Introduction

Carbon pricing policies aim to reduce emissions but impose transition risks on the economy and financial system. Existing data often lacks firm-level emissions, impeding accurate stress testing. This study utilises a unique dataset of 410,523 Hungarian firms to estimate CO2 emissions and assess the impact of the EU Emissions Trading System II (ETS II). It models direct costs and indirect supply chain contagion under varying carbon prices.

Results

Estimating firm level CO2 emissions with supply chain data

The study estimates emissions for all value-added tax (VAT) paying firms in Hungary based on supply chain purchases of oil and gas. While existing ETS I data covers only 119 firms accounting for 7% of national sales, this new method identifies 185,783 emitting firms responsible for 70% (EUR 179 billion) of sales. This reveals a substantial ‘hidden’ exposure to carbon pricing across all sectors, not just energy and manufacturing.

Estimating firms’ direct output losses from carbon pricing

Direct economic losses occur when carbon costs render firms unprofitable. At the EU ETS II price cap of 45 EUR/t, direct production losses are estimated at 1.3% of total sales. If prices rise to 200 EUR/t (the upper range for 2-degree targets), losses increase to 3.8%. These figures assume firms pass on a portion of costs to customers. Without cost pass-through, losses are lower, indicating the market’s ability to absorb some shocks directly.

Estimating direct financial system losses from carbon pricing

Bank exposure is significant, with loans to affected firms often exceeding banks’ CET1 capital. However, direct financial stability risks appear manageable at moderate prices. At 45 EUR/t, banking system-wide equity losses are estimated at 1.2%. Losses rise to 4.7% at 200 EUR/t. Notably, individual banks vary widely in exposure; at 45 EUR/t, most suffer less than 1% loss, though the most exposed bank faces a 6% equity reduction.

Amplification of transition risks due to supply chain contagion

Indirect risks from supply chain failures significantly amplify direct losses. In an ‘optimistic’ scenario where firms can substitute inputs, gross output losses at 45 EUR/t rise from 1.3% (direct) to 5.3% (total). At 200 EUR/t, total losses reach 12.3%.

In a ‘pessimistic’ scenario where essential inputs cannot be substituted, the model predicts a catastrophic jump in losses at just 30 EUR/t. Here, the failure of firms in the ‘systemic risk core’ triggers cascades, causing output losses to exceed 50% and bank equity losses to reach 43%. This highlights the critical vulnerability of the economy to specific, highly connected nodes in the supply network.

Discussion

Direct economic losses from planned carbon pricing (EU ETS II) appear manageable, provided firms can adapt. However, supply chain contagion acts as a massive amplifier of transition risk. Indirect network effects can increase losses by factors of 4 to 40 depending on input substitutability.

Banks face exposure not only from their immediate borrowers’ emissions but also from the embedded emissions in their borrowers’ supply chains. The findings suggest that policy makers and regulators must monitor systemically relevant firms to prevent cascading defaults. Furthermore, a predictable, gradually increasing carbon price implementation is preferable to sudden price shocks to mitigate these substantial indirect risks.