Who Buys Boston? Investors, Flippers, and the Cash Economy in Urban Residential Real Estate
Dataset: EDA Residential (Boston)49,301 transactions2000 – 2022Analysis: February 2026
This report presents an exploratory analysis of residential real estate transactions in Boston, Massachusetts from 2000 to 2022. Drawing on nearly 50,000 sales records, we investigate three interrelated questions: how investor participation has evolved over two decades of price appreciation; whether property flipping generates reliable returns and concentrates in specific neighborhoods; and what the dramatic rise in all-cash purchases reveals about who has access to Boston's housing market. Our analysis reveals a housing market increasingly shaped by capital-backed buyers operating outside traditional mortgage financing.
Research Questions
Prior to conducting visual analysis, I identified three questions motivated by current housing policy debates, academic literature on urban financialization, and personal observation of Boston's rapidly changing neighborhoods.
How has investor participation in Boston's residential market changed over time, and do investors pay premiums that push prices higher for everyone?
Motivation: A growing body of research documents the “financialization” of housing — the transformation of homes from use-value to investment assets (Aalbers, 2016). National reporting and academic work describe institutional investors systematically purchasing single-family and condo units in major cities. If investors crowd into Boston's already-tight market, their higher willingness-to-pay could accelerate price appreciation, making homeownership less accessible for owner-occupants. I wanted to see whether this pattern shows up in the data, particularly post-2008 when foreclosures and low interest rates created opportunities for bulk acquisition.
Do Boston property flips generate consistent profits, and has flipping activity concentrated in particular neighborhoods, potentially displacing existing residents?
Motivation: The popular narrative of “house flipping” — buying distressed properties, renovating, and reselling quickly — appears frequently in both popular media and policy discussions about gentrification. I was curious whether flipping in Boston is actually profitable (or whether the TV shows overstate returns), and whether flip activity is spatially concentrated in neighborhoods undergoing rapid change.
Has the prevalence of all-cash purchases increased over time, and is the cash premium concentrated among investor buyers, suggesting a structural disadvantage for mortgage-dependent buyers?
Motivation: Research by economists at Redfin (2022) and reporting in the Boston Globe (2023) highlight that all-cash offers are 2–4× more likely to succeed in competitive bidding situations, creating a structural advantage for wealthy individuals and institutions that can bypass mortgage contingencies. If cash purchases have risen significantly (especially post-2008 when credit tightened), this would then suggest that the composition of Boston's buyer pool has shifted in ways that disadvantage first-time and income-constrained buyers who rely on financing.
Phase One: Dataset Overview
Before investigating specific questions, I examined the overall shape and structure of the dataset to understand its contents, assess data quality, and identify any surprising patterns that might affect interpretation.
The dataset contains 49,301 residential transactions recorded by the Boston Assessing Department, spanning 2000–2022. Each record includes sale price, property characteristics (style, year built, square footage, bedrooms/bathrooms), location (latitude, longitude, ZIP code), transaction metadata (cash vs. financed, buyer/seller entity types), and derived variables indicating flip status and investor classification.
Data Quality Assessment
Field
Missing
Zero/Blank
Notes
price
0
0
Complete; 1 record = $999,999,999 (likely data error — excluded from price analyses)
intersf (interior sq ft)
0
431 zeros
Zero-sqft records excluded from $/sqft calculations
bedrooms
0
6,151 zeros (12.5%)
Many condos coded as 0-bedroom studios; not necessarily missing
bathrooms
0
3,445 zeros (7.0%)
Similar pattern to bedrooms
yearbuilt
66 missing
1,369 zeros
~3% of records lack construction year; excluded from era analysis
lat / lon
0
456 zeros
Un-geocoded records; excluded from any geographic analyses
zip
0
0
Complete; 38 unique ZIP codes present
investor_type_purchase
0
—
“Non-investor” is the default category, not missing
Key Finding: The overall data quality is high. The most substantive issue is the 12.5% of records with zero bedrooms, which reflects studio condominiums rather than missing data. Price records are complete. For analyses requiring interior square footage or geographic coordinates, I exclude the ~1% of affected records.
Figure 1
Distribution of Residential Sale Prices (Under $10M)
The price distribution is strongly right-skewed, with the majority of transactions clustered below $1M and a long tail extending into multi-million dollar territory. The modal price bucket ($0–$500K) contains 18,224 transactions (37%), while the next bucket ($500K–$1M) holds 16,650 (34%). Fewer than 5% of sales exceed $3M. This skew necessitates careful use of median rather than mean for price comparisons throughout this analysis.
Figure 2
Annual Transaction Volume & Median Sale Price, 2000–2022
Two clear disruptions punctuate an otherwise upward trend. Transaction volume peaked in 2004 (3,205 sales) during the pre-crisis boom, then fell sharply through 2009–2010. Median prices, however, declined only modestly during the financial crisis — from $550K in 2008 to $530K in 2009 — suggesting that Boston's urban condo market was relatively insulated from the nationwide crash. After 2012, prices entered a sustained appreciation cycle, doubling from $621K to a peak of $1.1M in 2019. The COVID dip in volume (2020: 1,516 sales, the lowest in the dataset) was followed by a price correction in 2021, likely reflecting the shift to remote work reducing demand for expensive urban condos.
Figure 3
Composition by Building Style & Median Price per Square Foot
Boston's residential transaction mix is dominated by row houses and mid-rise condominiums, reflecting the city's dense urban fabric. “Row Middle” units (townhouse interiors) account for 36% of all sales, while mid-rise and high-rise condos together represent another 40%. Notably, high-rise units command the highest median price per square foot (~$809/sqft), a premium of 37% over low-rise buildings ($586/sqft), suggesting that location and amenity effects outweigh the size advantage of lower-density buildings.
Figure 4
Median Sale Price by Building Construction Era
The relationship between building age and price is distinctly U-shaped — the oldest buildings (pre-1840) and the newest (post-1980) both command significant premiums. Pre-1840 structures fetch median prices above $1.5M, reflecting their rarity and historic Beacon Hill/Back Bay addresses. Mid-century buildings (1940–1959) are the most affordable (median $463K). Buildings constructed after 2000 command a $990K median, driven by new luxury condo development. This pattern complicates simple “newer is better” assumptions and highlights the premium placed on Boston's historic architectural stock.
Phase Two: Question Investigation
Q1 · Investor Participation Over Time
The dataset classifies each buyer as Non-investor, Small, Medium, Large, or Institutional based on portfolio size and organizational form. I examined how this composition has evolved over 23 years and whether investor buyers pay systematically different prices than owner-occupants.
Figure 5
Buyer Composition by Investor Type, 2000–2022
Investor participation rose substantially between 2011 and 2019, then partially retreated. In 2000, roughly 20% of purchases were by investors of any type. By 2013–2019, the investor share climbed to 34–37%, with small investors (1–9 units) driving most of the growth. Institutional buyers (large portfolio operators) remained a small fraction — typically 1–2% — but their presence grew meaningfully in the 2016–2019 period. Post-COVID (2021–2022), the investor share fell back to ~30%, possibly reflecting rising interest rates reducing the yield advantage of leveraged investment properties.
Figure 6
Median Sale Price by Buyer Investor Classification
Investor buyers pay substantially more than non-investors, with the gap widening by investor scale. Non-investor buyers — primarily owner-occupants — have a median purchase price of $582,500. Small investors pay $875K (50% more), medium investors $1.09M (87% more), and large investors $1.63M (180% more). This pattern likely reflects sorting: larger investors target premium properties expected to generate higher rents or appreciation. However, this does not necessarily mean investors are outbidding owner-occupants for the same units — they may be purchasing fundamentally different (larger, higher-floor) properties within the same buildings.
Unexpected Finding: Institutional buyers (54.8% cash rate) pay a median of $955K — slightly below large investors ($1.63M) — suggesting they may be purchasing in bulk at negotiated prices rather than competing at open-market auctions. This warrants further investigation in a follow-up analysis using repeat-sales matching.
Figure 7
Investor Market Share vs. Annual Price Appreciation, 2003–2022
Years with higher investor participation tend to coincide with stronger price appreciation, though causality is impossible to establish from this analysis alone. The scatter shows a positive association between the non-investor share and price level: years when investor share was highest (2013–2019) were also years of the steepest price growth. However, this correlation may reflect common underlying factors — a strong economy attracting both buyers and investors — rather than investors directly driving prices. The 2021 outlier (high prices, lower investor share) coincides with pandemic-driven demand from owner-occupants relocating within the city.
Q2 · Property Flipping: Returns and Geography
The dataset flags properties as “flipped” when the same address appears as both a buy-side and sell-side transaction within a defined horizon. I examined the profitability distribution and the geographic concentration of flip activity.
Figure 8
Property Flip Rate by Year, 2000–2022
Flipping was virtually nonexistent before 2011, then surged dramatically. Between 2000–2010, the flip rate never exceeded 0.6% of annual transactions. Beginning in 2011 — as Boston's prices began their sustained recovery — flipping activity jumped to 4.8%, and peaked at 9.1% in 2016. This timing is consistent with the post-crisis “buy low, renovate, sell high” opportunity created by a combination of discounted acquisition prices, cheap renovation financing, and rapidly appreciating end-markets. The rate stabilized at 5–7% through 2022, suggesting flipping has become a persistent structural feature rather than a one-time opportunistic response.
Figure 9
Distribution of Flip Profit Percentages (Among 1,718 Flagged Flips)
The median flip earns a 10.9% return, but the distribution is surprisingly wide — including many money-losing transactions. Roughly 15% of flips in the dataset resulted in a nominal loss (negative price change). The distribution peaks in the 0–20% range, with a long right tail extending to gains over 100%. These loss-making flips may represent distressed resales, estate liquidations, or market timing failures. The prevalence of negative-return flips suggests that house flipping in Boston is considerably less reliable than popular media portrayals suggest.
Figure 10
Flip Rate and Median Flip Profit by ZIP Code
Flipping is more geographically dispersed than expected, with the highest flip rates in expensive ZIP codes rather than lower-income ones. ZIP 02199 (Back Bay luxury) has the highest flip rate (5.6%) but negative median profits (-0.9%), suggesting these are often failed speculative purchases. ZIP 02113 (North End) shows the highest median flip profit (14.9%) at a moderate flip rate (3.5%). Notably, ZIP 02215 (Brighton/Allston) has the lowest median home price but strong flip profits (14.3%), potentially indicating value-add renovations in a transitional neighborhood. The absence of pronounced flip concentration in lower-income areas challenges the popular narrative of flippers targeting vulnerable communities — at least within this dataset's definition of “flip.”
Q3 · The Cash Economy: Who Pays Without a Mortgage?
Cash transactions — where no mortgage is recorded — represent a structural advantage in competitive markets. I examined how the cash sale rate has evolved and which buyer segments drive it.
Figure 11
All-Cash Purchase Rate by Year, 2000–2022
The financial crisis of 2008 permanently shifted Boston's cash-sale baseline upward. Before 2008, fewer than 1 in 4 transactions were cash (18–24%). During 2011–2013, the rate spiked to over 43% — likely as traditional mortgage lenders tightened credit, while investors with liquid capital seized distressed-market opportunities. Crucially, the rate never returned to pre-crisis levels: the 2015–2022 period settled at a structurally elevated 38–40%. This “sticky” shift suggests a lasting change in buyer composition, not merely a temporary adjustment, with lasting implications for mortgage-dependent buyers competing in the same market.
Figure 12
All-Cash Purchase Rate by Buyer Type
Cash purchase rates rise sharply with investor scale, revealing a stark two-tier structure in Boston's housing market. Non-investor (owner-occupant) buyers use cash in only 27.5% of transactions — likely reflecting high-net-worth individuals rather than typical homebuyers. In contrast, institutional investors pay cash in 54.8% of transactions, and large investors in 47.1%. This cash advantage is not just financial — in Boston's competitive bidding environment, cash offers can close in days rather than weeks, systematically favoring investors in multi-offer situations regardless of offered price.
Figure 13
All-Cash Purchase Rate by ZIP Code
Cash concentration is highest in the most expensive ZIP codes, with ZIP 02199 (Back Bay luxury) leading at 67.6% cash — nearly 7 in 10 transactions require no financing. Beacon Hill (02108, 45.2%) and the Financial District (02110, 41.9%) follow closely. The more affordable South End / Roxbury (02118) has the lowest cash rate at 19.8%, suggesting that traditional mortgage-financed buyers still dominate lower-price-tier neighborhoods. This geographic stratification means that competition from cash buyers is most acute precisely in the neighborhoods where prices are already highest.
Summary & Lessons Learned
This exploratory analysis reveals a Boston residential market that has undergone a structural transformation over 23 years — not just price appreciation, but a fundamental shift in who is buying and how.
Lesson 1 · Structural Shift
The 2008 financial crisis was an inflection point not just for prices but for buyer composition. Cash purchase rates nearly doubled and have never returned to pre-crisis levels, suggesting a lasting reorganization of the buyer pool toward capital-rich actors.
Lesson 2 · Investor Scale
Investor participation peaked at 37% in 2019, up from 20% in 2000. Larger investors pay more, use cash more, and may occupy a fundamentally different market segment than owner-occupants — making simple “investors vs. buyers” framing potentially misleading.
Lesson 3 · Flipping Reality
Property flipping surged post-2011 but is neither as reliable nor as neighborhood-specific as popular narratives suggest. ~15% of flips lose money, and flip activity is actually highest in already-expensive ZIP codes, not lower-income transitional areas.
Lesson 4 · Data Limitations
The dataset captures transaction prices but not rent levels, household income, or displacement rates — the variables most relevant to understanding housing affordability. Future analysis should link these records to ACS demographic data and rental listings.
Key open questions for future investigation include: Do investor purchases in a given ZIP code predict above-average price appreciation in subsequent years?Are short-term flips more profitable than long-term holds in Boston's market? And critically: How does investor activity vary within ZIP codes, and does it cluster around specific streets, transit stops, or school district boundaries? These questions point toward a richer spatial and longitudinal analysis that an expanded dataset could support.