Paper: arXiv 2505.01962
Authors: Haochuan Wang
Abstract
Understanding how market participants react to shocks like scheduled macroeconomic news is crucial for both traders and policymakers. We develop a calibrated data generation process DGP that embeds four stylized trader archetypes retail, pension, institutional, and hedge funds into an extended CAPM augmented by CPI surprises. Each agents order size choice is driven by a softmax discrete choice rule over small, medium, and large trades, where utility depends on risk aversion, surprise magnitude, and liquidity. We aim to analyze each agent’s reaction to shocks and Monte Carlo experiments show that higher information, lower aversion agents take systematically larger positions and achieve higher average wealth. Retail investors under react on average, exhibiting smaller allocations and more dispersed outcomes. And ambient liquidity amplifies the sensitivity of order flow to surprise shocks. Our framework offers a transparent benchmark for analyzing order flow dynamics around macro releases and suggests how real time flow data could inform news impact inference.
Complexity vs Empirical Score
- Math Complexity: 6.0/10
- Empirical Rigor: 3.0/10
- Quadrant: Lab Rats — high math complexity, limited empirical testing
Why this score: The paper employs advanced mathematical concepts like softmax discrete choice, an extended CAPM with macroeconomic shocks, and Monte Carlo simulations, which indicates a moderately high math complexity. However, it is entirely simulation-based with no real backtesting, datasets, or implementation details, resulting in low empirical rigor.
Research Flowchart
flowchart TD
A["Research Goal: <br> Model heterogeneous macro shock responses"] --> B["Methodology: <br> Extended CAPM + Agent-Based Modeling"]
B --> C{"Data / Inputs"}
C --> C1["Trader Archetypes:<br>Retail, Pension, Institutional, Hedge"]
C --> C2["Macro Shocks:<br>CPI Surprises"]
C --> C3["Market Conditions:<br>Ambient Liquidity"]
C --> D["Computation:<br>Softmax Discrete Choice Rule"]
D --> E["Simulation:<br>Monte Carlo Order Flow Dynamics"]
E --> F["Key Outcomes"]
F --> F1["High Info/Low Aversion:<br>Larger Positions, Higher Wealth"]
F --> F2["Retail Investors:<br>Under-react, Dispersed Outcomes"]
F --> F3["Liquidity Effect:<br>Amplifies Shock Sensitivity"]
Evidence Card
What the authors claim, on what data, and who has checked it. Read from the paper's own text by a language model: every quote is verbatim, nothing was re-run. How to read this card
Proposes
The paper proposes a calibrated data generation process (DGP) that embeds four trader archetypes (retail, pension, institutional, hedge funds) into an extended CAPM augmented by CPI surprises, using a softmax discrete‑choice rule for order‑size selection.
Source sentence
We develop a calibrated data generation process DGP that embeds four stylized trader archetypes retail, pension, institutional, and hedge funds into an extended CAPM augmented by CPI surprises. Each agents order size choice is driven by a softmax discrete choice rule over small, medium, and large trades, where utility depends on risk aversion, surprise magnitude, and liquidity.
Assumes 6
- Traders are assigned fixed risk‑aversion levels for each type. model
Source sentence
3 Data Generation Process 3.1 Market Participants Heterogeneity We classify traders into four types with fixed risk aversion levels: Retail, Pension Funds, Institutional Investors, and Hedge Funds.
- The model includes only a risk‑free asset and an inflation‑exposed risky asset, assumed independent. model
Source sentence
Simplified Two Assets Structure The model includes only a risk-free asset and an inflation-exposed risky asset, assuming independence between them.
- Risky‑asset returns follow a normal distribution with a fixed mean of 0.1% and standard deviations that vary with information level. statistical
Source sentence
Return Distribution Assumptions We assume risk-free asset has a constant Rf = 4% and that risky asset returns follow a normal distribution with a fixed mean 0.1% and varying standard deviations based on trader information levels. (eg.
- Liquidity is modeled as a normally distributed random variable (liquidityarr = 0.5 + 0.1·N(0,1)). model
Source sentence
Liquidity is modeled as a normally distributed random variable: liquidityarr = 0.5 + 0.1 · N(0, 1) to simulate period with different level of liquidity.
Show 2 more
- Macroeconomic shocks are independent and identically distributed normal shocks. statistical
Source sentence
However, here we assume shocks are independent and identically distributed (iid) normal shocks.
- Risk‑aversion is static over time for each trader type. model
Source sentence
Static Risk Aversion We assume traders have fixed risk aversion levels, however, in reality, risk pref-erences is dynamic and change with market conditions.
Data
- Calibrated data generation process (synthetic order‑flow data) synthetic unknown
Source sentence
We develop a calibrated data generation process DGP that embeds four stylized trader archetypes retail, pension, institutional, and hedge funds into an extended CAPM augmented by CPI surprises.
Costs
Modeled
Base transaction costs are specified per trader type: Retail 0.1, Pension 0.005, Institutional 0.005, HedgeFund 0.002.
Source sentence
Trader Type Risk Aversion Info Level Max Risk Base Transaction Cost Retail 3.0 0 0.25 0.1 Pension 2.0 0.7 0.4 0.005 Institutional 1.5 0.8 0.8 0.005 HedgeFund 1.0 1.0 1.0 0.002 Shock Model Currently, macroeconomic shocks are simulated as a combination of temporary and permanent com-ponents, where temporary shocks follow a mean-reverting normal process and permanent shocks follow a random walk.
Finds 3
- Higher‑information, lower‑aversion agents take systematically larger positions and achieve higher average wealth. performance
Source sentence
We aim to analyze each agent's reaction to shocks and Monte Carlo experiments show that higher information, lower aversion agents take systematically larger positions and achieve higher average wealth. We aim to analyze each agent’s reaction to shocks and Monte Carlo experiments show that (i) higher-information, lower-aversion agents take systematically larger positions and achieve higher average wealth; (ii) retail
- Retail investors under‑react on average, allocating smaller positions and exhibiting more dispersed outcomes. performance
Source sentence
Retail investors under react on average, exhibiting smaller allocations and more dispersed outcomes. We aim to analyze each agent’s reaction to shocks and Monte Carlo experiments show that (i) higher-information, lower-aversion agents take systematically larger positions and achieve higher average wealth; (ii) retail investors under-react on average, exhibiting smaller allocations and more dispersed outcomes; and
- Ambient liquidity amplifies the sensitivity of order flow to macro‑surprise shocks; with increasing liquidity all traders tend to increase trade size. performance
Source sentence
And ambient liquidity amplifies the sensitivity of order flow to surprise shocks. We aim to analyze each agent’s reaction to shocks and Monte Carlo experiments show that (i) higher-information, lower-aversion agents take systematically larger positions and achieve higher average wealth; (ii) retail investors under-react on average, exhibiting smaller allocations and more dispersed outcomes; and (iii) ambient
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 16 related papers). Coverage is partial, so this is neither replication nor refutation.
Code
No repository by the authors found in the paper.
Extracted 2026-11-14 · same card as JSON · A language model chose these items: check the source sentence before relying on one.