Dallas-Fort Worth Real Estate Market 2026

The Verus-AI forecast for Dallas-Fort Worth positions the metro as a structurally sound residential market entering 2026, a conclusion consistent with PwC's ranking of the region as its top-rated market. Across all 1,481 scored Dallas-Fort Worth tracts, the mean five-year forecast stands at +32.93%, and the median Verus-AI score is 59 on a 100-point scale. Those aggregate figures suggest broad, if uneven, appreciation potential rather than a market operating on a narrow base of outperforming locations.

At the apex of the distribution, two tracts, Tract 0548 and Tract 9205, share the peak Verus-AI score of 84 among the 20 tracts shown in the ranked tables, each graded B+ with a Low risk designation and a five-year forecast of 37.01%. That score of 84 also represents the maximum observed across all 1,481 scored Dallas-Fort Worth tracts, as confirmed by the scored-universe data. The separation between these two leading tracts and the broader field is meaningful: the 90th-percentile score across the scored universe sits at 70, placing the top-ranked tracts well clear of the upper decile threshold.

The distribution is not without stress points. All five of the bottom-ranked tracts in the ranked tables carry Verus-AI scores of 0, F grades, High risk designations, and five-year forecasts of -4.9%. The contrast between the top and bottom of the ranked set is sharp, but the negative-forecast cohort remains small in the context of the full scored universe, a point developed further in the forecast distribution section below.

The Verus-AI score is constructed from a history window spanning 2014 through 2024 and projects forward across a five-year forecast window of 2025 through 2029. Scores are expressed on a scale of 0 to 100. Confidence bands reported alongside point forecasts represent 80% intervals, meaning the model assigns roughly a one-in-five probability that outcomes fall outside the stated range. Readers should treat the bands as an honest representation of model uncertainty rather than as bounds on possible outcomes.

Dallas-Fort Worth comprises 1,822 constituent census tracts in total. Of those, 1,481, representing 81.3% of the tract universe, carry sufficient transaction history to be scored. The remaining 341 tracts were deemed unscoreable due to insufficient price history across the 2014-2024 observation window. The scored population of 1,481 tracts covers a total metro population of 9,519,999, with a population-weighted median home value of $339,600. Analysts working from the ranked tables below should note that the 20 tracts displayed, the top 15 and bottom 5 by Verus-AI score, represent a curated slice of that broader universe, not a complete enumeration.

The Verus-AI score integrates multiple signals, including income trajectory, rent-to-price yield, and modeled price appreciation, into a single composite. Grade designations (B+, F, and intermediate grades) and risk labels (Low, Moderate, High) are derived from score thresholds applied consistently across the scored universe. Where a rent-to-price figure is unavailable for a specific tract, as is the case for Tract 0317, the one tract in the ranked set for which that field is not populated, the score is computed from the remaining available inputs.

The shape of the five-year forecast distribution across all 1,481 scored Dallas-Fort Worth tracts is notably right-compressed. The 50th, 75th, and 90th percentiles all resolve to 37.01%, indicating that a large share of the scored universe clusters at or near the model's upper forecast band. The 25th percentile sits at 31.25% and the 10th percentile at 23.88%, meaning that even tracts in the lower decile of the forecast distribution are projected to appreciate in absolute terms over the five-year horizon. The mean five-year forecast of 32.93% sits modestly below the median of 37.01%, pulled down by the small cohort of tracts carrying negative forecasts.

That negative-forecast cohort is narrow. The floor of the distribution is -4.9%, shared by the five bottom-ranked tracts in the ranked tables as well as additional tracts in the broader universe. The concentration of negative forecasts at a single floor value, rather than a continuous distribution of declining projections, is a structural feature of the model output worth noting: it suggests the model identifies a discrete group of distressed locations rather than a gradual gradient of deterioration across the lower tail.

The forecast chart for the leading tract, Tract 0548, illustrates the trajectory clearly. Observed values rose from $187,900 in 2014 to $478,000 in 2024, a compounding appreciation over the ten-year history window. The Verus-AI point forecast carries that value to $654,901 by 2029, implying a five-year CAGR of 6.5%. The 80% confidence band at the terminal year spans from $570,768 to $751,436, a width of $180,668, reflecting the model's honest acknowledgment that the range of plausible outcomes widens materially as the horizon extends. Readers should consult the forecast chart directly to observe how band width expands year by year from 2025 onward.

Several patterns in the top-15 cohort are worth highlighting beyond the raw scores. First, the income growth figures across the top-ranked tracts are uniformly positive, ranging from 6.36% for Tract 0521 to 15.38% for Tract 1521 among the 20 tracts in the ranked set. That consistency of income momentum across the top tier is a meaningful signal; the model treats income trajectory as a forward-looking indicator of demand sustainability, not merely a contemporaneous measure of affordability.

Second, the rent-to-price yields in the top-15 cohort span a wider range than the score dispersion might suggest. Tract 5600 carries an annualized rent-to-price ratio of 10.16%, the highest among the top-15 tracts in the ranked set, against a current value of $206,374, the lowest current value in that cohort. Tract 9210, by contrast, carries a rent-to-price ratio of 5.59% against a current value of $672,300. The model assigns both tracts a score of 81 and a B+ grade, indicating that yield and price level are not the primary differentiators at the top of the distribution; income growth and forecast appreciation carry comparable weight.

Third, and perhaps most instructive for readers focused on risk, the bottom-ranked tracts in the ranked table present a heterogeneous picture despite sharing identical scores, grades, and forecasts. Tract 4203 recorded an income decline of -23.87% year-over-year, the sharpest income decline among the 20 tracts shown in the ranked tables, while Tract 3622 recorded a decline of -14.09%. The rent-to-price ratio for Tract 3713 stands at 79.93%, an extreme outlier relative to the rest of the ranked set, suggesting a deeply distressed price base rather than a high-yield investment profile. Each of the five bottom-ranked tracts carries a five-year forecast of -4.9% and an annualized CAGR of -1.0%, meaning the model anticipates modest but sustained value erosion in those locations rather than a sharp correction.

The momentum table surfaces a separate and important divergence. Tract 3627 leads the momentum table with a 29.08% income gain, the highest year-over-year income growth among the 10 tracts shown in the momentum table, yet carries a Verus-AI score of only 15 and a five-year forecast of -4.9%. Tract 1645 similarly records a 21.38% income gain alongside a Verus-AI score of 15 and a five-year forecast of -4.9%. The model's multi-factor architecture appears to discount short-term income surges when other structural signals, price level, historical appreciation trajectory, and risk indicators, do not corroborate the momentum reading.

The data does not support a market-wide crash scenario for Dallas-Fort Worth over the 2025-2029 forecast horizon. The population-weighted median home value of $339,600 and the mean Verus-AI score of 58.26 both indicate a market operating well above distress thresholds in aggregate.

The score distribution reinforces this reading. The 10th-percentile Verus-AI score across the scored universe is 47, and the minimum observed score is 0, but that minimum is confined to a small number of tracts, not representative of the distribution's central tendency. The interquartile range runs from 53 to 65, a relatively tight band that suggests the middle of the market is reasonably homogeneous in quality, without the wide dispersion that would characterize a market with systemic structural weakness.

That said, the 17 tracts carrying negative forecasts are not trivially small in absolute terms, and the -4.9% floor forecast, paired with a -1.0% annualized CAGR, represents a meaningful divergence from the metro mean for the households and investors concentrated in those locations. The model does not project a recovery for those tracts within the 2025-2029 window. Readers should consult the ranked table and the broader scored-universe data to identify whether specific tracts of interest fall within or near that distressed cohort before drawing conclusions about any particular location.

The Verus-AI forecast is a model-derived estimate grounded in the 2014-2024 history window. It does not incorporate forward-looking assumptions about interest rate policy, employment shocks, or supply-side changes that fall outside the historical pattern of the observation period. The 80% confidence bands reported for each tract's forecast, illustrated most clearly in the forecast chart for Tract 0548, where the terminal-year band spans $570,768 to $751,436, are the appropriate frame for interpreting the precision of any individual tract estimate. A one-in-five probability of outcomes outside that range is not negligible over a five-year horizon.

Coverage limitations are also relevant. The 341 unscoreable tracts, 18.7% of the 1,822 constituent tracts in the Dallas-Fort Worth metro, are excluded from all distributional statistics reported here. If the unscoreable tracts are systematically concentrated in lower-quality or higher-volatility locations, the scored-universe statistics may overstate the metro's aggregate quality. The data does not provide a basis for resolving that question, and readers should treat the 81.3% coverage figure as a material caveat rather than a minor footnote.

Income figures are year-over-year point estimates and are subject to revision as survey data is updated. The extreme income readings in the momentum table, including the 29.08% gain for Tract 3627 and the -23.87% decline for Tract 4203, may reflect data revisions, boundary changes, or single-year anomalies as much as genuine structural shifts. The Verus-AI model's apparent discounting of those extreme readings in its multi-year forecast is consistent with a methodology that weights sustained trend over single-period momentum, but it does not eliminate the possibility that those income signals are early indicators of directional change that the model has not yet fully incorporated. Institutional users are encouraged to treat the model output as one input among several rather than a standalone decision tool.