Econometric assessment of seasonal volatility and predictive uncertainty in Albania’s hydropower-dependent electricity market

The Albanian energy sector is highly reliant on hydropower, making the national energy balance vulnerable to seasonal variability and hydrological shocks. This study analyzes quarterly electricity production data for 2012Q1–2025Q3 using Seasonal–Trend decomposition via Loess (STL) and Seasonal Autoregressive Integrated Moving Average (SARIMA) modelling. STL is used to isolate trend and seasonal strength, while a SARIMA(0,1,2)(0,1,1)4 model is estimated to generate production forecasts for the 2025Q4–2030Q4 horizon (Q = calendar quarter). Results confirm a dominant seasonal component (FS = 0.635) and a recurring seasonal trough in Q3. A post-2022 increase in volatility is reflected in widening 95% prediction intervals, indicating that while the seasonal timing of stress is predictable, the magnitude of future outcomes becomes increasingly uncertain over the medium term. To operationalize this risk, the study implements a hydrological drought stress test on Albania’s quarterly net domestic production (Yp), using the SARIMA baseline as the reference path. The sensitivity results quantify sizable, seasonally concentrated supply gaps under proportional production shocks (− 10% and − 25%), reinforcing the structural nature of the recurring third-quarter (Q3) deficit and its implications for adequacy and import coverage. The findings inform resilience planning by quantifying the scale and timing of seasonal shortfalls and by motivating diversification and flexibility options that are seasonally complementary to hydropower. In particular, solar PV is discussed as a potential counter-cyclical contributor conditional on scale-up, grid hosting capacity, and balancing resources. The paper concludes that hydropower-centric strategies alone are insufficient for medium-term stability and highlights the role of diversified non-hydro resources, grid-efficiency improvements, and storage in hedging forecasted production troughs and reducing exposure to external balancing needs.