About | Property Analytics London

Our Methodology

Property Analytics uses a cross-sectional factor-based approach to analyse the London real estate market, similar to those widely used in equity market analysis. This methodology allows us to decompose property returns into specific factors that drive market prices.

Factor Model Development

Our model was developed through rigorous statistical analysis of London property transaction data from 1995. We employed multiple regression techniques to identify the key factors that consistently explain variations in property prices across different market cycles.

A Cross-sectional model

Decomposing returns into factors can be done in mainly three ways: a purely statistical approach, time-series regression, and cross-sectional regression. The first one is hard to interpret, and the second one requires knowing the factor returns a priori. The third one is the most powerful and therefore widely used for stocks. All it requires is knowing the characteristics of each property and regressing the prices on those to derive the factor returns. Fortunately, this is very easy for properties where we can use, for example, the number of rooms, but it's somewhat harder for stocks.

Factor Selection Process

We started with a comprehensive set of potential factors and systematically eliminated those with low explanatory power or high correlation with other factors. The final model includes 7 factors that collectively provide a robust framework for understanding London property prices:

Baseline market return
Total floor area
New build premium
Non-linear floor area
Number of habitable rooms
Energy efficiency
Construction period

Model Validation

We validate our model using rigorous statistical tests, like ratio between residuals and factor returns, low factor cross-correlations, and low autocorrelations. We also do out-of-sample testing, comparing its predictions against actual market transactions.

Data Sources

Our analysis is based on a comprehensive dataset of London property transactions and related information from multiple authoritative sources:

Primary Data Sources

HM Land Registry Price Paid Data
Energy Performance Certificate (EPC) data
Office for National Statistics (ONS) housing data
London Datastore housing and planning data
UK House Price Index

Supplementary Data

Bank of England interest rate data
UK Census demographic data
Transport for London accessibility data
Local planning authority records
Historical economic indicators

Data Processing

The raw data underwent extensive cleaning, normalization, and integration to create a unified dataset suitable for factor analysis. We employed various statistical techniques to handle missing values, outliers, and data inconsistencies.

Data Limitations

While our dataset is comprehensive, it has certain limitations:

Not all property transactions are captured in public records
Some property characteristics may have changed over time without being recorded
Energy performance data is not available for all properties, especially in earlier years
The model may not fully capture unique features of individual properties

About the Team

Property Analytics London was developed by a team of data scientists, economists, and real estate professionals passionate about bringing data-driven insights to the London property market.

Our Expertise

Our team combines expertise in:

Quantitative finance and factor modeling
Real estate economics and valuation
Data science and statistical analysis
Machine learning and predictive modeling
London property market dynamics

Our Mission

We believe that better data leads to better decisions. Our mission is to democratize access to sophisticated property market analysis, helping homebuyers, investors, and property professionals make more informed decisions in the London real estate market.

About Our Analysis