Most Undervalued Housing Markets in 2026

Across the 7,233 scoreable tracts across 13 major U.S. metros, the Verus-AI model identifies a cohort of census tracts where quality-adjusted pricing diverges sharply from what underlying fundamentals would imply. The leading entry, Houston's Tract 2300, carries a value premium of 94.19%, meaning the model estimates the tract's quality score places it near the 97th percentile of its metro peers, while its current value of $101,298 sits at only the 3rd percentile of metro values. That gap is the widest within-metro mispricing reading among the 15 most-undervalued tracts in the dataset.

The finding is not uniformly distributed. That concentration suggests the mispricing signal is partly a function of metro-level price dispersion: markets with wide internal value spreads generate more extreme quality-to-price gaps at both ends of the distribution. The ranked table below presents all 15 entries with their value premiums, Verus-AI scores, risk grades, and forecast values.

The Verus-AI score is a composite quality index scaled from 0 to 100, calibrated against a history window spanning 2014 to 2024 and applied across the 82,179 scoreable U.S. census tracts. The primary ranking metric on this page is the within-metro value premium: the gap, expressed in percentage points, between a tract's score percentile and its current-value percentile within its metro. A tract scoring at the 97th percentile for quality but priced at the 3rd percentile for value produces a value premium near 94%, as seen in the leading entry.

Forecasts cover the 2025 to 2029 window and are reported as 5-year percentage changes. Confidence bands are 80% intervals, they describe the range within which the model expects the outcome to fall eight times in ten, not a guaranteed corridor. A structural cap of 37.01% applies to the 5-year forecast; 23,955 tracts across the 82,179 scoreable U.S. census tracts reach this ceiling, which compresses the apparent upside for the highest-scoring tracts and should be understood as a conservative modeling choice rather than a statement about market potential. Coverage reaches 84.09% of all U.S. census tracts across 3,234 counties.

The within-metro value premium is intentionally a relative measure. A tract with a high value premium is not necessarily cheap in absolute terms; it is priced low relative to what its quality signals would predict within its own metro. Readers should consult the ranked table below alongside the absolute current-value column, which ranges from $96,690 for San Antonio's West Side to $275,000 for Austin's Round Rock among the 15 most-undervalued entries.

The ranked table below lists the 15 most-undervalued tracts across the 7,233 scoreable tracts across 13 major U.S. metros, sorted by value premium in descending order. The value premiums range from 89.94% for Dallas-Fort Worth's Tract 9101 to 94.19% for Houston's Tract 2300, a relatively narrow band of roughly four percentage points separating the top from the 15th entry, which indicates that the threshold for inclusion is tight and that the tracts listed are genuinely clustered rather than spread across a wide spectrum.

Houston's representation across four entries is notable given that the metro's mean Verus-AI score of 51.18 is the lowest among the 13 curated launch metros, just below Atlanta at 52.65. The metro's weighted median value of $272,400 is also among the lower readings in the dataset. The combination of a wide internal score distribution and a compressed absolute price level creates conditions where high-scoring tracts can appear dramatically underpriced relative to their quality percentile.

Austin's three entries, Round Rock, Taylor, and Tract 1300, are distinguished by their score percentiles. Taylor registers at the 99.14th percentile with a current value of $275,000. Both carry Low risk grades. The Dallas-Fort Worth entries span Moderate and Low risk grades, with value premiums between 89.94% and 92.57%. Charlotte's two entries, Tract 4000 and Tract 4500, both carry Verus-AI scores of 72 and Moderate risk grades, with current values of $163,200 and $166,133 respectively.

The 15 most-overvalued tracts across the 7,233 scoreable tracts across 13 major U.S. metros carry value premiums ranging from -93.97% to -98.44%, indicating that within the same metros where undervaluation appears, premium enclaves trade at valuations the model regards as substantially disconnected from underlying quality signals.

Dallas-Fort Worth's Tract 0600 carries a value premium of -98.44%, the most negative reading among the 15 most-overvalued tracts in the dataset, with a current value of $1,915,530, a Verus-AI score of only 16, and an Elevated risk grade. The intra-metro spread between these extremes is not a modeling artifact; it reflects the model's assessment that quality signals and price signals have diverged in opposite directions across different parts of the same metropolitan area.

Three Phoenix tracts, Tract 5004, Tract 5003, and Tract 5002, cluster among the most-overvalued entries within the 7,233 scoreable tracts across 13 major U.S. metros, with value premiums of -98.2%, -97.9%, and -97.2% respectively, and Verus-AI scores of 33, 33, and 39. Phoenix does not appear on the undervalued list at all, despite having the highest mean Verus-AI score of 64.02 among the 13 curated launch metros. This suggests that Phoenix's quality advantage is broadly priced in at the metro level, leaving fewer pockets where quality substantially outpaces price.

The forecast chart for the leading tract, Houston's Tract 2300, illustrates the model's trajectory from an observed value of $66,300 in 2014 to $101,298 in 2024, with a forecast path reaching $138,787 by 2029. The 80% confidence band at the terminal year spans $124,211 to $155,074, a width of $30,863. The band widens progressively across the five forecast years, as shown in the forecast chart, which is consistent with standard uncertainty accumulation over longer horizons. The observed series shows a period of relative stagnation from 2014 through 2019, followed by acceleration beginning in 2020.

The structured data does not include direct rental yield figures, so no yield calculations appear in this analysis. However, the absolute price levels of the most-undervalued tracts are relevant context for readers who apply yield frameworks independently. Among the 15 most-undervalued entries, current values range from $96,690 at San Antonio's West Side to $275,000 at Austin's Taylor. At these price points, even modest rental income streams would imply yield levels that are difficult to achieve in higher-priced submarkets within the same metros.

The within-metro value premium metric used here is a quality-adjusted price signal, not a yield signal, and the two can diverge. A tract with a high value premium may carry a high yield if rents track quality more closely than prices do, or it may carry a lower yield if the area's rental market is also depressed relative to quality. The Verus-AI score incorporates a range of quality signals over the 2014 to 2024 history window, but readers seeking to layer rental yield analysis onto these rankings should source rental data independently and treat the value premium as one input among several.

The value premium metric is a within-metro relative measure, and its interpretation depends on the stability of the metro-level quality distribution. If a metro undergoes structural change, demographic shifts, infrastructure investment, or sustained employment contraction, the percentile rankings that underpin the value premium can shift, potentially narrowing or reversing apparent mispricing without any change in absolute prices. The model's history window of 2014 to 2024 captures one full cycle of post-crisis recovery and pandemic-era price acceleration, but it does not guarantee that the relationships observed over that period will persist through the 2025 to 2029 forecast window.

Risk grades in the dataset range from Low to Elevated. Among the 15 most-undervalued tracts, the majority carry Low or Moderate risk grades, with no Elevated or High grades present. Among the 15 most-overvalued tracts, risk grades are more varied: Dallas-Fort Worth's Tract 0600 and Houston's River Oaks carry Elevated grades, Phoenix's Tract 8100 carries a High grade, and several others carry Moderate or Low grades. The presence of Low risk grades on some overvalued tracts, such as Dallas-Fort Worth's Tract 9700 and Tract 2100, is a reminder that the risk grade and the value premium are distinct signals that do not always align.

The forecast cap of 37.01% applies to a substantial share of tracts: 23,955 of the 82,179 scoreable U.S. census tracts reach this ceiling. The 80% confidence bands should be read as intervals, not as bounds on possible outcomes; realized values could fall outside the stated range in two of ten scenarios by construction. Climate, demographic, and forecast figures are model-derived estimates based on historical patterns and should not be interpreted as guarantees of future performance.