Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Modeling the Feedback of AI Price Estimations on Actual Market Values

Viorel Silaghi, Zobaida Alssadi, Ben Mathew, Majed Alotaibi, Ali Alqarni, Marius Silaghi

TL;DR

This work examines how public AI price estimates from Mass Real Estate Estimators (MREEs) like Zillow can trigger a recursive inflation in housing markets through feedback with homeowners. It introduces the Real Estate Prediction Problem (REPP) framework, modeling houses as nodes on a graph with price components $v$ (construction features) and $u$ (owner-estimated features), plus location and market-adjustment terms $oldsymbol{}$ and $oldsymbol{ ho}$ that update via transaction-driven rules $P$, $oldsymbol{}$, and $oldsymbol{ ho}$. The key finding is that estimation errors produce inflation proportional to their absolute size $|oldsymbol{ riangle}|$, propagating through neighboring prices and updating location estimates, creating a reinforcement loop. A policy-relevant insight is that restricting MREE visibility to opt-in listings can significantly dampen inflationary dynamics, though full opt-in can maximize inflation under certain conditions; the results highlight a potential mechanism for stabilizing AI-assisted housing markets amid pervasive price signaling. The framework and simulations offer a tractable, polynomial-time approach to analyze AI-human feedback in markets and inform governance of AI-assisted real estate data disclosures.

Abstract

Public availability of Artificial Intelligence generated information can change the markets forever, and its factoring into economical dynamics may take economists by surprise, out-dating models and schools of thought. Real estate hyper-inflation is not a new phenomenon but its consistent and almost monotonous persistence over 12 years, coinciding with prominence of public estimation information from Zillow, a successful Mass Real Estate Estimator (MREE), could not escape unobserved. What we model is a repetitive theoretical game between the MREE and the home owners, where each player has secret information and expertise. If the intention is to keep housing affordable and maintain old American lifestyle with broad home-ownership, new challenges are defined. Simulations show that a simple restriction of MREE-style price estimation availability to opt-in properties may help partially reduce feedback loop by acting on its likely causes, as suggested by experimental simulation models. The conjecture that the MREE pressure on real estate inflation rate is correlated with the absolute MREE estimation errors, which is logically explainable, is then validated in simulations.

Modeling the Feedback of AI Price Estimations on Actual Market Values

TL;DR

This work examines how public AI price estimates from Mass Real Estate Estimators (MREEs) like Zillow can trigger a recursive inflation in housing markets through feedback with homeowners. It introduces the Real Estate Prediction Problem (REPP) framework, modeling houses as nodes on a graph with price components (construction features) and (owner-estimated features), plus location and market-adjustment terms and that update via transaction-driven rules , , and . The key finding is that estimation errors produce inflation proportional to their absolute size , propagating through neighboring prices and updating location estimates, creating a reinforcement loop. A policy-relevant insight is that restricting MREE visibility to opt-in listings can significantly dampen inflationary dynamics, though full opt-in can maximize inflation under certain conditions; the results highlight a potential mechanism for stabilizing AI-assisted housing markets amid pervasive price signaling. The framework and simulations offer a tractable, polynomial-time approach to analyze AI-human feedback in markets and inform governance of AI-assisted real estate data disclosures.

Abstract

Public availability of Artificial Intelligence generated information can change the markets forever, and its factoring into economical dynamics may take economists by surprise, out-dating models and schools of thought. Real estate hyper-inflation is not a new phenomenon but its consistent and almost monotonous persistence over 12 years, coinciding with prominence of public estimation information from Zillow, a successful Mass Real Estate Estimator (MREE), could not escape unobserved. What we model is a repetitive theoretical game between the MREE and the home owners, where each player has secret information and expertise. If the intention is to keep housing affordable and maintain old American lifestyle with broad home-ownership, new challenges are defined. Simulations show that a simple restriction of MREE-style price estimation availability to opt-in properties may help partially reduce feedback loop by acting on its likely causes, as suggested by experimental simulation models. The conjecture that the MREE pressure on real estate inflation rate is correlated with the absolute MREE estimation errors, which is logically explainable, is then validated in simulations.
Paper Structure (14 sections, 1 theorem, 9 equations, 5 figures, 1 table)

This paper contains 14 sections, 1 theorem, 9 equations, 5 figures, 1 table.

Table of Contents

  1. Introduction
  2. Absolute MREE errors propagating into inflation
  3. Background and Concepts
  4. Case studies of Overestimation and Underestimation Effects
  5. Overestimation of House A by the MREE
  6. Underestimation of House A by the MREE
  7. Assumptions that conditioned the repeated inflation
  8. Model Integrating Information
  9. Studied REPP functions
  10. Model With Information Opt-In
  11. Simulations Discussion
  12. Simulation I: Inflation based on estimated price.
  13. Simulation II: Inflation with opt-in MREE listings.
  14. Conclusions

Key Result

Theorem 1

Assuming $P$ and $\Lambda$ are polynomial, the REPP problem can be solved in polynomial time.

Figures (5)

  • Figure 1: Zillow map with listings (Feb 2022)
  • Figure 2: Zillow map with estimations (Feb 2022)
  • Figure 3: Market inflation as function of absolute error range, various neighborhood sizes: 5 houses to 25 houses
  • Figure 4: Market inflation as function of time in days for the two extreme absolute error ranges considered: $5k houses and $50k houses (green) at neighborhood distance 25
  • Figure 5: Market inflation as function of time at different percentages of opt-in for listings absolute error range $10k at neighborhood distance 10. The numbers next to collors specify the percentage of homes who do not opt in.

Theorems & Definitions (2)

  • Definition 1: REPP
  • Theorem 1