Estimated Dynamic Equilibrium Model: Supply and Demand as a Sample Path of a Stochastic Process
Pith reviewed 2026-05-19 14:17 UTC · model grok-4.3
The pith
Sequentially recycling upper-tail market-clearing prices from noisy bids generates upward price drift even with zero-mean errors.
A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.
Core claim
In the Estimated Dynamic Equilibrium Model, treating supply and demand as a sample path of a coupled stochastic process reveals that sampling clearing prices from the upper tail of noisy bid distributions and recycling them as future valuation inputs creates an order-statistic bias. This bias causes expected prices to drift upward despite strictly zero-mean estimation errors, and compounding the bias across periods yields exponential price growth in simulations.
What carries the argument
The feedback loop in which market-clearing prices sampled from the upper tail of noisy bid distributions are recycled directly into the next period's heterogeneous valuations.
If this is right
- Expected prices exhibit upward drift and can grow exponentially over successive periods.
- Controlled simulations produce six distinct regimes from band-stability to runaway bubbles under a single set of agent rules.
- The framework recovers Walrasian equilibrium as a limiting case when the upper-tail recycling is absent.
- Machine-learning valuation algorithms that rely on recent clearing prices may amplify the inherent statistical bias.
- The mechanism provides a potential account for varying empirical results on the effects of opinion divergence in markets.
Where Pith is reading between the lines
- Similar feedback from tail-sampled prices could operate in other markets with noisy valuations, such as labor or asset markets, potentially leading to persistent mispricing.
- Introducing external information or arbitrage opportunities that reset the reference price might eliminate the drift, offering a testable way to stabilize prices.
- Extending the model to include strategic agent behavior or correlated errors could reveal interactions that either dampen or accelerate the price growth.
- If valuation models in practice use recent transaction prices, they may contribute to observed market trends beyond fundamentals.
Load-bearing premise
The previous market-clearing price, selected from the upper tail of bids, serves directly as the reference point for constructing the next period's valuations without any intervening factors that would interrupt the feedback.
What would settle it
Running a simulation or observing a market where the reference valuation for the next period is instead drawn from the median or mean of bids rather than the upper-tail clearing price, and checking whether the upward price drift disappears.
Figures
read the original abstract
We introduce the Estimated Dynamic Equilibrium Model (EDEM), an agent-based framework that treats supply and demand as a coupled stochastic process driven by heterogeneous, noisy agent valuations. The model's primary technical contribution is the identification of a generative mechanism for persistent disequilibrium: when market-clearing prices are sequentially sampled from the upper tail of noisy bid distributions and recycled as inputs for future valuations, expected prices drift upward despite strictly zero-mean estimation errors. We derive this order-statistic bias in closed form for i.i.d. uniform bids and use simulations to show that compounding this bias across epochs yields exponential price growth without requiring assumptions of investor optimism or irrationality. This framework extends Miller's divergence-of-opinion theory to a dynamic setting, recovering Walrasian equilibrium and Miller's static premium as limiting cases. Through controlled experiments and sensitivity analysis on a simulated real-estate neighborhood, we identify six distinct regimes-ranging from band-stability to runaway bubbles-emerging from a single agent ruleset. These results offer a potential explanation for the contradictory findings in the empirical divergence-of-opinion literature and suggest that machine-learning valuation algorithms may inadvertently amplify this inherent statistical bias.
Editorial analysis
A structured set of objections, weighed in public.
Referee Report
Summary. The manuscript introduces the Estimated Dynamic Equilibrium Model (EDEM), an agent-based framework that models supply and demand as a coupled stochastic process driven by heterogeneous noisy agent valuations. Its central claim is that sequentially sampling market-clearing prices from the upper tail of noisy bid distributions and recycling those prices as anchors for subsequent valuations produces an upward drift in expected prices due to order-statistic bias, even when estimation errors are strictly zero-mean; this bias is derived in closed form for i.i.d. uniform bids and, when compounded across periods via simulation, yields exponential price growth and six distinct regimes (from band-stability to runaway bubbles) without invoking investor optimism or irrationality. The model recovers Walrasian equilibrium and Miller's static premium as limiting cases and is illustrated on a simulated real-estate neighborhood.
Significance. If the iterative bias mechanism holds, the paper supplies a purely statistical generative account of persistent disequilibrium and asset-price bubbles that extends Miller's divergence-of-opinion theory to a dynamic setting while remaining consistent with rational agents. The closed-form single-period derivation and the emergence of multiple regimes from a single ruleset constitute concrete strengths; the controlled experiments on simulated real-estate data further ground the analysis. These elements could help reconcile contradictory empirical findings on divergence of opinion and highlight potential amplification effects in machine-learning valuation systems.
major comments (2)
- [§3] §3 (closed-form derivation for i.i.d. uniform bids): the single-period order-statistic bias for the expected clearing price is correctly obtained from standard uniform-order-statistic results, but the claim that this bias compounds exponentially across epochs presupposes that each period's bids remain unconditionally i.i.d. uniform. The recycling rule (prior clearing price used as the reference point for the next period's heterogeneous valuations) necessarily induces serial dependence and shifts the support of the bid distribution, so the single-period formula does not automatically iterate; the exponential-growth result therefore requires an additional argument that the i.i.d. property is restored each period.
- [Simulation experiments] Simulation experiments (real-estate neighborhood): the reported regimes (band-stability to runaway bubbles) are described qualitatively; no sample sizes, number of Monte-Carlo replications, standard errors on growth rates, or sensitivity checks to the bid-distribution family are provided. Without these quantities it is impossible to assess whether the observed exponential drift survives the dependence induced by recycling or is an artifact of redrawing from a fixed support.
minor comments (2)
- [Abstract] The abstract states that six regimes are identified but neither enumerates nor briefly characterizes them; a short list or one-sentence description would improve readability.
- [Model setup] Notation for the recycling rule and the precise mapping from prior clearing price to next-period valuation anchors should be introduced with an equation or diagram in the model-setup section to avoid ambiguity when the dependence issue is discussed.
Simulated Author's Rebuttal
We thank the referee for the constructive comments, which have helped us improve the clarity and rigor of the manuscript. We address the major comments point by point below.
read point-by-point responses
-
Referee: [§3] §3 (closed-form derivation for i.i.d. uniform bids): the single-period order-statistic bias for the expected clearing price is correctly obtained from standard uniform-order-statistic results, but the claim that this bias compounds exponentially across epochs presupposes that each period's bids remain unconditionally i.i.d. uniform. The recycling rule (prior clearing price used as the reference point for the next period's heterogeneous valuations) necessarily induces serial dependence and shifts the support of the bid distribution, so the single-period formula does not automatically iterate; the exponential-growth result therefore requires an additional argument that the i.i.d. property is restored each period.
Authors: We thank the referee for pointing out the need for an explicit argument regarding the iteration under serial dependence. Although the recycling of the clearing price as an anchor induces dependence across periods, the agent valuations are redrawn each period from a distribution centered at the anchor with i.i.d. noise terms whose distribution does not depend on history. This restores the i.i.d. uniform property conditionally on the previous price, allowing the single-period bias formula to be applied iteratively. We will add a subsection in the revised §3 deriving the multi-period expectation under this conditional i.i.d. structure to rigorously support the exponential growth claim observed in the simulations. revision: yes
-
Referee: [Simulation experiments] Simulation experiments (real-estate neighborhood): the reported regimes (band-stability to runaway bubbles) are described qualitatively; no sample sizes, number of Monte-Carlo replications, standard errors on growth rates, or sensitivity checks to the bid-distribution family are provided. Without these quantities it is impossible to assess whether the observed exponential drift survives the dependence induced by recycling or is an artifact of redrawing from a fixed support.
Authors: We agree that the simulation section would benefit from more quantitative details to allow readers to evaluate the robustness of the results. In the revised manuscript, we will include the number of Monte Carlo replications used (1000), the sample size of the simulated real-estate neighborhood, standard errors for the growth rate estimates across regimes, and additional sensitivity analyses varying the bid distribution family (including normal and beta distributions in addition to uniform). These additions will demonstrate that the upward drift and regime structure persist under the serial dependence induced by price recycling. revision: yes
Circularity Check
No significant circularity: bias derived from standard order statistics; dynamic growth follows explicitly stated recycling rule
full rationale
The paper derives the single-period upward bias in closed form under i.i.d. uniform bids using order statistics, then shows compounding via simulations that implement the recycling of prior clearing prices as valuation anchors. This generative setup produces the drift by construction of the stated rules (zero-mean noise plus upper-tail sampling plus feedback), but the mathematical steps do not reduce to a fitted parameter, self-definition, or unverified self-citation chain. The central result remains independently verifiable from the model equations without tautological collapse. No load-bearing self-citations or ansatzes are identified in the abstract or description.
Axiom & Free-Parameter Ledger
axioms (1)
- domain assumption Agent bids are i.i.d. uniform random variables
invented entities (1)
-
Estimated Dynamic Equilibrium Model (EDEM)
no independent evidence
Lean theorems connected to this paper
-
IndisputableMonolith/Cost/FunctionalEquation.leanwashburn_uniqueness_aczel unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
E[max_i β_i / v_t(h(s))] = 1 + σ(n−1)/(n+1) ... compounded across epochs yields exponential price growth
-
IndisputableMonolith/Foundation/ArithmeticFromLogic.leanembed_add unclear?
unclearRelation between the paper passage and the cited Recognition theorem.
vt+T(h) = v_t(h) · r̄_t ... r̄_t = average of min winning ratios
What do these tags mean?
- matches
- The paper's claim is directly supported by a theorem in the formal canon.
- supports
- The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
- extends
- The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
- uses
- The paper appears to rely on the theorem as machinery.
- contradicts
- The paper's claim conflicts with a theorem or certificate in the canon.
- unclear
- Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.
Reference graph
Works this paper leans on
- [1]
-
[2]
and Kim, Chansog (Francis) and Pantzalis, Christos , title =
Doukas, John A. and Kim, Chansog (Francis) and Pantzalis, Christos , title =. Journal of Financial and Quantitative Analysis , volume =. 2006 , doi =
work page 2006
- [3]
- [4]
- [5]
-
[6]
The American Economic Review , volume =
Mayshar, Joram , title =. The American Economic Review , volume =
- [7]
- [8]
-
[9]
Epstein, Larry G. and Wang, Tan , title =. Econometrica , volume =
-
[10]
Proceedings of the 14th Python in Science Conference (SciPy 2015) , pages =
Masad, David and Kazil, Jacqueline , title =. Proceedings of the 14th Python in Science Conference (SciPy 2015) , pages =
work page 2015
-
[11]
Grimm, Volker and Railsback, Steven F. and Vincenot, Christian E. and Berger, Uta and Gallagher, Cara and DeAngelis, Donald L. and Edmonds, Bruce and Ge, Jiaqi and Giske, Jarl and Groeneveld, J. The. Journal of Artificial Societies and Social Simulation , volume =. 2020 , doi =
work page 2020
-
[12]
Doyne and Hinterschweiger, Marc and Low, Katie and Tang, Daniel and Uluc, Arzu , title =
Baptista, Rafa and Farmer, J. Doyne and Hinterschweiger, Marc and Low, Katie and Tang, Daniel and Uluc, Arzu , title =. Bank of England Staff Working Paper , number =
-
[13]
Geanakoplos, John and Axtell, Robert and Farmer, J. Doyne and Howitt, Peter and Conlee, Benjamin and Goldstein, Jonathan and Hendrey, Matthew and Palmer, Nathan M. and Yang, Chun-Yi , title =. American Economic Review , volume =
-
[14]
Zillow Group Reports Third Quarter 2021 Financial Results, Announces Wind Down of. 2021 , month = nov, day =
work page 2021
-
[15]
Arbuzov, Mikhail , title =
- [16]
discussion (0)
Sign in with ORCID, Apple, or X to comment. Anyone can read and Pith papers without signing in.