Financial Service - ReLIFE-Project-EU/relife-wiki GitHub Wiki

<<<<<<< HEAD

title: "{Financial Service}"

=======

Financial Service

68a1c8eaf7115d9e292915662c02136c8d074971

Introduction

The Financial Service is designed to support a wide range of stakeholders – from professionals such as investors, financing institutions, and ESCOs to non-professionals like homeowners and building managers – in assessing the investment potential of deep renovation projects.

The service leverages both technical and financial data (e.g., building characteristics, market conditions) to evaluate multiple renovation and financing scenarios. Its analytical modules calculate key financial indicators (CAPEX, OPEX, PP, DPP, NPV, IRR, ROI, ARV), simulate different funding options, forecast future market parameters, and assess investment risk through Monte Carlo simulations.

Due to its broad, general-purpose application, the Financial Service serves all three ReLIFE open-access tools.

Go To Top

Sections

Funding Options

A core functionality of the Financial Service is the ability to select among different funding options and simulate the corresponding financial indicators. Three main funding options are supported and can be applied individually or in combination:

Funding Option Description
Loan A borrowed amount repaid over a specified period with interest. The full amount is provided upfront.
Subsidy A government-provided grant that does not require repayment, typically covering a percentage of the total CAPEX.
On-Bill An upfront amount enabling renovation, repaid through the monthly energy bill as a percentage of the energy savings achieved.

Combining funding options allows users to build comprehensive, customised financing packages tailored to their financial constraints.

Financial Indicators

A set of financial indicators is calculated to assess and compare renovation scenarios. These indicators serve both professional and non-professional users:

Indicator Description
CAPEX (Capital Expenditure) Total upfront investment required to implement the renovation measures
OPEX (Operating Expenses) Annual costs for energy consumption and maintenance after renovation
PP (Payback Period) Time required to recover the initial investment through energy savings
DPP (Discounted Payback Period) Payback Period adjusted for the time value of money
NPV (Net Present Value) Present value of all future cash flows from the renovation, discounted at a predefined rate
IRR (Internal Rate of Return) Discount rate that makes the NPV of the investment equal to zero; a higher IRR indicates a more attractive project
ROI (Return On Investment) Ratio of net profit to the total investment cost
ARV (After Renovation Value) Estimated market value of the property following renovation

After Renovation Value (ARV)

The ARV represents the estimated market price of a property following renovation. Because this value is primarily determined by market dynamics, a Machine Learning (ML)-based approach using a Random Forest regression model is employed for its estimation.

Data and analysis:

  • Initial analysis was conducted on a dataset of ~170,000 properties listed for sale in Athens, Greece (2024–2025)
  • The analysis confirmed a strong direct correlation between EPC rating and property price per mΒ², consistent across construction decades and geographic regions
  • Data cleaning was performed using quantile filtering (25th–75th percentiles) and the DBSCAN clustering algorithm to remove outliers (e.g., coastal properties, auction entries, land-with-building listings)

Model details:

  • Algorithm: Random Forest (100 estimators)
  • Features: general location, sub-location, EPC rating, building type, recent renovation status, number of floors, floor number, floor area, building age
  • Training/testing split: 70% / 30% with stratification across location, EPC rating, construction decade, and floor area
  • Performance: RMSE = 281.34 €/mΒ², RΒ² = 0.90

Key finding: EPC rating is a reliable predictor of ARV. The greatest ARV increases occur when buildings transition from very low to very high EPC ratings. The approach will be extended to additional EU countries as data becomes available, with ongoing work to harmonise EPC rating scales under a common framework (ISO 52016-1:2017).

Forecasting Module

The Forecasting Module predicts future market parameters over the lifetime of a renovation project, enabling more accurate calculation of financial indicators and investment risk assessment. Three key parameters are forecasted, each with 80% prediction intervals to quantify uncertainty:

Energy Prices

Fuel prices (natural gas, heating oil):

  • Data sources: yfinance commodity futures prices (tickers: HO, NG) + EU retail prices per country
  • Model: RANSAC Robust Regression – handles outliers caused by geopolitical events and economic shocks
  • Output: Point forecasts and 80% confidence intervals per country, saved as CSV files

Electricity prices:

  • Data source: EU electricity retail prices per country
  • Model: RANSAC fitted on the electricity price trend and transformed month feature
  • Output: Point forecasts and 80% confidence intervals per country, saved as CSV files

Inflation Rates

  • Model: Unobserved Components Model (UCM) – captures long-term trend, stochastic cycles, and damped cyclical patterns
  • Historical data: 1960–2023 (World Bank)
  • Output: Point forecasts and 80% confidence intervals per country as CSV files
  • Note: Projects with longer lifetimes produce wider prediction intervals, reflecting the increasing uncertainty of long-term inflation forecasting

Interest Rates (Euribor)

  • Model: Classical time series decomposition – separates and forecasts the linear trend and cyclical component of the 1-year Euribor
  • Data: 1-year Euribor rates from the European Central Bank (ECB), December 1994 – December 2024
  • Output: Point forecasts and 80% confidence intervals as CSV files

Risk Assessment Framework

The Risk Assessment Framework evaluates the uncertainty surrounding investment outcomes, giving stakeholders a comprehensive view of potential financial performance under varying market conditions.

Methodology:

  1. Integration with the Forecasting Module – Predictions for energy prices, inflation, and interest rates are used as probabilistic inputs (assumed Normal distributions).
  2. Scenario construction – A Monte Carlo simulation approach is used (5,000–10,000 scenarios), sampling randomly from the forecast distributions for each year and each market parameter.
  3. Calculation of financial indicators – For each simulated scenario, annual cash flows are constructed (including energy savings, maintenance costs, and financing payments), and all financial indicators are recalculated.
  4. Results – Outputs include:
    • Histogram plots showing the distribution of each indicator's possible values and their frequency of occurrence
    • Summary tables detailing the range of outcomes and their probability of occurring

This framework allows users to evaluate not only expected investment outcomes but also the likelihood of different scenarios, supporting informed and risk-aware decision-making. Funding options can be applied individually or in combination within the simulations to assess how different financing strategies affect both returns and risk.

ReLIFE Financial Service: component and data flow diagram

Detailed view of the components of the methodological framework of ReLIFE Financial Service and the corresponding data flow from the user to the final results.

Data Requirements

Data required by the Financial Service is divided into two categories:

Technical Data (sourced from user input and the Forecasting Service):

  • Building characteristics: location, floor area, building type, year of construction, EPC rating before renovation
  • Energy efficiency data: energy consumption before and after renovation, EPC rating after renovation, annual energy savings (kWh/year)

Financial Data (from literature, market research, and user input):

  • Renovation costs: CAPEX per measure and country (from market data)
  • Annual maintenance costs
  • Market parameters: energy prices, inflation rates, Euribor interest rates (forecasted by the Forecasting Module)
  • Financing inputs: loan amount, loan term, subsidy percentage, on-bill repayment rate

Go To Top

How To Cite

Please refer to the How To Cite section on the Welcome Page.

Go To Top

Authors And Reviewers

Authored by the Decision Support Systems Laboratory (DSS Lab) at the National Technical University of Athens (NTUA).

NTUA Team

  • Evangelos Spiliotis
  • Daniela Stoian
  • Dimitrios Avgoloupis
  • Efstathios Stamatopoulos
  • Sokratis Divolis

Reviewers:

Go To Top

License

The ReLIFE open source projects are licensed under the EUPL-1.2 license. Please check each repository for project-specific details.

Go To Top

Acknowledgement

This work is carried out within the ReLIFE project and is co-funded by the European Union (CINEA) under Grant Agreement No. 101167067.

Co-funded by the European Union. Views and opinions expressed are however those of the author(s) only and do not necessarily reflect those of the European Union or CINEA. Neither the European Union nor CINEA can be held responsible for them.

Go To Top