The question

Buyers who use a mortgage and buyers recorded as paying cash can appear to pay different prices for the same home. Is that contrast about financing, or about which people and transactions appear in each group?

The paper follows repeat sales of the same New York City houses and condominiums in 2016–2025, linked to deed and mortgage records, and asks how the price contrast changes as company buyers and possible related-party transfers are separated.

Main findings

  • 9.25log points: financed house purchases above cash purchases of the same house (95% interval 7.17–11.35)
  • 4.87log points after excluding the 358 pairs with a same-surname deed, a marker of possible transfers between relatives (2.69–6.76)
  • 1.53log points when the cash buyer is a company; the interval (−1.79 to 5.44) includes zero and allows a positive gap

House estimates have 95% bootstrap intervals. Financing is classified from linked deed and mortgage records, without verifying purchase finance at every endpoint.

Financed-minus-cash price contrast, log points, with 95% intervals Headline 9.25; excluding same-surname deeds 4.87; company cash buyers 1.53; other cash buyers 13.77; condominiums 0.77. -4 0 4 8 12 16 Log points (financed minus cash, same property) All house pairs (headline): 9.25 log points, 95% interval 7.17 to 11.35 All house pairs (headline) 9.25 Excluding same-surname deeds: 4.87 log points, 95% interval 2.69 to 6.76 Excluding same-surname deeds 4.87 Cash buyer is a company: 1.53 log points, 95% interval -1.79 to 5.44 Cash buyer is a company 1.53 Cash buyer is another party: 13.77 log points, 95% interval 10.96 to 15.98 Cash buyer is another party 13.77 Condominiums (common design): 0.77 log points, 95% interval -0.01 to 1.56 Condominiums (common design) 0.77
Financed-minus-cash contrasts in log points. House rows have 95% bootstrap intervals under the headline specification. The condominium row uses the common borough–quarter design and a parcel-clustered analytic interval; it is not a like-for-like contrast with the restricted house sample. Exact values appear in the table.
Show as a table
SampleEstimate95% interval
All house pairs (headline)9.257.17 to 11.35
Excluding same-surname deeds4.872.69 to 6.76
Cash buyer is a company1.53−1.79 to 5.44
Cash buyer is another party13.7710.96 to 15.98
Condominiums (borough–quarter design)0.77−0.01 to 1.56

Why it matters

  • A cash discount is often read as the value of closing certainty, and proposals to tax cash purchases, including one considered in New York in May 2026, treat it as the cash buyer's advantage. The measured gap depends heavily on who pays cash, which makes it a poor basis for a tax or subsidy on the payment method.
  • The company-cash estimate is smaller and imprecise. Company buyers include LLCs used by households, so they cannot all be treated as large institutional investors.
  • Financing checks found two mortgages missed by the linkage rule. Correcting those two sampled labels moves the main contrast from 9.25 to 9.23 log points.
Housing-finance researchersPolicy analystsUsers of deed records

What the results can tell us

These are associations from linked records. A missing mortgage match does not prove that a purchase was unfinanced, and a shared surname does not establish a family relationship. The estimates do not identify a causal effect of financing or the value of closing certainty. Document checks identify bundle prices and contradictory mortgage index dates. In a new stratified random sample of 40 sale endpoints, AI readings support purchase financing for 22; 18 remain unresolved. These checks support sensitivity analysis, not independently verified financing accuracy.

Cite this paper

BibTeX citation
@misc{loschi2026house,
  author       = {Loschi, Pablo},
  title        = {Who Gets the House? Related-Party Transfers and the
                  Cash--Mortgage Price Gap in New York City},
  year         = {2026},
  howpublished = {Working paper and replication package, Zenodo},
  doi          = {10.5281/zenodo.22421850}
}