we-solve (beta)

Econ-485 · economics · Economics

Climate/ESG asset pricing and real effects

LLM status Open unverified

Statement

Atlas ID: F8; original number: 98; field: Finance. Classification: Editorial research formulation (theoretical/empirical); not a certified unsolved theorem.

Definitions and assumptions

Firms $i$ choose investment $I_i$, abatement $a_i$ and financing within technologies producing cash flows and physical emissions $E_i(I_i,a_i)$. Investors maximize expected utility over budget-feasible portfolios, possibly valuing environmental characteristics directly. Model $\theta$ specifies climate-state risks, tastes, disclosure and rating measurement, ownership and a market-clearing equilibrium selection. For ordinary financial claims, a pricing kernel $m_{t+1}$ satisfies $P_{it}=\mathbb E_t[m_{t+1}X_{i,t+1}]$, with $X_{i,t+1}$ the next-period payoff, when its assumptions apply; investor-specific nonpecuniary benefits or participation constraints require the corresponding modified optimality condition. Assume finite discounted cash flows and expected emissions. Fix a finite horizon $H$ and define $E_{i,\theta}(z)=\sum_{t=0}^{H}E_{it,\theta}(z)$ as total physical emissions in tonnes of carbon-dioxide equivalent under intervention $z$. Investment $I_{i,\theta}(z)$ uses a common monetary numeraire and the same horizon.

Formal research question

For an externally assigned financing-demand intervention $z\in\{0,1\}$, identify effects on financing costs, investment and emissions: \[ \Delta_E=\sum_i\mathbb E_\theta[E_{i,\theta}(1)-E_{i,\theta}(0)], \qquad \Delta_I=\sum_i\mathbb E_\theta[I_{i,\theta}(1)-I_{i,\theta}(0)]. \] The summation boundary includes affected competitors or identifies a narrower direct effect explicitly. Decompose observed green price differences into climate-risk compensation, investor tastes and expected cash-flow changes using independently informative data. Determine when shifts in investor demand cause physical mitigation after accounting for new issuance, secondary trading, firms’ substitution and general equilibrium leakage.

Interpretation and scope

A "green premium" needs a sign convention: a higher price usually means a lower expected financial return conditional on comparable payoffs. Ex post green outperformance can arise from unexpected preference or climate news and does not establish higher ex ante required returns. Rating shocks may change information as well as demand, undermining an exclusion restriction based solely on financial costs. Firms can relabel assets, sell emitting operations or shift production, so disclosure improvements are not physical emissions reductions. Costs of capital are endogenous to risk and firm choices; a causal financing channel requires a defensible intervention and mediation assumptions. Welfare requires climate damages and any preference benefits, not the emissions statistic alone. State the intervention's assignment unit and law, consistency between observed and assigned outcomes, and positivity of both assignments on the target state support. A causal design additionally requires random assignment, conditional exchangeability, or another justified identification restriction; market spillovers belong in the assigned regime.

Source audit and status

[S1] concerns physical climate risks and market efficiency, [S2] carbon emissions and returns, and [S3] equilibrium sustainable investing. They do not jointly identify the causal physical-emissions effect of all ESG finance. This is an editorial asset-pricing and real-effects agenda; no universal sign or currently open theorem is asserted.

References

    },leftmargin=*]

  1. Harrison Hong, Frank Weikai Li, and Jiangmin Xu (2019). Climate Risks and Market Efficiency. Journal of Econometrics 208(1), 265–281. https://www.sciencedirect.com/science/article/pii/S0304407618301817. Locator and role: Publisher or author abstract / bibliographic record; Physical climate risk, food-sector returns and market efficiency. Provides background or a benchmark result; does not establish that the editorial question remains unresolved in 2026.

  2. Patrick Bolton and Marcin Kacperczyk (2021). Do Investors Care about Carbon Risk?. Journal of Financial Economics 142(2), 517–549. https://doi.org/10.1016/j.jfineco.2021.05.008. Locator and role: Publisher or author abstract / bibliographic record; Carbon emissions and stock returns; estimated association does not identify emissions response to financing costs. Provides background or a benchmark result; does not establish that the editorial question remains unresolved in 2026.

  3. Ľuboš Pástor, Robert F. Stambaugh, and Lucian A. Taylor (2021). Sustainable Investing in Equilibrium. Journal of Financial Economics 142(2), 550–571. https://doi.org/10.1016/j.jfineco.2020.12.011. Locator and role: Publisher or author abstract / bibliographic record; Equilibrium asset pricing with ESG tastes and climate-risk hedging. Provides background or a benchmark result; does not establish that the editorial question remains unresolved in 2026.

Common conventions: Source mathematical conventions

Unless an entry states otherwise, agent and firm sets are finite. State and action spaces are measurable, strategies and outcome maps are measurable, probability laws are specified, and expectations exist for the targets under discussion. Infinite-horizon discount factors lie in $(0,1)$ and discounted objectives are finite. Norms, tolerances and complexity input sizes must be specified for computational comparisons. Constants or primitives that appear locally are inputs, not empirical facts asserted by this catalogue.

Identification. A statistical model $m\in\mathcal M$ implies an observable law $P_m$ and target $\theta(m)$. At law $P$, its identified set is \[ \Theta(P;\mathcal M)=\{\theta(m):m\in\mathcal M,\ P_m=P\}. \] Point identification holds when this set is a singleton, or equivalently when $P_{m_1}=P_{m_2}$ implies $\theta(m_1)=\theta(m_2)$ for all compatible models. Sharp bounds mean bounds attained, or approached when the boundary is not attained, by compatible models. Writing this set is a definition, not a solution: the substantive task is to establish informative restrictions and derive or estimate the set. An unrestricted-confounding model may give only trivial bounds.

Inference. Identification, estimability and valid uncertainty quantification are separate questions. Fix $\alpha\in(0,1)$. A confidence procedure $C_n$ over a declared law class $\mathcal P$ has asymptotic uniform coverage at level $1-\alpha$ when \[ \liminf_{n\to\infty}\inf_{P\in\mathcal P}\Pr_P\{\theta(P)\in C_n\}\geq1-\alpha. \] For set-identified targets the coverage event must be defined separately, for example coverage of the full identified set. If an entry uses identification-indexed models instead of uniquely indexed laws, the same coverage convention is applied over those models. Pointwise limits do not establish uniform validity, and asymptotic validity does not guarantee finite-sample precision. Sample sizes, dependence structures and weak-identification sequences are part of each application.

Counterfactuals. $Y_i(a)$ is a potential outcome under a specified intervention, including its timing, implementation and versions. It is not $Y_i$ conditional on voluntarily choosing $a$. A full intervention vector permits interference; shorthand scalar treatment notation does not silently impose no interference. Assignment, selection, attrition, anticipation and measurement must be specified. A claim about an instrument's local effect is not automatically a population policy effect.

Economic equilibrium. A counterfactual model includes best responses, constraints, feasibility, market clearing, state transitions and expectations consistent with the stated information. When several equilibria exist, either a declared selector is an input or the target is set-valued. Comparisons across policy regimes must use the stated selection convention. Reduced-form relationships are not presumed invariant after an equilibrium-changing intervention.

Optimization and welfare. Optimization is over the stated feasible set. If attainment is not established, $\sup$ or $\inf$ and $\varepsilon$-optimal choices replace an asserted maximizer or minimizer. For example, with value $V=\sup_{a\in\mathcal A}W(a)<\infty$, an $\varepsilon$-optimal set is $\{a\in\mathcal A:W(a)\geq V-\varepsilon\}$ for $\varepsilon>0$. Benefits, costs, risk and health or environmental outcomes can be aggregated only using declared units and conversion weights. Interpersonal utility weights, the treatment of fiscal transfers and foreign profits, discounting, and the behavioral welfare criterion are explicit inputs. A comparative-static derivative additionally requires existence and appropriate differentiability of the selected counterfactual.

Sources. Local labels [S1], [S2], and so forth restart in every question. Locators identify the relevant abstract, model, section or result at the level actually checked; a bibliographic or abstract-level verification is not represented as inspection of an entire proof. Working papers are distinguished from later publications and changing versions. Source summaries are paraphrased. All URL checks and editorial formulations refer to the audit date stated above.

Statement licence: CC-BY-4.0 · Source file

Contributions 1

Econ-485-A1 · Exploration · LLM-generated · gpt 6 astra ultra ·

GPT 6 Astra Ultra attempt on Econ-485

Green asset prices and physical emissions: separating demand, issuance and production substitution

Problem ID: G12-481. Status: unresolved; restricted research attempt.

A price premium is only one equation

The target asks for causal financing-demand effects on investment and physical emissions, including competitors, and for a decomposition of green prices. This attempt proves an exact asset-market decomposition, an observational nonidentification result, and an equilibrium threshold for whether favorable green financing reduces aggregate emissions. The models are deliberately separate: a security-price change is not automatically a financing concession on marginal real investment.

Pástor, Stambaugh and Taylor [PST] provide the central equilibrium background. Their June 4, 2020 manuscript's introduction, including its statements about climate hedging and the two real-impact channels, was inspected. Our physical-tonnes accounting below is a restricted derivation, not their social-impact definition or a refutation of their model. Their published article appeared in 2021. No empirical estimates or universal ESG welfare claims are made here.

There are $n$ risky claims with terminal payoffs $X\in{\mathbb{R}}^n$, common mean $\mu$ and positive definite covariance $\Sigma$. The riskless gross return is normalized to one. Investor $j$ has unrestricted risky holdings $x_j$, risk tolerance $t_j>0$, and a per-unit nonpecuniary holding benefit $\eta_j\in{\mathbb{R}}^n$, in terminal-money units. A riskless position finances any risky portfolio, so the budget constraint is satisfied by adjusting that position; there are no borrowing or short-sale constraints. The investor maximizes the declared mean--variance objective \[ x_j^\mathsf T(\mu-P+\eta_j) -\frac{1}{2t_j}x_j^\mathsf T\Sigma x_j. \] Net risky supply is $s$. Aggregate wealth and the riskless supply are chosen consistently with initial endowments; riskless positions then clear by the sum of the individual budget identities.

Proposition (Exact clearing-price decomposition): Write $T=\sum_jt_j$ and $\bar\eta=T^{-1}\sum_jt_j\eta_j$. The unique risky-market clearing price is \[ P=\mu-\Sigma s/T+\bar\eta. \tag{1} \] For a fixed vector $v$ of green-minus-brown claim quantities, $v^\mathsf TP$ is the sum of an expected-cash-flow term, a risk-supply term, and a weighted-taste term. A change in tastes alone produces $\Delta P=\Delta\bar\eta$ when all other displayed primitives are fixed.
Proof: Positive definiteness makes each objective strictly concave and coercive. Its unique maximizer is $x_j=t_j\Sigma^{-1}(\mu-P+\eta_j)$. Summing and equating to $s$ yields (1). Conversely (1) makes these unique optimal portfolios sum to supply. Multiplying by $v^\mathsf T$ and subtracting two versions of (1) proves the decompositions. ∎
Proposition (Price and payoff data do not separate risk tolerance from tastes): Even when $\mu,\Sigma,s$ and $P$ are observed exactly, they need not identify the risk-supply and taste terms in (1). If unrestricted tastes are admitted, every $T>0$ is observationally compatible by choosing $\bar\eta=P-\mu+\Sigma s/T$.
Proof: For any specified $T>0$ use one investor with tolerance $T$ and taste vector equal to the displayed value. Proposition Proposition (Exact clearing-price decomposition) produces exactly the observed price and holds the same payoff law and supply fixed. Two different tolerances give different risk terms whenever $s\ne0$, since $\Sigma$ is invertible. Both economies have identical stated data. ∎

Climate-risk attribution requires an additional restriction on $\Sigma$ or its factor exposures and prices of risk. Labeling a security "green" does not supply one. If a taste intervention is randomized and excluded from cash-flow beliefs, covariance and participation, its price effect identifies the change in weighted taste conditional on this model; a rating disclosure that also changes beliefs fails that exclusion.

A real financing intervention with competitors

Now consider two differentiated-output producers. A representative quasilinear consumer has gross benefit \[ V(q_1,q_2)=\alpha_1q_1+\alpha_2q_2 -\tfrac12 b_1q_1^2-\tfrac12 b_2q_2^2-\rho q_1q_2, \] where $b_i>0$, $\rho\geq0$, $b_1b_2>\rho^2$. Quantities are nonnegative and the numeraire endowment is large enough for an interior demand solution. Inverse prices are $P_i=\alpha_i-b_iq_i-\rho q_j$. A unit of capacity investment produces one unit of output during the horizon, so $I_i=q_i$, with no resale or transfers of existing capacity. Each firm chooses quantity as a Cournot producer. Its real capacity cost is $c_iq_i+\kappa_iq_i^2/2$, $\kappa_i\geq0$.

The intervention is a funded concession $z d$ per unit of new capacity for firm 1, $d>0$, $z\in[0,1]$. It is paid by external investors or a declared program, which incurs the same cash outlay $zdq_1$. This is the financing wedge to be measured; we do not identify it with $\Delta P$ without an issuance model. Both firms pay a fixed real carbon charge $\tau\geq0$ per tonne. They choose abatement $a_i$ with real cost $k_i a_i^2/2$, $k_i>0$, and physical emissions $E_i=e_iq_i-a_i\geq0$, $e_i>0$. Thus profit is \[ (\alpha_i-b_iq_i-\rho q_j)q_i-c_iq_i-\tfrac12\kappa_iq_i^2 -\tau(e_iq_i-a_i)-\tfrac12k_ia_i^2+{\mathbf 1}\{i=1\}zdq_i. \] Let $D_i=2b_i+\kappa_i$ and $H=D_1D_2-\rho^2>0$. Assume the solution below satisfies $q_i>0$ and $0\leq\tau/k_i<e_iq_i$ for every $z\in[0,1]$. The latter restriction makes physical abatement feasible and avoids corner changes.

Theorem (A physical mitigation threshold): The unique equilibrium on this domain is \[ \begin{pmatrix}q_1(z)\\q_2(z)\end{pmatrix} =\begin{pmatrix}D_1&\rho\\\rho&D_2\end{pmatrix}^{-1} \begin{pmatrix}\alpha_1-c_1-\tau e_1+zd\\ \alpha_2-c_2-\tau e_2\end{pmatrix}, \qquad a_i=\tau/k_i. \] For the binary comparison $z=1$ versus $0$, \[ \Delta I=\frac{d(D_2-\rho)}{H},\qquad \Delta E=\frac{d(e_1D_2-e_2\rho)}{H}. \tag{2} \] Hence favorable financing of the lower-intensity firm reduces combined emissions exactly when $e_1/e_2<\rho/D_2$. Merely having $e_1<e_2$ is insufficient. If $\rho=0$, combined emissions strictly increase.
Proof: The abatement first-order condition is $k_ia_i=\tau$. The quantity first-order conditions are $D_iq_i+\rho q_j= \alpha_i-c_i-\tau e_i+{\mathbf 1}\{i=1\}zd$. Strict concavity of each firm's objective on its convex feasible set makes the displayed feasible interior point a best response. For uniqueness, the quantity game has a strictly concave potential with Hessian negative of the positive definite matrix displayed, and abatement adds independent strictly concave terms. The feasible constraints $a_i\leq e_iq_i$ are convex and individual; the first-order optimality inequalities for two equilibria, when added, would contradict strict monotonicity of the negative gradient unless the profiles coincide. Inverting the matrix gives $q_1'=dD_2/H$, $q_2'=-d\rho/H$. These derivatives are constant. Summing investments and physical emissions over both firms, with $a_i'=0$, and integrating over $[0,1]$ gives (2) and the sign claims. ∎

This result accounts for displaced competitor production. In particular, reported emissions of a reclassified asset need not follow (2): changing labels leaves $e_iq_i-a_i$ unchanged. If capacity and abatement are contractually fixed and only existing securities trade, changing $\eta_j$ in (1) changes prices without changing emissions or investment. That special case and Theorem Theorem (A physical mitigation threshold) establish why secondary-market premiums alone cannot identify the canonical physical effect.

A design that measures the stated total effect

Let independent markets $c$ contain all affected producers and consumers, and let $Z_c$ assign the entire financing regime with known $0<\pi_c<1$, independently of all potential outcomes. Define $Y_c(z)=\sum_{i\in c}E_i(z)$ with the same physical boundary at both assignments, including closures as zero and separately tracking entries. For a fixed number $C$ of markets set \[ \widehat\Delta_E=\frac1C\sum_c \left\{\frac{Z_cY_c}{\pi_c}- \frac{(1-Z_c)Y_c}{1-\pi_c}\right\}. \]
Proposition: With consistency, no interference across these markets, and finite mean potential emissions, ${\mathbb{E}}\widehat\Delta_E=C^{-1}\sum_c {\mathbb{E}}[Y_c(1)-Y_c(0)]$. The same formula applies to investment.
Proof: Conditional on each pair of potential outcomes, independence implies ${\mathbb{E}}[Z_cY_c/\pi_c]=Y_c(1)$ and ${\mathbb{E}}[(1-Z_c)Y_c/(1-\pi_c)]=Y_c(0)$. Integrate and sum. ∎

An exclusion restriction is unnecessary for this assignment effect, but is necessary for calling it an effect mediated solely by financing costs. Capital-market spillovers crossing assigned markets require a larger assignment unit or an explicit joint exposure model. Welfare additionally requires damage valuations, ownership, investor preference benefits and real resource costs; concession payments cancel only when payer and recipient have the same welfare weight and both lie within the accounting boundary. Neither a positive price premium nor negative $\Delta E$ alone establishes a welfare improvement. Estimating the issuance pass-through and production-substitution parameters is the central remaining task.

Completion Estimate

45%

This is a post hoc, heuristic estimate for the full catalogue target. The attempt proves a clearing-price decomposition, its taste-risk confound, a competitor-adjusted emissions threshold, and randomized market-level effect identification. Full ESG financing attribution still needs issuance pass-through, independently informative climate-risk data, mediator exclusions, and validated production and abatement responses.

References

Go to contribution →