Time-Varying Lasso
“Time-Varying Lasso” (Härdle, Wang, Zboňáková, 2016):
1️⃣ High-dimensional financial data require adaptive shrinkage methods to track evolving systemic risks.
2️⃣ The Lasso (Tibshirani, 1996) selects and shrinks coefficients through an ( L_1 ) penalty ( \lambda ).
3️⃣ In dynamic settings, the penalty parameter ( \lambda_t ) itself varies over time.
4️⃣ Its evolution reflects changes in volatility, multicollinearity, and the active set size ( q_t ).
5️⃣ Simulation studies show λ responds systematically to shocks in σ², β-shifts, q-changes, and κ(X′X).
6️⃣ Real-data application on 100 US financial firms links λ-dynamics to the Financial Risk Meter (FRM).
7️⃣ λ co-moves with the VIX and network risk measures, capturing market stress episodes.
8️⃣ In quantile Lasso, λ also depends on the check-function asymmetry (τ) and conditional tail behavior.
9️⃣ The BIC-based moving-window estimation reveals λ as a real-time systemic-risk indicator.
🔟 Conclusion: Time-varying penalization unifies shrinkage, forecasting, and network-risk monitoring
Wolfgang Karl HÄRDLE attained his Dr. rer. nat. in Mathematics at Universität Heidelberg in 1982 and in 1988 his habilitation at Universität Bonn. He is Ladislaus von Bortkiewicz Professor of Statistics at Humboldt-Universität zu Berlin and the director of the Sino German Graduate School (洪堡大学 + 厦门大学) IRTG1792 on “High dimensional non stationary time series analysis”. He directs IDA Institute for Digital Assets,
University of Economic Studies, Bucharest, RO. His research focuses on data analytics, dimension reduction and quantitative finance. He has published over 30 books and more than 300 papers in top statistical, econometrics and finance journals. He is highly ranked and cited on Google Scholar, REPEC and SSRN. He has professional experience in financial engineering, S.M.A.R.T. (Specific, Measurable, Achievable, Relevant, Timely) data analytics, machine learning and cryptocurrency markets. He has created the www.quantlet.com platform, a cryptocurrency index, CRIX www.royalton-crix.com He is 玉山学者 (Yushan Scholar), web page hu.berlin/wkh