Shortermism and excessive risk taking in optimal execution with a target performance

Emilio Barucci; Yuheng Lan

arxiv: 2505.15611 · v1 · submitted 2025-05-21 · 💱 q-fin.MF · q-fin.TR

Shortermism and excessive risk taking in optimal execution with a target performance

Emilio Barucci , Yuheng Lan This is my paper

Pith reviewed 2026-05-22 14:01 UTC · model grok-4.3

classification 💱 q-fin.MF q-fin.TR

keywords optimal executionperformance barriersshort-termisminventory coststarget remunerationstochastic controlrisk taking

0 comments

The pith

Remuneration based on hitting performance barriers makes brokers pursue more aggressive yet less risky execution strategies over short horizons, while producing poorer and more dispersed results over long horizons.

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

The paper studies the optimal execution problem in which a broker is paid according to whether accumulated wealth plus inventory value reaches an upper performance barrier before hitting a lower downside barrier. The performance measure incorporates trading profits, current market valuation of remaining shares, and a quadratic penalty on inventory holdings. Solving the associated stochastic control problem shows that, relative to the classical case without barriers, the incentive structure produces faster trading that also reduces downside risk when the horizon is short. When the horizon lengthens, the same incentives instead deliver lower expected performance and greater dispersion in outcomes.

Core claim

Over a short horizon, this type of remuneration leads, at the same time, to a more aggressive and less risky strategy compared to the classical one, and over a long horizon the performance turns to be poorer and more dispersed.

What carries the argument

A performance process defined as accumulated trading wealth plus market-valued inventory minus quadratic inventory costs, which defines upper and lower barriers that the broker's optimal trading control must approach or avoid.

If this is right

Short-horizon brokers liquidate positions more rapidly while keeping the probability of large losses lower than in the standard model.
Long-horizon execution under these incentives produces both lower average terminal wealth and higher variance of terminal wealth.
The incentive structure creates a tension between immediate trading aggression and long-run performance stability that is absent from barrier-free execution.

Where Pith is reading between the lines

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

Compensation contracts that reward barrier hitting may unintentionally shorten the effective planning horizon of execution desks.
Similar barrier-type incentives could be analyzed in other market-making or portfolio problems to check whether the short-versus-long horizon reversal persists.

Load-bearing premise

The broker can continuously adjust trading speed to steer the performance process toward the upper barrier while steering away from the lower barrier, without needing a fully specified price or liquidity model.

What would settle it

Numerical solution of the control problem or Monte Carlo simulation of trading paths under the barrier incentives versus the classical problem, checking whether short-horizon aggressiveness and risk reduction, and long-horizon degradation and dispersion, appear at the predicted magnitudes.

read the original abstract

We deal with the optimal execution problem when the broker's goal is to reach a performance barrier avoiding a downside barrier. The performance is provided by the wealth accumulated by trading in the market, the shares detained by the broker evaluated at the market price plus a slippage cost yielding a quadratic inventory cost. Over a short horizon, this type of remuneration leads, at the same time, to a more aggressive and less risky strategy compared to the classical one, and over a long horizon the performance turns to be poorer and more dispersed.

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

This adds dual performance and downside barriers to optimal execution but the claims on short-horizon aggressiveness with lower risk and long-horizon deterioration rest on price dynamics and impact rules that are not detailed enough to verify.

read the letter

This paper extends the optimal execution problem by adding a target performance barrier and a downside barrier that the broker tries to hit or avoid. The key takeaway is that short horizons make the optimal strategy more aggressive but less risky than the standard approach, while long horizons lead to worse and more spread-out performance. What is new is the use of these dual barriers in a setup with quadratic inventory costs, building directly on Almgren-Chriss style models. It does a good job of showing how remuneration structures with performance targets can alter trading behavior in ways that depend on the time frame. The modeling seems solid in framing the problem as a stochastic control task with barriers. The qualitative results on horizon effects are interesting and could have implications for how firms set incentives for execution desks. The main soft spot is that the underlying price process and the exact form of temporary and permanent market impact are not spelled out in the abstract or summary. This makes it hard to confirm how the Hamilton-Jacobi-Bellman equation is derived and solved, or why the short-horizon strategy is both faster and safer. If those elements are standard Brownian motion with linear impact, the claims might hold, but without seeing the details it's difficult to assess robustness. The lack of numerical illustrations or comparisons to real data also leaves the practical relevance a bit open. This work is aimed at researchers in mathematical finance who focus on optimal trading and incentive design. Someone familiar with execution algorithms would find value in the barrier extension and the horizon comparison. It is worth sending to peer review. The idea is coherent and the results are potentially useful, so referees could help strengthen the assumptions and add verification steps.

Referee Report

1 major / 1 minor

Summary. The paper examines an optimal execution problem in which a broker seeks to reach an upper performance barrier while avoiding a lower downside barrier. Performance is defined as accumulated trading wealth plus the mark-to-market value of remaining inventory, subject to quadratic inventory penalties. The central claim is that, relative to the classical (no-barrier) execution problem, short-horizon strategies under this remuneration structure are simultaneously more aggressive and less risky, whereas long-horizon performance is both poorer on average and more dispersed.

Significance. If the derivations and comparisons hold, the results would illustrate how barrier-based performance targets can induce short-termism and altered risk profiles in execution algorithms, with potential implications for incentive design in brokerage and market-making. The work would benefit from explicit verification that the claimed qualitative differences survive under standard price dynamics and impact assumptions.

major comments (1)

The model formulation does not specify the SDE for the unaffected asset price (drift, volatility, jumps) or the map from trading rate to cash proceeds and price impact (temporary/permanent, linear/nonlinear). Without these elements the Hamilton-Jacobi-Bellman equation cannot be written or solved, rendering the comparison of aggressiveness (|u|) and risk (hitting probabilities or variance) to the classical case unverifiable. This is load-bearing for the short-horizon and long-horizon claims.

minor comments (1)

Notation for the performance process and barrier levels should be introduced with explicit definitions before the optimization problem is stated.

Simulated Author's Rebuttal

1 responses · 0 unresolved

We thank the referee for their careful reading and constructive feedback on our manuscript. We address the major comment regarding model specification below and will revise the paper to improve clarity and verifiability of the results.

read point-by-point responses

Referee: The model formulation does not specify the SDE for the unaffected asset price (drift, volatility, jumps) or the map from trading rate to cash proceeds and price impact (temporary/permanent, linear/nonlinear). Without these elements the Hamilton-Jacobi-Bellman equation cannot be written or solved, rendering the comparison of aggressiveness (|u|) and risk (hitting probabilities or variance) to the classical case unverifiable. This is load-bearing for the short-horizon and long-horizon claims.

Authors: We agree with the referee that explicit specification of the price dynamics and impact map is necessary for full verifiability. In the revised manuscript we will add a dedicated paragraph in Section 2 stating that the unaffected mid-price follows the geometric Brownian motion dS_t = μ S_t dt + σ S_t dW_t (no jumps), and that trading at rate u_t generates cash proceeds at the rate u_t (S_t - κ u_t) dt under linear temporary impact with coefficient κ > 0 and no permanent impact. These are the exact assumptions underlying the HJB equation and the numerical comparisons already presented. We will also include a brief remark confirming that the same dynamics and impact are used for the classical (no-barrier) benchmark, so that the reported differences in aggressiveness and risk remain directly comparable. This addition will not alter any results or proofs but will make the derivations self-contained. revision: yes

Circularity Check

0 steps flagged

No significant circularity detected

full rationale

The paper defines the performance process explicitly as accumulated wealth plus mark-to-market inventory valuation minus quadratic inventory costs, then formulates the broker's problem as an optimal control task to hit an upper performance barrier before a lower one. The qualitative comparisons (more aggressive/less risky over short horizons, poorer/dispersed over long horizons) are obtained by solving the resulting optimization problem and contrasting it with the classical no-barrier execution case. No equation reduces a claimed prediction to a fitted parameter by construction, no uniqueness theorem is imported via self-citation, and no ansatz is smuggled in; the derivation remains self-contained within the stated control framework and barrier setup.

Axiom & Free-Parameter Ledger

2 free parameters · 2 axioms · 0 invented entities

The central claim rests on the definition of performance including quadratic costs and the existence of optimal strategies for the barrier problem, which are standard assumptions in the field but not independently verified here.

free parameters (2)

performance barrier level
The target performance level is likely a parameter in the model.
downside barrier level
The loss avoidance level is a key parameter.

axioms (2)

domain assumption The asset price follows a stochastic process allowing for optimal control.
Standard in optimal execution models.
domain assumption Inventory costs are quadratic in remaining shares.
Specified in abstract as yielding quadratic inventory cost.

pith-pipeline@v0.9.0 · 5608 in / 1249 out tokens · 47028 ms · 2026-05-22T14:01:13.203944+00:00 · methodology

discussion (0)

Lean theorems connected to this paper

Citations machine-checked in the Pith Canon. Every link opens the source theorem in the public Lean library.

IndisputableMonolith/Cost/FunctionalEquation.lean washburn_uniqueness_aczel unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

The optimal strategy for Problem P1 is v*(t) = (2γ−b)/(2l) Q*(t), Q*(t) = exp((b−2γ)/(2l) t) Q0. The value function is J(y) = (e^{−λy} − e^{−λk}) / (e^{−λh} − e^{−λk}) with λ = (2γ−b)^2 / (2l σ^2).
IndisputableMonolith/Foundation/AlphaCoordinateFixation.lean J_uniquely_calibrated_via_higher_derivative unclear

?

unclear
Relation between the paper passage and the cited Recognition theorem.

Performance Y(t) = X(t) + Q(t)(S(t) − γ Q(t)) with linear impact f(v)=b v, g(v)=l v and quadratic slippage.

What do these tags mean?

matches: The paper's claim is directly supported by a theorem in the formal canon.
supports: The theorem supports part of the paper's argument, but the paper may add assumptions or extra steps.
extends: The paper goes beyond the formal theorem; the theorem is a base layer rather than the whole result.
uses: The paper appears to rely on the theorem as machinery.
contradicts: The paper's claim conflicts with a theorem or certificate in the canon.
unclear: Pith found a possible connection, but the passage is too broad, indirect, or ambiguous to say the theorem truly supports the claim.