Buying a Mall: Can Acquisition Valuation Influence Its Future Value?

From an investment perspective, this suggests that part of a shopping center’s future value may already be embedded in the price paid at acquisition, while another part responds to factors that the initial price does not capture. The result does not imply that assets acquired at higher valuations lose value or that those acquired at lower valuations systematically generate higher returns. Rather, it shows that initial valuation differences tend to be only partially preserved.

What explains that behavior is a different question. The data identify the phenomenon, but they do not establish its cause. Even so, at least four economic hypotheses are consistent with the evidence.

The first—and perhaps the one most consistent with the findings—is that the market had already priced in a significant portion of the expected growth for assets acquired at higher valuations. In that case, a higher valuation would reflect not only the asset’s current characteristics, but also expectations regarding its future performance. If those expectations were already embedded in the acquisition price, the potential for further value appreciation would naturally tend to be smaller.

A second explanation is that assets acquired at lower valuations had greater room to improve occupancy, rents, or commercial quality through renovations, repositioning strategies, or changes in tenant mix. If so, part of the initial gap between assets would naturally tend to narrow.

A third hypothesis is that the real estate market tends to moderate extreme valuation differences as it incorporates new information about asset performance, a pattern consistent with convergence processes documented in other markets.

Finally, changes in cap rates or financing conditions could also contribute to this behavior. Evaluating that possibility, however, would require additional information beyond the scope of this analysis.

These hypotheses, individually or in combination, extend beyond the specific case of FIBRA Shop and suggest that the conditions under which an asset enters a portfolio appear to shape part of its subsequent performance. However, they do not explain it entirely. From that perspective, when major portfolio managers—such as Mexican REITs—change the composition of their portfolios, the relevant question is not only what they buy or sell, but also at what valuation they do so. That valuation not only determines the acquisition cost; it may also define part of the value creation potential the asset will retain in the years ahead.

To learn more about Mexico’s retail real estate market, visit SiiLA Market Analytics or contact us at contacto@siila.com.mx.

***

¹ The analysis covers the 18 FIBRA Shop shopping centers with a single identifiable acquisition date. Acquisition values were adjusted to constant June 2026 pesos using Mexico’s National Consumer Price Index (INPC) and compared with appraised values reported for Q2 2026. The relationship between changes in asset value and portfolio holding period, gross leasable area (GLA), and initial real valuation per square meter was evaluated using correlations and regression models with HC3 robust standard errors, complemented by robust regressions, bootstrap resampling (5,000 iterations), nonlinear specifications, influential observation analysis, and sensitivity tests. Holding period (Pearson p=0.84; Spearman p=0.76; regression p=0.91) and asset size (p=0.41) showed no statistically significant associations. A permutation test also indicated that the relationship between acquisition price and the value creation multiple was driven primarily by a mathematical identity. Accordingly, the economic analysis focused on the relationship between real acquisition valuation and Q2 2026 valuation, both expressed per square meter.

² Elasticity was estimated using a log-linear model relating real acquisition valuation and Q2 2026 valuation, both expressed per square meter. The estimated coefficient was β=0.49 (p=0.013) and β=0.55 (p=0.005) after simultaneously controlling for portfolio holding period and gross leasable area (GLA). In both cases, the null hypothesis H₀: β=1 was rejected (p<0.05), indicating that initial valuation differences are not fully reflected in the valuations observed in Q2 2026. The stability of the estimates was confirmed using HC3 robust standard errors, robust regressions, bootstrap resampling (5,000 iterations), influential observation analysis, nonlinear specifications, and sensitivity tests.