A Kalman-smoother-based estimator applied to Swedish administrative data reveals a steep, convex negative gradient in the marginal propensity to consume with respect to cash-on-hand, from 0.7 to 0.3 across deciles.
hub
Parisa Foroutan and Salim Lahmiri
12 Pith papers cite this work, alongside 393 external citations. Polarity classification is still indexing.
12
Pith papers citing it
393
external citations · external index
hub tools
representative citing papers
MACROCAST is the first leakage-free time series foundation model for real-time macroeconomic forecasting, trained exclusively on synthetic series and vintage data, outperforming AR(1), Chronos-2, BVAR, and DFM benchmarks on FRED-MD.
ReSGA, a large autoencoder, outperforms prior methods on joint VaR-ES forecasting for US equities and converts the edge into economic gains via a size-enhanced momentum strategy, with gains attributed to data complexity.
Bitcoin price power law is rejected on distributional series and not robust to time-origin shifts, but dominates medium-horizon forecasts against baselines because it avoids committing to specific wave shapes.
GRAFT improves electric load forecasting accuracy by aligning multi-source daily texts with half-hour load series and using cross-attention fusion, outperforming baselines on a new Australian benchmark across hourly to monthly horizons.
Develops a robust two-step two-sample IV estimator and associated overidentification and weak-instrument tests that use only six summary statistics and remain efficient under heteroskedasticity and sample heterogeneity.
Benchmark of 15 time-series architectures on equity portfolios finds no model dominates, with TransEnc-8 at 0.352 rank-1 acceptability and all promoted models showing negative net Sharpe at 20 bps costs under constraints.
NSTM maps European bidding zones into a network via metric graph and neighborhood measure, outperforming independent local models in day-ahead price forecasting across 39 zones.
A Dirichlet-based Bayesian forecasting model predicts the origin-country mix of Airbnb bookings and beats naive baselines for Europe, but not consistently across all regions.
Raw IFS forecasts outperform raw AIFS for wind speed at all horizons, but post-processing with EMOS or QR reduces the gap, leaving IFS ahead mainly at short leads.
A literature survey finds no peer-reviewed Bitcoin price models beat the naive baseline at medium horizons and proposes methodological improvements including walk-forward testing and Diebold-Mariano tests.
citing papers explorer
-
Identifying the MPC-Liquidity Gradient in High-Quality Data
A Kalman-smoother-based estimator applied to Swedish administrative data reveals a steep, convex negative gradient in the marginal propensity to consume with respect to cash-on-hand, from 0.7 to 0.3 across deciles.
-
MACROCAST: A Vintage-Consistent Time Series Foundation Model for Real-Time Macroeconomic Forecasting
MACROCAST is the first leakage-free time series foundation model for real-time macroeconomic forecasting, trained exclusively on synthetic series and vintage data, outperforming AR(1), Chronos-2, BVAR, and DFM benchmarks on FRED-MD.
-
ReSGA: A Large Tail Risk Model for Learning Value-at-Risk and Expected Shortfall
ReSGA, a large autoencoder, outperforms prior methods on joint VaR-ES forecasting for US equities and converts the edge into economic gains via a size-enhanced momentum strategy, with gains attributed to data complexity.
-
Bitcoin's Power Law: Weak Structure, Strong Forecasts
Bitcoin price power law is rejected on distributional series and not robust to time-origin shifts, but dominates medium-horizon forecasts against baselines because it avoids committing to specific wave shapes.
-
GRAFT: Grid-Aware Load Forecasting with Multi-Source Textual Alignment and Fusion
GRAFT improves electric load forecasting accuracy by aligning multi-source daily texts with half-hour load series and using cross-attention fusion, outperforming baselines on a new Australian benchmark across hourly to monthly horizons.
-
Two-Sample IV: Efficient Two-Step Estimation and Tests for Overidentification and Weak-Instruments
Develops a robust two-step two-sample IV estimator and associated overidentification and weak-instrument tests that use only six summary statistics and remain efficient under heteroskedasticity and sample heterogeneity.
-
Benchmarking Deep Time Series Models for Equity Portfolios
Benchmark of 15 time-series architectures on equity portfolios finds no model dominates, with TransEnc-8 at 0.352 rank-1 acceptability and all promoted models showing negative net Sharpe at 20 bps costs under constraints.
-
Networked Spatial Effects in European Electricity Price Forecasting
NSTM maps European bidding zones into a network via metric graph and neighborhood measure, outperforming independent local models in day-ahead price forecasting across 39 zones.
-
Forecasting the Evolving Composition of Inbound Tourism Demand: A Bayesian Compositional Time Series Approach Using Platform Booking Data
A Dirichlet-based Bayesian forecasting model predicts the origin-country mix of Airbnb bookings and beats naive baselines for Europe, but not consistently across all regions.
-
AI and physics-based weather forecasting: A comparative study
Raw IFS forecasts outperform raw AIFS for wind speed at all horizons, but post-processing with EMOS or QR reduces the gap, leaving IFS ahead mainly at short leads.
-
Bitcoin Price Prediction: Peer-Reviewed Evidence and Social Media Discourse
A literature survey finds no peer-reviewed Bitcoin price models beat the naive baseline at medium horizons and proposes methodological improvements including walk-forward testing and Diebold-Mariano tests.
- When Should Forecasting Models Be Re-Specified? A Cost-Sensitive Trigger for Adaptive Model-Form Updating