How the Trajectory Index works
A plain-English explanation of every layer of the scoring model, from raw government data to a single 0 to 100 district score.
The formula, in plain English
The Trajectory Index is a weighted average of six domain scores, each of which is itself a weighted average of underlying signals. Every raw signal is first normalised nationally on a 0 to 100 scale, where 50 equals the national median. A score of 50 does not mean average performance in absolute terms. It means exactly average relative to all 2,391 scored districts simultaneously.
Normalisation happens fresh each month when new data arrives. A district's score can change even without local events if the national distribution shifts. This is intentional: the score answers the question “is this district an above-average opportunity right now, compared to everywhere else?” rather than “has this district improved since last year?”
A final gravity blend applies a 3% pull toward neighbouring district scores, producing trajectory_score_exp, the variant used on the map. This smooths out isolated statistical anomalies and reflects the real-world observation that areas rarely surge in isolation from their neighbours.
The six domains
Weights reflect how predictive each domain has been in historical validation. Demand Pressure carries the most weight because it is the most robust leading indicator of price growth.
Price Opportunity
21%Value gap · price CAGR · rent CAGR · price-to-income
Measures how far below its fundamental ceiling a district is currently priced. A high score means the market price has not yet caught up with the area's supply-demand dynamics. Key inputs are how much cheaper the district is relative to the county median (value gap), the annualised rate of price appreciation (price CAGR), rent growth momentum (rent CAGR), and the local ratio of house prices to incomes. Districts with wide value gaps and accelerating rents score highest.
Demand Pressure
27%Population growth · sales volume · rail usage growth · net migration · pop. acceleration · score momentum · neighbour score
The single highest-weighted domain, capturing the breadth and depth of people wanting to live in an area. Strong demand pressures prices upward regardless of current fundamentals. Signals span demographic momentum (population growth and acceleration, net internal migration), market activity (transaction volumes), connectivity demand (rail entry/exit growth), score persistence over three years, and whether neighbouring districts are also rising. Demand Pressure is the strongest leading indicator in the model.
Transformation
19%Business formation · IMD improvement · vacant trend · claimant rate trend · planning activity · tenure shift · reno rate
Captures structural change in progress: the early signals of an area moving up the value curve before prices reflect it. Rising business formation indicates improving economic confidence. Falling deprivation ranks (IMD improvement) and declining claimant rates signal an improving labour market. Declining vacant properties and increasing planning applications for cafés, restaurants and mixed-use suggest incoming investment. Tenure shift (rising homeownership share) and renovation activity complete the picture.
Economic Baseline
14%Household income · claimant rate · EPC average rating · housing supply
The foundation layer: is the local economy healthy enough to sustain price appreciation? Districts need a floor of income depth, employment strength, and housing quality to attract sustained demand. Income and claimant rate measure the economic capacity of residents. EPC average rating is a proxy for housing stock quality. Housing supply captures the ratio of new-build completions to housing need. High baseline scores give other signals more weight in a healthy macro environment.
Safety
10%Crime rate · crime trend
Safety consistently features in tenant and buyer decision-making, and it has a measurable drag effect on price growth. This domain uses two signals: the absolute crime rate (crimes per 1,000 population, sourced from Police.uk monthly data) and the three-year trend. An area with high crime but a strong downward trend can still score well here; conversely, a low-crime area where crime is rising gets penalised. Safety receives a deliberate 10% weight: important but not determinative on its own.
Risk Ceiling
9%Flood risk · isolation · new-build share · NSIP burden · new-build pipeline
Acts as a cap on otherwise strong scores where structural risks exist. High flood exposure (Flood Zone 2+ coverage from Environment Agency data) can constrain insurance and mortgage availability. An elevated isolation score reflects poor transport connectivity. A high share of new-build sales or a large incoming development pipeline signals potential future oversupply. NSIP burden captures nationally significant infrastructure projects with negative amenity impact in proximity. Districts with extreme risk readings are pulled back from top-decile placement regardless of other domain scores.
The cycle indicator: why market regime matters
Property markets alternate between two broad regimes. In a momentum market, typically a period of rising national house prices, well-priced areas with strong demand signals tend to compound quickly. Capital chases quality. In a catch-up market, flat or softening national prices, previously overlooked districts with high value gaps tend to outperform, as buyers widen their search and yield-hunters seek underpriced stock.
The model encodes this as a single national variable, cycle_momentum, a continuous 0 to 1 index derived from national price trend, transaction volume velocity, and mortgage approval data. A reading near 0 signals a catch-up environment; near 1 signals a momentum environment.
Domain weights are modulated by this reading at each monthly recompute. In a catch-up environment, Price Opportunity and Transformation receive marginally more weight. In momentum, Demand Pressure and Economic Baselinereceive more weight. The adjustment is intentionally modest: the fundamentals drive the score, the cycle indicator fine-tunes it.
The context score: same signal, different meaning
A new café opening in Middlesbrough carries a different signal than a new café opening in Mayfair. Raw signals need to be interpreted in the context of an area's existing character: its urbanity, density, connectivity, and current price level.
ctx_score is an internal composite (0 to 1) that characterises each district on a spectrum from sparse rural to dense urban core. It is derived from population density, rail connectivity, employment density, and ONS urban classification. High context scores are assigned to well-connected urban districts; low scores to rural and semi-rural areas.
The context score modulates how signals within Transformation and Demand Pressure are weighted. Business formation in a dense urban district with high connectivity is weighted more heavily than the same signal in an isolated rural area, where it is more likely to be noise. Without this adjustment, the model would systematically over-score sparse rural districts that occasionally show high percentage changes from a small base.
Validation
We test the model by asking whether high-scoring districts in a given year subsequently saw stronger price growth than low-scoring districts. All tests use out of sample data: the scoring engine only sees data that existed at the prediction date. The primary metric is Spearman rank correlation (ρ), which is robust to outliers and non-linear price distributions.
| Window | Metric | ρ | n |
|---|---|---|---|
| 2010 to 2015 | HPSSA 5yr | +0.441 | 2,291 |
| 2013 to 2015 | HPSSA 2yr | +0.557 | 2,291 |
| 2013 to 2018 | HPSSA 5yr | +0.554 | 2,291 |
| 2022 to 2024 | Land Registry | +0.188 | 2,292 |
| 2022 to 2024 | HPSSA | +0.236 | 2,292 |
top vs bottom decile
Top-decile districts appreciated 41% faster cumulatively than bottom-decile districts over the same period (2013 to 2018).
better than chance
Top-10% scoring districts had a 50.8% hit rate for being top-10% price performers over the next 2 years. Random baseline: 10%.
of weak areas declined
62% of bottom-decile districts saw price declines in 2022 to 2024, validating the model as a risk filter, not just a growth finder.
Correlation data uses ONS HPSSA annual median prices and Land Registry Price Paid Data. For full methodology details see the Model Validation page. n = 2,291 to 2,518 districts depending on window (districts with price CAGR above 50% excluded as outliers).
What the score is not
Not a property valuation
The score operates at postcode district level (e.g. E17, not E17 4BP). It captures area-level forces, not property-level factors like condition, lease length, service charges, or specific micro-location. Two properties 300 metres apart in the same district will have the same Trajectory Score but could differ materially in investment merit. Always combine district signals with a property-level assessment.
Not financial advice
Praesago is an intelligence tool, not a financial adviser. Scores and reports are data summaries derived from publicly available government datasets. They do not account for your personal financial situation, tax position, risk tolerance, or investment timeline. Always take independent financial advice before committing capital to property investment.
Probabilistic, not deterministic
A high Trajectory Score does not guarantee price appreciation. It means that, based on the current balance of publicly observable signals, this district exhibits the characteristics that have historically preceded above-average performance. Individual district outcomes are subject to local shocks, national macro events, policy changes, and factors not captured in the current signal set. The model gives you an edge at scale; it does not eliminate risk.
Not a timing tool
The score says where to look, not when to buy. National monetary policy, credit conditions, and macro sentiment dictate timing. A high-scoring district in a rising-rate environment may still underperform a lower-scoring district in a falling-rate environment over a short window. The index is calibrated to 2 to 5 year holding periods, not quarterly moves.
Data freshness and update cadence
The scoring engine recomputes all 2,391 district scores each month as new data is ingested. Different signals update on different schedules: Land Registry transactions arrive monthly (with a 3 to 6 month lag), police crime data arrives monthly, ONS earnings and population estimates update annually. The engine always uses the most recent available vintage of each signal.
Each district score carries a confidence rating (1 to 6) reflecting data completeness. Rural districts with thin transaction volumes or sparse police data carry lower confidence scores. The confidence rating is shown on every district report and data panel so you can weight the score accordingly.
See it in action
Explore live scores for 2,391 districts on the map, or run a full intelligence report on any UK postcode.