Finance · Machine learning · Luxembourg

Alessandro Tugnetti

Data Scientist · PhD in Finance

I am a data scientist based in Luxembourg with a research background in finance and economics. I work with complex data, statistical models and machine learning to investigate economic questions, evaluate forecasts and communicate results. I completed my PhD in Finance at the University of Luxembourg in April 2026.

My research focuses on digital-asset markets and long-run economic growth. I combine econometric analysis and machine learning with careful data preparation and out-of-sample evaluation. Alongside research, I contribute to QFI alumni career research and community initiatives, and I am continuing my professional development through WeSTEM+’s Applied Data & AI Academy.

Let’s connect. Get in touch

~ News

[October]
Started WeSTEM+’s Applied Data & AI Academy, with completion expected in April 2027.
[27 April]
Defended my PhD dissertation, Essays in Empirical Finance, at the University of Luxembourg.

Selected Publications

All research

Eur. Financ. Manag.
Pricing dynamics and herding behaviour of NFTs

Gilbert Fridgen, Roman Kräussl, Orestis Papageorgiou, and 1 more author

European Financial Management, 2025

Abs DOI Bib PDF

This paper analyzes the sales of 875,389 art nonfungible tokens (NFTs) on the Ethereum blockchain to identify the key determinants influencing NFT pricing and market dynamics. We find that market liquidity and trade volume are strong predictors of NFT prices. Contrarily, social media activity negatively correlates with prices. Introducing an artist ranking system, our study reveals a “superstar effect”, with a few artists dominating sales, and herding behaviour within the NFT market.
@article{fridgen2025pricing, title = {Pricing dynamics and herding behaviour of NFTs}, author = {Fridgen, Gilbert and Kr{\"a}ussl, Roman and Papageorgiou, Orestis and Tugnetti, Alessandro}, journal = {European Financial Management}, volume = {31}, number = {2}, pages = {670--710}, year = {2025}, doi = {10.1111/eufm.12506}, url = {https://onlinelibrary.wiley.com/doi/full/10.1111/eufm.12506}, dimensions = {true}, }
J. Econ. Surv

Non-fungible tokens (NFTs): A review of pricing determinants, applications and opportunities

Roman Kräussl and Alessandro Tugnetti

Journal of Economic Surveys, 2024

Abs DOI PDF

This paper provides a review of the development of the non‐fungible tokens (NFTs) market, with a particular focus on its pricing determinants, its current applications, and its future opportunities. We investigate the current state of the NFT markets and highlight the perception and expectations of investors toward these products. We summarize and compare the financial and econometric models that have been used in the literature for the pricing of non‐fungible tokens with a special focus on their predictive performance. We design a framework that can help to understand the price formation of NFTs. We further aim to shed light on the value‐creating determinants of NFTs in order to better understand investors’ behavior on the blockchain.

Opfield Machine Learning Demo · Opfield Networks

Half a face. Can the network remember the rest?

Loading the portrait…

(Reconstruction steps: 0/3000)

A Hopfield network is an associative memory inspired by the Ising model. Its neurons interact to reach stable, low-energy states, allowing a partial pattern to recover a stored memory.

This network stores a simple 50 × 50 blue-and-white pixel portrait. Each pixel is one binary neuron. The left half starts blank. Click Retrieve Pattern to recall it from the visible half.

E = −½ ∑_{i ≠ j} W_ij s_i s_j

A small educational demonstration inspired by Morris Huang’s website. The network recalls a learned image; it does not generate a new face. Recall is computed first; 3000 reconstruction steps then display the recovered pixels.