A New Asset-Pricing Model Adds Spatial Interactions to High-Dimensional Markets

Asset prices are not always explained by one market force alone. This paper adds another layer: spatial interactions, a way to let high-dimensional assets influence one another inside pricing models. The authors define a new quantity called spatial rho, a counterpart to market beta, and build a Spatial Capital Asset Pricing Model before extending it to a broader Spatial Arbitrage Pricing Theory with multiple factors. For observable factors, they propose a generalized shrinkage Yule-Walker estimator that uses ridge regression, a stabilizing tool for cases with many variables. When the factors are hidden, they first use autocovariance-based eigenanalysis to extract them, then apply the same estimation strategy. The paper also establishes asymptotic properties for these estimators when both the dimension and sample size grow. Simulated and real-data examples are used to show the method’s effectiveness and usefulness.

A single market beta can miss the way assets move in packs. If one name slips, related names often feel it too. This model tries to catch that second push. It adds spatial links, so assets can tug on one another inside the pricing rule. The key new signal is spatial rho. It plays the same role for connections that beta, the usual market sensitivity score, plays for the market. That matters when the asset list keeps growing. It also matters when simple stories break down under crowd motion. If you watch a portfolio swing as a group, this model is aimed at that pattern. It gives that pattern a name and a way to test it.

What spatial rho adds to asset pricing

SCAPM is the first step. It stands for Spatial Capital Asset Pricing Model. It adds a beta-like measure for links between assets. The model calls that measure spatial rho. SAPT is the wider frame. It stands for Spatial Arbitrage Pricing Theory. Arbitrage means a price gap that should not last. SAPT brings in several factors instead of one. It extends under a complete market setting. That means the factor mix can pin down prices. The model works with both seen factors and hidden ones. It also gives asymptotic results. Those results say the estimators stay well behaved as the number of assets and the sample size grow. Simulated and real data examples then show the approach is useful.

How the fit works when factors are hidden

For seen factors, the model uses generalized shrinkage Yule-Walker estimation. Yule-Walker is a way to fit a model from time-linked variation. Shrinkage means it pulls wild estimates back toward calm values. The ridge part, or ridge regression, does that with a small penalty. That helps when many assets move at once. When factors are hidden, the model starts with autocovariance-based eigenanalysis. Autocovariance tracks how a series moves with itself over time. Eigenanalysis ranks the main directions in that pattern. Those directions reveal the hidden factors. The same SYW fit then uses the estimated factors. That two-step path keeps the fit stable.

Why these links matter in crowded markets

This matters because asset returns rarely move one at a time. They often move in groups. A model with spatial rho can describe that shared motion. It can also do it when some drivers are visible and others are not. That makes the frame broader than a single market beta. It also gives a clear route for estimation in crowded data. The shrinkage step reins in noise. The factor-extraction step can recover hidden structure. Together, those pieces turn a hard many-asset problem into one you can fit and test. That is the paper's main promise.

What to test next

The sharp next test is a market with even more moving parts. The model has already faced simulated and real data examples. The open question is how far spatial rho travels when the asset list grows again. That test would probe the surprise at the center of the frame. A beta-like signal can now sit beside a web of links. If that holds, pricing models can treat crowd motion as part of the signal. They would not see it as background noise alone. That is a concrete change in how crowded markets get read.