Paper: arXiv 2509.05922
Authors: Peilin Rao, Randall R. Rojas
Abstract
This paper provides robust, new evidence on the causal drivers of market troughs. We demonstrate that conclusions about these triggers are critically sensitive to model specification, moving beyond restrictive linear models with a flexible DML average partial effect causal machine learning framework. Our robust estimates identify the volatility of options-implied risk appetite and market liquidity as key causal drivers, relationships misrepresented or obscured by simpler models. These findings provide high-frequency empirical support for intermediary asset pricing theories. This causal analysis is enabled by a high-performance nowcasting model that accurately identifies capitulation events in real-time.
Complexity vs Empirical Score
- Math Complexity: 8.5/10
- Empirical Rigor: 9.0/10
- Quadrant: Holy Grail — high math complexity, high empirical rigor
Why this score: The paper employs advanced causal machine learning (DML, APE) with high-dimensional data, justifying a high math score, while its detailed backtesting, extensive data sourcing, and implementation-focused pipeline indicate strong empirical rigor.
Research Flowchart
flowchart TD
A["Research Goal: Identify Causal Drivers<br>of Market Troughs"] --> B["Data: High-Frequency Market Data<br>Options Implied Volatility<br>Liquidity Indicators"]
B --> C{"Methodology: Causal ML"}
C --> D["Computational Process 1<br>Flexible DML Framework"]
C --> E["Computational Process 2<br>Nowcasting Capitulation Events"]
D --> F["Key Findings: Robust Causal Estimates"]
E --> F
F --> G["Outcomes"]
G --> H["1. Volatility of Options-Implied<br>Risk Appetite is a Key Driver"]
G --> I["2. Market Liquidity is a Key Driver"]
G --> J["3. Misrepresented by<br>Simpler Linear Models"]
G --> K["4. Support for<br>Intermediary Asset Pricing Theories"]
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 high‑performance nowcasting model for market troughs and a comparative causal analysis using DML‑APE to identify causal drivers.
Source sentence
We have created a transparent, high performance nowcasting model of market troughs with outstanding out-of-sample accuracy.
Assumes 6
- The outcome variable is binary (trough vs non‑trough). model
Source sentence
the binary outcome and non-linear interactions problem
- The treatment (risk‑appetite measure) follows a heteroskedastic Gaussian process conditional on confounders. statistical
Source sentence
Given a standard and flexible assumption that the treatment is a heteroskedastic Gaussian process conditional on confounders
- Trades are entered at the closing price of the ES contract on the signal day. execution
Source sentence
The long position is entered at the closing price of the ES contract on day t.
- A flat round‑trip commission and slippage cost of $5.00 per contract is applied. execution
Source sentence
a round-trip commission and slippage cost of $5.00 is deducted from the profit or loss of each contract traded.
Show 2 more
- SMOTE is used to address class imbalance in training data. data
Source sentence
We use SMOTE (Synthetic Minority Ov…
- Feature aggregation (mean, std, trend, last value) yields stationary features. statistical
Source sentence
An Augmented Dickey-Fuller (ADF) test demonstrated that the percentage of stationary features increased from 90.6% in the parent set to 100% in the final aggregated set
Data
2023-07 – 2025-06 · E-mini S&P 500 futures · futures · daily
- SPX.EOD options commercial
Source sentence
SPX End-of-Day Option Chains Apr 2013 - Jun 2025 Daily
- CBOE.SPX.OPNINT options commercial
Source sentence
SPX Option Open Interest Apr 2013 - Jun 2025 Daily
- CME.ES.OHLCV.1M market-prices commercial
Source sentence
E-mini S&P 500 Futures OHLCV Apr 2013 - Jun 2025 1‑Minute
- CME.ES.BBO.1S order-book commercial
Source sentence
E-mini S&P 500 Futures BBO Apr 2013 - Jun 2025 1‑Second
- ^VIX macro public
Source sentence
CBOE Volatility Index (VIX) Apr 2013 - Jun 2025 Daily
- EFFR macro public
Source sentence
Effective Federal Funds Rate Apr 2013 - Jun 2025 Daily
Costs
Partial
A flat round‑trip commission and slippage cost of $5.00 per contract is deducted from each trade.
Source sentence
a round-trip commission and slippage cost of $5.00 is deducted from the profit or loss of each contract traded.
Finds 4
Sharpe 2.03
- The SVM classifier achieves an out‑of‑sample ROC AUC of 0.89. predictive-accuracy
Source sentence
produced an ROC AUC of 0.89 out-of-sample
- Volatility of options‑implied risk appetite and market illiquidity are identified as causal drivers of market troughs under the DML‑APE specification. other
Source sentence
we found that it is not simply the level of fear, but the volatility of options‑implied risk appetite (e.g., GEX, risk‑neutral skewness) and the instability of market illiquidity (Amihud illiquidity) that causally drive market troughs.
- The DML‑APE model corrects the sign of several key causal effects and discovers new causal pathways missed by the PLR model. theoretical
Source sentence
The DML‑APE framework not only corrected the sign of some key causal effects, but also discovered new causal pathways related to market volatility.
- A simple E‑mini S&P 500 futures backtest using the nowcast signal detects V‑shaped price reversals but struggles in grinding bear markets. performance
Source sentence
confirmed the economic significance of the nowcast signal as a market capitulation detector, which performed well in its ability to detect V‑shaped price reversals, though it struggles in grinding bear markets.
Checked by others
No other paper in the archive has been found to support or contradict these findings (compared with 36 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.
Cited References (25)
- Illiquidity and stock returns: cross-section and time-series effects (2002) — link
- Modeling and Forecasting Realized Volatility (2003) — link
- Value and Momentum Everywhere (2013) — link
- Stock Return Characteristics, Skew Laws, and the Differential Pricing of Individual Option Contracts (2003) — link
- The Financial Accelerator in a Quantitative Business Cycle Framework (1999)
- Put–Call Ratios and Market Timing Effectiveness (1988) — link
- Expected stock returns and variance risk premia (2009) — link
- Market Liquidity and Funding Liquidity (2009) — link
and 17 more
- Cyclical Analysis of Time Series: Selected Procedures and Computer Programs (1971)
- Predicting excess stock returns out of sample: Can anything beat the historical average? (2008) — link
- Double/debiased machine learning for treatment and structural parameters (2018) — link
- Long Story Short: Omitted Variable Bias in Causal Machine Learning (2022) — link
- Making Sense of Sensitivity: Extending Omitted Variable Bias (2020) — link
- Flow Toxicity and Liquidity in a High-Frequency World (2012) — link
- Business Conditions and Expected Returns on Stocks and Bonds (1989) — link
- Taming the Factor Zoo: A Test of New Factors (2020) — link
- The Leverage Cycle (2010) — link
- A Comprehensive Look at the Empirical Performance of Equity Premium Prediction (2008) — link
- Empirical Asset Pricing via Machine Learning (2020) — link
- Intermediary Asset Pricing (2013) — link
- Credit Cycles (1997) — link
- Consumption, Aggregate Wealth, and Expected Stock Returns (2001) — link
- A Unified Approach to Interpreting Model Predictions (2017)
- The Information in Option Volume for Future Stock Prices (2006) — link
- The Investor Fear Gauge (2000) — link
0 of 25 cited works are in the corpus; resolved via arXiv source bundles, Semantic Scholar and OpenAlex. Corpus-wide: 2596 linked papers.