Providing Certainty

Andrew B. Choi; Chengyang Zhu; Christoph Schlom

arxiv: 2606.28583 · v1 · pith:6FXJCQHPnew · submitted 2026-06-26 · 💰 econ.TH

Providing Certainty

Andrew B. Choi , Christoph Schlom , Chengyang Zhu This is my paper

Pith reviewed 2026-06-30 00:35 UTC · model grok-4.3

classification 💰 econ.TH

keywords moral hazardinvestment timinginformation arrivalcommitmentstate-contingent paymentsenvironmental subsidiesR&D incentivesdynamic contracting

0 comments

The pith

A principal's optimal contract reduces the agent's payoff sensitivity to the state to prevent waiting for more information, even though this is inefficient.

A machine-rendered reading of the paper's core claim, the machinery that carries it, and where it could break.

The paper sets up a moral hazard problem in which an agent chooses the timing of an irreversible investment while public signals about a payoff-relevant state arrive gradually. The principal designs a state-contingent payment rule that the agent anticipates will be followed. To stop the agent from delaying in hopes of better information, the rule is chosen to shrink the variation in the agent's net payoff across states. Both parties would strictly prefer a rule with greater state dependence, yet the incentive to wait forces the principal to supply more certainty instead.

Core claim

In this setting the principal's optimal policy deliberately lowers the dependence of the agent's payoff on the realized state. This commitment discourages the agent from postponing investment until further signals arrive. The resulting contract is inefficient: both the principal and the agent would be better off under a less certain payment schedule. The analysis also identifies conditions under which the agent earns positive rent and under which moral hazard causes investment to occur later than in the first-best benchmark.

What carries the argument

The principal's ex-ante commitment to a state-contingent payment schedule that reduces the sensitivity of the agent's payoff to the state.

If this is right

The agent earns positive rent under the optimal policy in some parameter regions.
Moral hazard leads to delayed investment relative to the first-best timing.
The same logic governs the design of environmental subsidies and R&D incentive programs.
Both parties would gain if a less certain contract could be implemented without inducing delay.

Where Pith is reading between the lines

These are editorial extensions of the paper, not claims the author makes directly.

Policy makers facing similar timing problems may deliberately use fixed rather than performance-based payments to accelerate action.
The result highlights a general tension between dynamic information arrival and the value of commitment in contract design.
Empirical tests could compare subsidy structures across jurisdictions that differ in the credibility of long-term commitments.

Load-bearing premise

The principal can credibly commit in advance to a payment schedule that will be honored after the state is revealed.

What would settle it

Direct observation that a principal switches to a more state-dependent schedule once credible commitment is removed or once the agent doubts future honoring would falsify the claim that certainty is optimal.

Figures

Figures reproduced from arXiv: 2606.28583 by Andrew B. Choi, Chengyang Zhu, Christoph Schlom.

**Figure 1.** Figure 1: Signals and histories. to represent the two possible varieties of histories. History a indicates that there has been no breakdown. History bt indicates that a breakdown occurred in period t. Let H := {a, b1, . . . , bT } denote the set of all possible histories, and let h denote a generic element of H [PITH_FULL_IMAGE:figures/full_fig_p007_1.png] view at source ↗

**Figure 2.** Figure 2: Solving an inner problem (Example 1) The figure is based on Example 1 with T = 5. Panel (a) depicts a solution y to the inner problem [P(1)]. Since the agent does not stop if the history is b1, the policy y(b1) is irrelevant and may be set to 0. For t ≥ 2, the principal provides certainty by raising the policy under the bad state (y(bt) > y(a)), and certainty is front-loaded (y(bt) is nonincreasing in t). … view at source ↗

**Figure 3.** Figure 3: Solving Example 1 (T = 2) Panel (a) shows the solution y (t) to the inner problem [P(t)], for t = 0, 1, 2. Panel (b) shows the value of each inner problem. (Parameters: (p1, p2) = (0.8, 0.98), δ = 0.95, I = 5.5, s = 0.8, V = 0.002 and c(y) = 4y 3 .) Assumption 3 says that the principal’s ex post optimal policy does not depend on the agent’s investment timing (as long as the investment occurs at some point)… view at source ↗

**Figure 4.** Figure 4: Equilibrium payoff in Example 1 (constant ϕ(t), T = 5) Under the optimal policy y ∗ , the agent chooses τ1 and obtains U1(y ∗ ). As depicted by the dashed horizontal line, this payoff is equal to the agent’s maxmin payoff U4(0), which is what he gets if the principal sets y ≡ 0, and the agent best responds by choosing τ4. (Parameters: (pt) 5 t=1 = (0.45, 0.8, 0.7, 0.8, 0.98), δ = 0.95, I = 5.5, s = 0.8, V … view at source ↗

**Figure 5.** Figure 5: Equilibrium payoff in Example 2 (decreasing ϕ(t), T = 5) The agent waits until period 2 under ynull (point A), but the optimal policy y ∗ makes him invest earlier, in period 0 (point C). Hence, the agent obtains rent (C is above A). The intuition is that, in order to prevent the agent’s deviation away from C, it is optimal for the principal to increase y(a), which makes U2(y ∗ ) (point B) strictly higher t… view at source ↗

**Figure 6.** Figure 6: State-measurable policy rule (Example 1 with increasing A(t), T = 5) The principal can implement the optimal policy rule by waiting until some period and then announcing a state-measurable policy rule. (Parameters: pt = 0.9 for all t, δ = 0.95, I = 5, s = 0.55, V = 2 and c(y) = y 2 . Since s > (1 − δ)I and pt is constant, Proposition 5 applies.) and (7) is equivalent to 1 − λ > δptpt+1. (8) In particular, … view at source ↗

**Figure 7.** Figure 7: Solutions to inner problems (Example 1, decreasing A(t), T = 5) The solutions to inner problems are nested. For example, y (2) coincides with y (3) at every history except b3. (Parameters: (pt) 5 t=1 = ( 1 5 , 3 5 , 3 4 , 5 6 , 8 9 ), δ = 0.9, I = 3, s = 0.5, V = 0.1 and c(y) = y 2 . The waiting premium is decreasing: A(1) = 0.90, A(2) = 0.88, A(3) = 0.85, A(4) = 0.77, and A(5) = 0.59.) 6.2 Investment Timi… view at source ↗

read the original abstract

We introduce a moral hazard model in which public information about a payoff-relevant state arrives over time, an agent decides when to make an irreversible investment, and a principal commits to a state-contingent policy to incentivize investment. To discourage the agent from waiting for more information, the principal's optimal policy provides certainty, reducing the degree to which the agent's payoff depends on the state. This is inefficient -- both players would be better off with less certainty. We study when the agent receives positive rent, and when moral hazard delays investment. Our results apply to environmental subsidies and R&D incentives.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, simulated authors' rebuttal, and a circularity audit. Tearing a paper down is the easy half of reading it; the pith above is the substance, this is the friction.

Desk Editor's Note private letter to a colleague

The paper shows that with gradual public information arrival, the optimal contract provides certainty to deter waiting for investment, creating inefficiency under standard commitment.

read the letter

The main thing to know is that in this moral hazard setup, the principal optimally designs payments with less state dependence to stop the agent from delaying an irreversible investment until more information arrives. Both parties would prefer less certainty, but the incentive constraint forces the inefficiency.

The model adds gradual information arrival and the agent's timing choice to standard moral hazard, which is a clean extension. The inefficiency result follows directly once the principal commits to the state-contingent schedule, and the conditions for positive rent or delayed investment are derived from the usual participation and incentive constraints. The subsidy and R&D applications fit the setup without forcing it.

The commitment assumption does most of the work, as the reader noted; if the principal cannot credibly promise payments after the state realizes, the certainty provision unravels. The abstract gives no equations, so the exact proof strategy is unclear, but nothing in the described logic looks circular or fitted. No major internal contradictions appear.

This is for contract theorists working on dynamic incentives and information timing. It is coherent enough on its own terms to merit referee time, even if the contribution stays incremental.

Referee Report

2 major / 2 minor

Summary. The paper introduces a moral hazard model with public information arriving over time about a payoff-relevant state. An agent chooses the timing of an irreversible investment, while a principal commits ex ante to a state-contingent payment schedule. The central result is that the principal's optimal policy provides certainty—reducing the agent's payoff sensitivity to the realized state—to deter waiting for additional information. This policy is inefficient, as both parties would prefer less certainty. The analysis also characterizes conditions for positive agent rents and moral-hazard-induced delays, with applications to environmental subsidies and R&D incentives.

Significance. If the derivation holds, the paper contributes a clean mechanism-design insight into dynamic contracting under evolving information and irreversible actions. It shows how commitment to reduced state-dependence can substitute for direct timing incentives, generating an inefficiency that is not present in static moral-hazard settings. The applications to subsidy design are direct and falsifiable in principle.

major comments (2)

[§2] §2 (Model): The commitment assumption—that the principal can credibly pre-commit to a state-contingent schedule even after the state is publicly revealed—is load-bearing for both the optimality of certainty provision and the inefficiency result. If renegotiation is possible post-revelation, the agent's continuation value changes and the deterrence of waiting may fail; the paper should state explicitly whether this is ruled out by assumption or derived from equilibrium.
[§3] §3 (Main Result): The claim that the optimal policy reduces state-dependence (abstract, paragraph 2) requires an explicit comparison of the agent's value function under the certainty contract versus the full-information benchmark. Without the precise functional form of the payment schedule or the agent's outside option, it is unclear whether the inefficiency is strict or only weak.

minor comments (2)

[§2] Notation for the arrival process of public information and the agent's stopping time should be introduced with a timeline diagram to clarify the extensive form.
[§5] The applications paragraph would benefit from a short numerical example mapping the model's parameters to a concrete subsidy design (e.g., renewable-energy investment).

Simulated Author's Rebuttal

2 responses · 0 unresolved

We thank the referee for the constructive comments. We respond to each major comment below and will revise the manuscript accordingly where appropriate.

read point-by-point responses

Referee: [§2] §2 (Model): The commitment assumption—that the principal can credibly pre-commit to a state-contingent schedule even after the state is publicly revealed—is load-bearing for both the optimality of certainty provision and the inefficiency result. If renegotiation is possible post-revelation, the agent's continuation value changes and the deterrence of waiting may fail; the paper should state explicitly whether this is ruled out by assumption or derived from equilibrium.

Authors: The model is explicitly a commitment setting in which the principal designs and commits to the state-contingent schedule ex ante; this commitment is maintained after the state is realized. Renegotiation is ruled out by assumption, consistent with the standard mechanism-design approach to timing problems with irreversible actions. We will add a clarifying sentence in §2 stating that the analysis assumes no renegotiation after information arrival. revision: yes
Referee: [§3] §3 (Main Result): The claim that the optimal policy reduces state-dependence (abstract, paragraph 2) requires an explicit comparison of the agent's value function under the certainty contract versus the full-information benchmark. Without the precise functional form of the payment schedule or the agent's outside option, it is unclear whether the inefficiency is strict or only weak.

Authors: The proof of the main result (Theorem 1) derives the optimal contract by solving the principal's problem and directly compares the agent's continuation value under this contract to the value under the full-information benchmark contract. The comparison shows that the reduction in state-dependence is strict and generates a Pareto inefficiency. We will insert an explicit statement of this value-function comparison, together with the relevant functional forms, into the main text of §3. revision: yes

Circularity Check

0 steps flagged

No significant circularity; derivation self-contained

full rationale

The paper introduces a moral-hazard timing model with public information arriving over time and ex-ante commitment to state-contingent payments. The central claim—that the principal optimally supplies certainty to deter waiting, producing inefficiency—follows directly from the stated assumptions and standard contract-theory logic once commitment is granted. No equations, fitted parameters, or self-citation chains are visible in the provided text that would reduce any prediction to an input by construction. This is the expected outcome for a self-contained theoretical model whose results do not rely on internal re-labeling or load-bearing self-reference.

Axiom & Free-Parameter Ledger

0 free parameters · 0 axioms · 0 invented entities

Abstract-only review supplies no explicit free parameters, axioms, or invented entities; the model is described at the level of standard moral hazard assumptions.

pith-pipeline@v0.9.1-grok · 5611 in / 1018 out tokens · 25835 ms · 2026-06-30T00:35:55.293273+00:00 · methodology

discussion (0)

Reference graph

Works this paper leans on

3 extracted references

[1]

, T − t}

By part (b), we then have u(0, y(bt+r), t) ≤ 0 for all r ∈ { 1, . . . , T − t}. But since Ut ≥ 0, we must have u(1, y(a), t) ≥ 0. We thus have u(1, y(a), t) ≥ u(0, y(bt+1), t), as desired. Second, suppose that the constraint Ut ≥ Ut+s holds as a strict inequality for every s ∈ { 1, 2, . . . , T − t}. Then, the last statement in part (b) implies y(bt+1) = ...
[2]

The optimal policy is y(1) = 20 9 , y (10) = 4
[3]

That is, the agent receives positive rent

Importantly, the agent’s equilibrium payoff is pay(1) + (1 − p)ay(10) − I = 2 27 > 0, so the IR constraint is slack. That is, the agent receives positive rent. Note that, although the principal’s payoff is decreasing in y, limited liability of the agent does not bind –y is strictly positive under both states. Also note that, under the null policy rule whi...

[1] [1]

, T − t}

By part (b), we then have u(0, y(bt+r), t) ≤ 0 for all r ∈ { 1, . . . , T − t}. But since Ut ≥ 0, we must have u(1, y(a), t) ≥ 0. We thus have u(1, y(a), t) ≥ u(0, y(bt+1), t), as desired. Second, suppose that the constraint Ut ≥ Ut+s holds as a strict inequality for every s ∈ { 1, 2, . . . , T − t}. Then, the last statement in part (b) implies y(bt+1) = ...

[2] [2]

The optimal policy is y(1) = 20 9 , y (10) = 4

[3] [3]

That is, the agent receives positive rent

Importantly, the agent’s equilibrium payoff is pay(1) + (1 − p)ay(10) − I = 2 27 > 0, so the IR constraint is slack. That is, the agent receives positive rent. Note that, although the principal’s payoff is decreasing in y, limited liability of the agent does not bind –y is strictly positive under both states. Also note that, under the null policy rule whi...