VERUS·AI RESEARCH Tract-level housing forecasts

Verus-AI Research

Springfield Real Estate (Jacksonville, 2026)

Springfield (Jacksonville) at the tract level: 9 scored Census tracts ranked by the Verus-AI score, with comparable neighborhoods.

Pop-weighted median value $110,878
Median Verus-AI score 46/100
Scored tracts 9
Mean 5-Yr forecast +34.4%

Where the model sees value

Springfield, also known as Historic Springfield, is an urban neighborhood within Jacksonville, composed of 9 census tracts and a total population of 25,598.

The neighborhood-level mean Verus-AI score of 47.44 and median of 46.0 place the aggregate squarely in the lower-middle tier of the scored distribution, which runs from a minimum of 35.0 to a maximum of 67.0 across the 9 tracts shown in the ranked table. The mean 5-year forecast of 34.39% suggests meaningful nominal appreciation potential in aggregate, but that average is pulled downward by two tracts whose forecasts diverge sharply from the majority, a point the forecast chart makes visually apparent. The neighborhood's internal heterogeneity is not a minor footnote; it is the defining analytical challenge for any tract-level assessment of Springfield.

Median household income varies as dramatically as home values. Tract 12031001100 reports a median household income of $115,385, while tracts 12031000200 and 12031001600 report $16,088 and $16,292, respectively. That income stratification maps closely, though not perfectly, onto the score rankings, and it helps explain why the neighborhood's population-weighted median value of $110,878 understates conditions in the upper tracts while overstating them in the lower ones. Readers should treat the neighborhood-level aggregate as an orientation figure rather than a reliable proxy for any individual tract.

Verus-AI score distribution across Springfield's scored tracts
Verus-AI score distribution across Springfield's scored tracts00122355167Verus-AI score (0-100)
Verus-AI score distribution across Springfield's scored tracts: across the 9 scored Springfield tracts.

Springfield tracts ranked by Verus-AI score

Springfield tracts ranked by Verus-AI score
Rank Tract Verus-AI Score Grade 5-Yr Forecast Current Value Gross Rent Yield
1 12031001100 67 C+ +37.0% $414,195 2.9%
2 12031001200 60 C +37.0% $349,828 4.4%
3 12031001401 50 D+ +37.0% $193,541 6.7%
4 12031002802 48 D +32.4% $76,500 15.4%
5 12031000101 46 D +37.0% $132,400 9.8%
6 12031000102 44 D- +37.0% $124,600 11.2%
7 12031002901 41 D- +37.0% $110,878 12.2%
8 12031000200 36 F +18.1% $88,600 10.7%
9 12031001600 35 F +37.0% $76,400 7.7%
Statistically comparable neighborhoods
Neighborhood Metro Similarity Verus-AI Score Current Value
Tucker Atlanta 99.6% 76 $469,200
Brushy Creek Austin 98.7% 72 $484,000
Bloomingdale Tampa 98.6% 70 $463,400
Springfield summary
Metric Value
Scored tracts 9
Population (scored + unscored) 25,598
Population-weighted median value $110,878
Mean Verus-AI score 47.4 / 100
Median Verus-AI score 46.0 / 100
Forecast spread (p10 to p90) +29.5% to +37.0%
Designated Opportunity-Zone tracts 2 of 9
Most common FEMA climate rating Relatively Low

What is driving the spread

The Verus-AI score distribution across the 9 tracts shown in the ranked table spans from 35 to 67, a range of 32 points on a 100-point scale. The leading tract, 12031001100, at a score of 67 and a grade of C+, sits above the 90th percentile of the score distribution among the 9 tracts shown in the ranked table, where the 90th percentile stands at 61.4.

Of the 9 ranked tracts, 6 carry a risk designation of High and 3 carry Moderate. The three Moderate-risk tracts are 12031001100, 12031002802, and 12031000200. Notably, tract 12031000200 carries both a Moderate risk designation and an F grade, an unusual pairing that reflects the model's separation of risk volatility from absolute quality. The grade distribution across the 9 tracts is concentrated in the lower tiers: 1 C+, 1 C, 1 D+, 2 D, 2 D-, and 2 F. No tract in Springfield carries a grade of B or above.

The forecast picture is similarly bifurcated. Seven of the 9 tracts carry a 5-year forecast of 37.01%, implying a compound annual growth rate of 6.5%. Tract 12031002802 forecasts 32.35% over the same horizon, and tract 12031000200 forecasts only 18.06%, the lowest 5-year forecast among the 9 tracts shown in the ranked table, at a CAGR of 3.38%.

Rent-to-price yields tell a complementary story. Tract 12031002802 carries the highest annual rent-to-price ratio among the 9 tracts shown in the ranked table, at 15.42%, while the leading tract, 12031001100, carries the lowest at 2.94%. This inverse relationship between current value and yield is consistent with a neighborhood where lower-priced tracts offer income-oriented return profiles and higher-priced tracts are priced more for appreciation. Whether that yield premium adequately compensates for the income deterioration visible in several lower-scoring tracts is a question the income year-over-year figures sharpen considerably.

Income trends are among the more concerning signals in the data. Among the 9 tracts shown in the ranked table, the sharpest income decline belongs to tract 12031000200, at -21.26% year-over-year. Tract 12031001600 shows -17.55%, and tract 12031001401 shows -14.66%. Three of the five lowest-scoring tracts recorded negative income growth in the most recent annual period available in the model's data window; the remaining two recorded 0.0%.

The leading tract's historical value series, shown in the forecast chart, illustrates the trajectory that underlies its score. Tract 12031001100 was valued at $196,300 in 2014 and reached $414,195 by 2024, a gain of roughly $218,000 over the history window. The Verus-AI forecast projects a terminal value of $567,483 by 2029, with an 80% confidence interval spanning $462,722 to $695,963. The width of that terminal band, $233,241, reflects genuine uncertainty over a 5-year horizon and should be read as an 80% interval, not a guarantee. The forecast chart renders the band's widening profile year by year, which is the appropriate way to interpret the uncertainty structure.

The Verus-AI model identifies three comparable neighborhoods for Springfield: Tucker (Atlanta metro, similarity 99.6%, score 76, current value $469,200), Brushy Creek (Austin metro, similarity 98.7%, score 72, current value $484,000), and Bloomingdale (Tampa metro, similarity 98.6%, score 70, current value $463,400). All three carry scores above the 90th percentile of the score distribution among the 9 tracts shown in the ranked table, where that threshold stands at 61.4.

The comparison is instructive precisely because it is unflattering in aggregate. The three comparables score between 70 and 76, while Springfield's neighborhood-level mean score is 47.44 and its median is 46.0.

That gap suggests the leading Springfield tract may carry relative value against its structural peers, though the higher risk profile and narrower income base of the Jacksonville market warrant caution in drawing that inference too directly. The similarity percentages are high (98.6% to 99.6%), indicating the model finds strong structural alignment, but structural similarity does not imply identical return outcomes.

Of those, 7 carry a Relatively Low rating, 1 carries a Very Low rating, and 1 carries a Relatively Moderate rating. The modal climate rating of Relatively Low is the most common designation across all 9 scored Springfield tracts. The analysis does not extend these ratings to insurance or lending implications; they are model-derived designations and should be interpreted within that scope.

Two of Springfield's 9 tracts carry Opportunity Zone designations. The OZ overlay is relevant context for certain capital-gains-deferral strategies, though the analysis does not assess eligibility, compliance, or tax outcomes, those determinations require legal and tax counsel. The presence of 2 OZ-designated tracts within a 9-tract neighborhood is a structural fact worth noting for readers who track that overlay, particularly given that several of the lower-scoring tracts carry income profiles and current values that are broadly consistent with the distressed-community criteria that historically underlie OZ designations.

The combination of predominantly Relatively Low climate ratings and 2 OZ-designated tracts creates an overlay picture that is relatively benign on the climate dimension while flagging economic distress on the opportunity-zone dimension. That combination is not unusual for urban neighborhoods with wide internal score dispersion, but it does reinforce the tract-level framing: climate and OZ conditions are not uniform across Springfield, and aggregating them to a neighborhood level obscures meaningful variation.

Exhibit 1
Five-year forecast for the top-scoring tracts001100+37.0%001200+37.0%001401+37.0%002802+32.4%000101+37.0%000102+37.0%002901+37.0%000200+18.1%001600+37.0%
Exhibit 2
Observed and forecast median value, leading tract 12031001100$138K$288K$438K$587K$737K201420242029ObservedForecast

The forward view

The Verus-AI scores and forecasts presented here are model-derived estimates built on a history window spanning 2014 to 2024 and a forecast window covering 2025 to 2029. The score scale runs from 0 to 100. Confidence intervals are 80% intervals, meaning the model assigns an 80% probability that the realized outcome falls within the stated band; they are not guarantees, and outcomes outside the band are plausible. The forecast chart for tract 12031001100 illustrates how the band widens from $80,633 in the first forecast year to $233,241 at the terminal year, a widening that is structurally expected and should not be read as model instability.

They represent a single-period observation and should not be extrapolated as multi-year trend rates. The figures for tracts 12031000200 (-21.26%) and 12031001600 (-17.55%) are particularly sharp and may reflect data revisions, survey volatility, or genuine economic deterioration; the model does not adjudicate among those explanations, and readers should treat single-period income figures with appropriate caution.

Rent-to-price ratios are annualized figures derived from model inputs and do not account for vacancy, operating expenses, financing costs, or local landlord-tenant regulations. They are gross yield proxies, not net return estimates. The 15.42% figure for tract 12031002802 and the 12.19% figure for tract 12031002901 are high enough to warrant scrutiny of the underlying assumptions before any investment decision is made.

The comparable neighborhoods, Tucker, Brushy Creek, and Bloomingdale, are matched on structural similarity, not on regulatory, tax, or market-liquidity dimensions. Similarity percentages of 99.6%, 98.7%, and 98.6% indicate strong model-level alignment but do not imply that outcomes in those markets will replicate in Springfield. Climate ratings are model-derived designations; the analysis does not assess physical hazard severity beyond the label provided, and no insurance or lending implications are drawn from them.

Exhibit 3
Statistically comparable neighborhoods (similarity %)Tucker+99.6%Brushy Creek+98.7%Bloomingdale+98.6%
Exhibit 4
Five-year forecast dispersion across Springfield's scored tracts025810+18.1%+27.5%+37.0%Five-year forecast appreciation

Questions

What is the overall Verus-AI score for Springfield in Jacksonville?
The mean Verus-AI score across Springfield's 9 scored tracts is 47.44, with a median of 46.0 on a 100-point scale. Individual tract scores range from 35 to 67, reflecting wide internal dispersion rather than a uniform neighborhood profile.
Which Springfield tract has the highest score and what does its forecast look like?
Tract 12031001100 carries the highest Verus-AI score among the 9 tracts shown in the ranked table, at 67 (grade C+), with a current value of $414,195 and a 5-year forecast of 37.01%, implying a CAGR of 6.5%. The Verus-AI model projects a terminal value of $567,483 by 2029, with an 80% confidence interval of $462,722 to $695,963.
What are the lowest-scoring tracts in Springfield and what risks do they carry?
The two lowest-scoring tracts among the 9 tracts shown in the ranked table are 12031001600 (score 35, grade F, risk High) and 12031000200 (score 36, grade F, risk Moderate). Both recorded sharp income declines in the most recent annual period: -17.55% and -21.26%, respectively, the two steepest declines among all 9 scored Springfield tracts.
How does Springfield compare to its model-identified comparable neighborhoods?
The three comparable neighborhoods, Tucker (Atlanta, score 76), Brushy Creek (Austin, score 72), and Bloomingdale (Tampa, score 70), all score above Springfield's neighborhood mean of 47.44 and above the 90th-percentile threshold of 61.4 among the 9 ranked tracts. Only Springfield's leading tract, 12031001100 at a score of 67, approaches the comparables' range; the remaining 8 tracts score below all three comparables.
Are any Springfield tracts designated as Opportunity Zones?
Two of Springfield's 9 tracts carry Opportunity Zone designations. The analysis notes this as a structural overlay but does not assess tax eligibility, compliance, or investment outcomes, which require separate legal and tax counsel.
What are the climate risk ratings for Springfield tracts?
All 9 of Springfield's scored tracts have received climate ratings: 7 carry a Relatively Low rating, 1 carries a Very Low rating, and 1 carries a Relatively Moderate rating. These are model-derived designations; the analysis does not extend them to insurance or lending implications.

Methodology

Forecasts are produced by the Verus-AI model from tract-level Census demographic, employment, and market inputs. The five-year figure is a cumulative point forecast for 2025-2029; confidence bands reflect in-sample model uncertainty only and do not capture macroeconomic shocks, policy changes, or idiosyncratic events. Gross rent yield is derived from ACS tract-level median gross rent; tracts with suppressed or sentinel ACS rent values are shown as n/a. Rankings reflect the model's point estimates (model data as of 2026-05-10) and are not investment advice. Tracts retired in the post-2020 Census geometry are excluded where coverage is insufficient. Five-year forecast appreciation is capped by the model at +37.01%, displayed as +37.0%; a tract at that ceiling carries the model's maximum, and its true expectation may be higher.