+
Sources: HouseCanary, Federal Reserve, Census Bureau 2013
Typical Down Payment % Median Annual HH Income & Net Worth (K)
X%
Net worth (K)
Annual HH Income (K)
Avg down payment %
Major Risk: Forecasted Interest Rate Increase - 4.3% to 6% Other Risks
Age Group Personal Balance 4.3% Rate 6% Rate Impact to Demand Rationale
Millennials (15-34)
Income $59k Worth $10k Debt Equity
30% 35% -36% • No/low savings • Anemic income/career growth • No / low gain from stock & housing
rebound • Heavy debt load (school, personal) • Hard to underwrite given QM
Young Gen X (35-44)
Income $83k Worth $47k Debt Equity
21% 25% -24% • Rocked by housing downturn • Limited savings • Heavy debt
Old Gen X (45-54)
Income $88k Worth $105k Debt Equity
18% 21% --% • Home equity • Stock market rebound • High income/career trajectory
Boomers (55+)
Income $67k Worth $180k Debt Equity
17% 19% --% • Large savings & net worth • Limited housing / personal debt • Large down payments • Large equity investments
% Income Used for Housing
Source: HouseCanary, BLS, Census Bureau, Nov 2014
Young Households Greatly Impacted by Future Rate Hikes
Pinpointing Migration of 55+ is Critical for Investment
Understanding Who & How Much Consumers Can Afford
0"
2,000"
4,000"
6,000"
8,000"
10,000"
12,000"
14,000"
16,000"
18,000"
0)49k" 50)99k" 100)149k" 150)199k" 200)249k" 250)299k" 300)399k" 400)499k" 500)749k" 750)999k" >1mil"
HH"Age:""65+" HH"Age:"45)64"
HH"Age:"15")"44" Renters"
Supply"
0"
2,000"
4,000"
6,000"
8,000"
10,000"
12,000"
14,000"
16,000"
18,000"
0)49k" 50)99k" 100)149k" 150)199k" 200)249k" 250)299k" 300)399k" 400)499k" 500)749k" 750)999k" >1mil"
HH"Age:""65+" HH"Age:"45)64"
HH"Age:"15")"44" Renters"
Supply"
Demand 4.3% Interest Rate
Demand 6.0% Interest Rate
# of Households
House Price
# of Households
House Price
33
Insight
Huge opportunity with 55+ over next several decades
55+ opportunity remains robust as interest rates increase
Operationalizing the opportunity requires deep insight
• Where target consumers live
• Where target consumers move within area
• Where target consumers move from outside area
• What target consumers can afford
Impact
• Leveraged data to identify priority submarkets for
overall growth strategy
• Capitalized on underserved consumer segment
• Drove underwriting with targeted age segment needs
$500M Identified Growth Opportunity at a Local Level
What to Build Case Study 3: Product Optimization
34
• Company trying to establish competitive differentiation within master plan segmentation.
• Focus was on maximizing pricing power to serve established families among heavy competition.
Product Optimization Target Product Sizes with Imbalanced Supply & Attribute Premiums
-20% -15% -10% -5% 0% 5% 10% 15% 20% 25% 30%
1500-1999sf
2000-2499sf
2500-2999sf
3000-3499sf
3500-3999sf
4000-4499sf
4500-4999sf
5000sf+
Submarket
1-‐Story Prem
3-‐Car Prem
0.0 5.0 10.0 15.0 20.0
1500-1999sf
2000-2499sf
2500-2999sf
3000-3499sf
3500-3999sf
4000-4499sf
4500-4999sf
5000sf+
Zip Code MOS: 6.2
MSA MOS: 7.2 % Price Premium Paid for 1-story homes & 3-car garages by Square Footage Range
35
Healthy Months of Supply in Targeted Size Ranges 3-Car Garages Matter; Single Stories Don’t
Product Optimization Maximum Functionality Drives Price Based on Consumer Profile
$97
$100
$112
$98 $97
$100
$105
$129
$80
$90
$100
$110
$120
$130
3 4 5 6 2 3 4 5
Price per Foot
Bedroom Count Bathroom Count
+4%
+11% -12%
+5%
+23%
+4%
+29%
Source: MLS, December 2014, SFD only, 2000-4000sf 36
Product Components Premiums Leverage Location, Product Footprint and Configuration to Maximize Revenue
Comparable Price Subject Property Price Prem / (Dscnt) %
Bath Count 4 v.5 Bathrooms
Bed Count 4 v. 5 Bedrooms
New v. Resale Built Before 2010 v. Built Since
Optimized Product 2-car, 4 bed, 4-bath v. 3-car, 5 bed, 5 bath new
Location MSA v. Zip
$102/ft $100/ft (2%) $96/ft $125/ft 30% $96/ft $107/ft 11% $99/ft $119/ft 21% $99/ft $107/ft 8% $86/ft $169/ft 96%
Source: MLS, December 2014, SFD only, 2000-4000sf, 5,000-14,999sf lots only
Garage Count 2 v. 3-Car Garage
37
Drive Pricing Strategy through Product Attributes
38
Clo
sin
g P
rice
(s
quar
e fe
et)
House size (square feet)
70’ Lots Positioning Strike Zone
60’ Lots
$400,000%
$450,000%
$500,000%
$550,000%
$600,000%
$650,000%
$700,000%
$750,000%
$800,000%
$850,000%
2000% 2500% 3000% 3500% 4000% 4500%
Single%Stories% 36Car%Garages% 26Car%Garages%
Submarket%(ALL)% Built%Since%2010% Shea%BackCountry%6%Shadow%Walk%(40'%Wide%Product)%
Shea%BackCountry%6Water%Dance%(45'%Wide%Product)% Log.%(Single%Stories)% Log.%(36Car%Garages)%
Log.%(26Car%Garages)% Log.%(Submarket%(ALL))% Log.%(Built%Since%2010)%
Competitor 1
Competitor 2
39
Insight
Target consumers placed less value on bedroom counts
and higher value on additional garage space.
Combination of the two yielded increased value, allowing
the client to set themselves apart from the competition.
Client created a unique vantage point allowing them to:
• better underwrite the land for its optimal value, and
• design more consumer-refined product for the market
Impact
• Maximize pricing power
• Increase revenue disproportionately v. cost
$50K Per Home in Additional Price
$8M+ Total Impact
Where We Are Going With Big Data
About HouseCanary
40
41
Overview
The most advanced tools for real estate professionals
Answers five key questions for builders, developers, banks, and funds:
Market timing: When to invest?
• Analyze the housing cycle
• Identify downturn risk
• Pinpoint market changes
Location: Where to invest? • Target sub-markets
• Understand local market changes
• Define where and why the market is changing
Product design: What to build?
Pricing: What to charge?
Marketing: Who to sell to?
42
All in one from an iPad • Communication • Data integration: Market data, property data, maps • Photos, Property measurement • Comp analysis, adjustments, valuation
Flips the focus onto value • From 90% time on form filling to 90% focusing on appraising
value & risk
Scientific analytics at your fingertips…no PhD degrees needed • Valuation and quantified comps analysis • Analytically derived value adjustments • Value drivers and hierarchy
Leave your work at the job site • Enter the site, do what you need to, submit in just 30 minutes
Seamless, paperless integration • Links to UADP forms in MISMO format • All mobile, all paperless, all stored in the cloud • Platform is 'form' agnostic, streamlining the production of
alternative valuation products
Overview
43
NATIONAL HOME CONSTRUCTION DATABASE
Powered by
What did HouseCanary win?
• Exclusive arrangement to build the National Home Construction
Database (NHCD) awarded 11/2014
• We will build a service where builders can upload new home
sales data, in return receiving aggregated / anonymized data
• Competitive bid against top data companies
Why is this valuable to the new home market?
• First look at proprietary new home sales data
• First residential real estate dataset compiled nationally through
single entity
• Opens many other business opportunities over time
Overview
Creating and owning the “New Home MLS”
Jeremy Sicklick CO-FOUNDER & CEO
office: +1-866-729-7770
mobile: +1-213-422-2577
email: [email protected]
For Questions
300 Brannan St. #501
San Francisco, CA 94107
1617 Pacific Coast Highway, #H
Redondo Beach, CA 90277
JP Ackerman PRESIDENT, STRATEGIC REAL ESTATE PRODUCTS
office: +1-866-729-7770
mobile: +1949-444-3600
email: [email protected]