Italian SME M&A Post-Acquisition Performance Predictor
An AI-powered dissertation prototype with two heads: a ranking model scores a target firm against candidate acquirers across ROA, TURN, and ROS, while a screening model shortlists the matches most likely to land in the top quartile of realized post-acquisition performance.
Executive SummaryRead the full dissertation summary as a PDF.
Below, the full AI training pipeline is shown. Learn more about the training and evaluation approach by clicking on each stage.
Input layer
↓
↓
↓
Processing layer
↓
↓
↓
↓
Model layer
↓
↓
↓
↓
↓
↓
Output layer
↓
↓
↓
⚠
Estimation Sample
Base is the headline sample. Observed is a robustness specification that relaxes the materiality criterion and keeps every deal with an observed post-deal outcome.
Pre-fill the form with a real precedent deal from the database
Example deal loaded ·
scenario
Actual realized scores:
ROA: ·
TURN: ·
ROS:
Click Run Match to see the model's prediction for this deal
Target Company
▾+ 19 more fields
Match Results
Scenario:
High ≥ 70 Medium 45–70 Low < 45Scale: 0–100
ROA Score
Predicted change in Return on Assets: profitability of deployed capital
TURN Score
Predicted change in Asset Turnover: how efficiently assets generate revenue
ROS Score
Predicted change in Return on Sales: margin efficiency after the deal
Shortlist signal
Screening head's probability the match lands in the top quartile of realized performance (base rate ≈ 25%)
Rank
Acquirer
ROA Score
TURN Score
ROS Score
Overall Score ⓘ
Shortlist ⓘ
Performance Radar
Tick acquirers into the chart. By default, only the winner is shown. The overall score equals the surface a candidate covers here: balanced profiles cover more ground than lopsided ones.