Full Text Available
Note: Clicking the button above will open the full text document at the original institutional repository in a new window.
Using Artificial Intelligence for Balanced Concession Period Determination in Public-Private Partnership Projects
PPP Infrastructure projects are highly impacted by inflation. A single inflation shock could completely reshape the project’s financial agreements, shift IRR, and eventually force concession periods to be extended or renegotiated. This factor, which often dictates the final concession period, is oft...
Saved in:
| Main Author: | |
|---|---|
| Format: | Thesis |
| Published: |
AUC Knowledge Fountain
2026
|
| Subjects: | |
| Tags: |
No Tags, Be the first to tag this record!
|
| Summary: | PPP Infrastructure projects are highly impacted by inflation. A single inflation shock could completely reshape the project’s financial agreements, shift IRR, and eventually force concession periods to be extended or renegotiated. This factor, which often dictates the final concession period, is often treated in most PPP planning as a fixed parameter or a simplified assumption rather than a dynamic driver, which leads to a critical gap that exposes all partners to severe financial risks and inaccurate concession period determination. To address this blind spot, this research develops a data-driven AI- backed decision support system to determine optimal concession period for PPP projects based on accurate forecasting of future inflation. As such, this research combines a two-stage methodology that forecast inflation in Egypt and fuses it into a decision support system for concession determination. A multivariant LSTM model is rigorously trained, validated and tested using 11 important macroeconomic indicators in Egypt from 2005 till 2023, generating monthly inflation till 2040. This forecast is then embedded into a decision support engine to determine the best concession period that delivers the strongest and optimal balance between the investor’s return and the government’s affordability for a real PPP case study. The results showed how LSTM was able to capture Egypt’s dynamic inflation shocks easily more than any modelling technique as when it was validated and compared with the actual inflation from 2023 till 2025, it was able to capture the downwards trend in 2024 and 2025 with MAE 2.95 and R2 of 77% variance. When these forecasts were fed into the decision support engine, it made the optimal concession period visible, balancing between the investor’s return and government payments which will not be achieved if the inflation was assumed rather than forecasted. The merit of this research is to deal with inflation not as an assumed indicator but a data driven framework that strengthen the financial decision planning of PPP projects in Egypt and bridges the persistent gap providing an advanced inflation forecasting tool combined with a concession period determination engine in a single model. Which will eventually help the government to provide a balanced concession period determination during the PPP planning phase. |
|---|