Journal of Monetary Economics · 2025
Policy transition risk, carbon premiums,
and asset prices
A two-sector macro-finance model of climate, policy uncertainty, and financial markets
Christoph Hambel (Tilburg University & Netspar) ·
Frederick van der Ploeg (University of Oxford & UvA & CEPR)
The question
How does uncertainty about future climate policy shape asset prices and the green transition?
Climate policy is inherently uncertain — carbon pricing may be introduced, reversed, or tightened depending on elections and political shifts. This paper asks how financial markets price this policy transition risk, and whether it accelerates the green transition by generating a carbon premium (higher expected returns on brown than on green assets).
Three main contributions
①
Carbon premiums from policy risk
Policy transition risk generates positive carbon premiums — especially when the economy is carbon-intensive and close to the temperature cap. This endogenously accelerates the green transition.
②
Asset price jumps at policy tips
When climate policy tips to carbon pricing, green share prices rise sharply and brown prices fall. Policy shocks cause much larger price reactions than climate or technology tipping events.
③
Precautionary savings and rates
Transition risk increases precautionary saving and depresses the risk-free rate, especially when temperature is near its cap and fossil fuel phase-out becomes imminent.
The model — key equations
Two-sector macro-finance model with policy tipping
Sectoral production (eq. 2.1) & energy composite (eq. 2.2)
\[Y_n = A_n K_n^{1-\eta_n} E_n^{\eta_n} \Lambda_n(T), \qquad E_n = \bigl(\kappa_{1,n} G_n^{\rho_n} + \kappa_{2,n} F_n^{\rho_n}\bigr)^{1/\rho_n}\]
Sector 1 (green), sector 2 (brown). Damage function \(\Lambda_n(T)\) reduces TFP with temperature. \(E_n\) is a CES composite of renewables \(G_n\) and fossil fuel \(F_n\); elasticity of substitution \(\zeta_n = 1/(1-\rho_n)\). The brown sector is significantly more fossil-intensive.
Capital dynamics — green & brown (eq. 2.3)
\[dK_1 = \Bigl(I_1 - \tfrac{1}{2}\varphi_1\tfrac{I_1^2}{K_1} + R - \tfrac{1}{2}\kappa\tfrac{R^2}{K_1} - \delta_1^k K_1\Bigr)dt + K_1\sigma_1\,dW_1 - K_{1-}\ell\,dN\]
\[dK_2 = \Bigl(I_2 - \tfrac{1}{2}\varphi_2\tfrac{I_2^2}{K_2} - R - \delta_2^k K_2\Bigr)dt + K_2\sigma_2\bigl(\rho_{12}\,dW_1 + \sqrt{1-\rho_{12}^2}\,dW_2\bigr) - K_{2-}\ell\,dN\]
\(R\): capital reallocation from brown to green, with quadratic intrasectoral adjustment cost \(\kappa\). Investment costs \(\varphi_n\). \(N\): macroeconomic disaster Poisson process (rate \(\lambda\), loss \(\ell\)) hitting both sectors equally. \(\rho_{12}\): diffusive correlation between sectors.
Temperature dynamics (eq. 2.5)
\[dT = \vartheta\,\nu(F_1 + F_2)\,dt + \sigma_T\,dW_3\]
Temperature rises in cumulative net emissions (TCRE \(\vartheta\)). Stochastic term \(\sigma_T\,dW_3\) captures climate uncertainty. Emission intensity \(\nu\) declines with economic growth. In the CAP state a 2°C cap is enforced; breach triggers an immediate fossil fuel ban.
Epstein-Zin recursive utility (eq. 2.6)
\[J = \sup_{F_n,G_n,I_n,R}\mathbb{E}_t\!\int_t^\infty f\!\bigl(C_s,\,J(s,K_{1s},K_{2s},T_s,X^p_s)\bigr)\,ds, \qquad f(C,J) = \delta\theta J\!\left[\frac{C^{1-1/\psi}}{[(1-\gamma)J]^{1/\theta}} - 1\right]\]
Separates risk aversion \(\gamma\) from the elasticity of intertemporal substitution \(\psi\), with \(\theta\equiv(1-\gamma)/(1-1/\psi)\). State variables: \(K_1,K_2,T,X^p\). Collapses to CRRA when \(\gamma=1/\psi\). Calibrated to match a risk-free rate of 0.8%/yr and equity premium of ~6.5%/yr.
Policy tipping — the Markov chain
Stochastic transitions between three policy regimes
Climate policy evolves as a Markov chain with reversible transitions. Financial markets continuously price in the risk of switching between regimes.
State 1
BAU
No carbon pricing. Policy makers ignore climate externalities. Green transition driven only by asset diversification and falling renewable energy costs.
State 2
PIGOU
Modest carbon pricing. Internalises all global warming externalities including recurring disaster and tipping risk. No binding temperature cap imposed.
State 3
CAP
Ambitious carbon pricing. Enforces a hard 2°C temperature cap (Paris Agreement). If cap is breached, fossil fuel use is immediately banned and stranded-asset risk materialises.
Transition intensities calibrated to Moore et al. (2022). Endogenous probabilities: probability of tipping to active policy rises to 75% if temperature exceeds 1.5°C.
Key results — asset pricing
How policy uncertainty is priced in financial markets
Carbon premium rises near the temperature cap
The carbon premium is small when the economy is far from the cap, but increases sharply as temperature approaches 2°C. At that point, the risk of fossil fuel phase-out makes brown assets much riskier than green ones.
Policy shocks dominate climate and tech shocks
A BAU→CAP transition causes an immediate +14% jump in green share prices and −27% drop in brown prices. Climate tipping events cause only 3–5% drops in both assets. Technology breakthroughs have even smaller effects.
Risk-free rate falls with transition risk
Policy uncertainty raises precautionary savings, curbing the risk-free rate. The mean rate stays near 0.8%/yr but the 5% quantile falls sharply in paths where temperatures breach 2°C, as policy makers urgently phase out fossil fuels.
More ambitious taxes under transition risk
When finally implemented, carbon taxes are ~50% higher than under Pigouvian pricing alone — policy makers must compensate for time lost in BAU. Average initial tax: $60/tCO₂ (CAP state) vs. $45/tCO₂ (PIGOU alone).
Simulation results — full model
Energy transition and asset pricing with all extensions
20,000 sample paths simulated to 2100. The full model combines policy transition risk (three states, endogenous probabilities), recurring climate disasters, irreversible climate tipping, and negative emission technology. Key qualitative findings:
The green transition is faster than under pure BAU but slower than under a guaranteed CAP, with wide uncertainty bands. Climate disasters and tipping reduce long-run output, pushing policy makers to implement higher carbon taxes. Once negative emission technology becomes available, net emissions can turn negative and temperature stabilise or fall.
BAU mean temp 2100: 3.9°C
~45% paths: below 2°C cap
Physical risks substantially raise the optimal carbon tax. In the full model, climate disasters roughly double the Pigouvian carbon tax ($45 → $91/tCO₂ in 2025), and adding tipping pushes it to $121/tCO₂. Under the CAP state with transition risk, policy makers must additionally compensate for time lost in BAU, driving taxes even higher when finally implemented.
PIGOU + disasters + tipping: $121/tCO₂
At cap (conditional): $134/tCO₂
The carbon premium is on average smaller than in the core model — society spends less time in the CAP state. It is near zero under BAU and under pure physical risks (disasters and tipping hit both sectors symmetrically). The premium spikes when temperature approaches 2°C and the economy is in the CAP state, where brown assets face acute fossil fuel phase-out risk.
Physical risks alone: small premium
Near temp. cap (CAP state): up to 6%
Policy transitions dominate all other news. A BAU→CAP tip causes +14% green and −27% brown share price jumps. Climate tipping events cause only 3–5% drops in both sectors. Negative emission technology causes a small brown price rise (+3.8%) and green price drop (−1.8%), as the fossil fuel phase-out risk is reduced. Price impacts diminish as the transition nears completion.
BAU→CAP: green +14%, brown −27%
Climate tip: both −3 to −5%
Why it matters
Practical relevance for investors, central banks, and policy makers
Central banks & financial stability
Policy transition risk lowers the risk-free rate through precautionary savings. Central banks conducting climate stress tests should incorporate stochastic policy scenarios — not just fixed NGFS pathways — to capture the full range of transition outcomes.
Green transition speed
Carbon premiums are not merely a financial curiosity — they incentivise firms to reallocate capital from brown to green faster. The model shows that the threat of carbon pricing accelerates decarbonisation even before it is implemented.
Empirical carbon premium puzzle
The model explains why carbon premiums (Bolton & Kacperczyk 2021, 2023) are particularly pronounced since 2015 and rise with temperature: the mechanism is policy risk, not investor preferences. Premiums should be highest where policy reversal risk is greatest.
Carbon pricing strategy
Delayed climate action is costly: when finally implemented, optimal carbon taxes must be 50% higher to compensate for foregone abatement. This provides a quantitative argument for front-loading carbon pricing and locking in policy credibility early.