Multimodal Deep Reinforcement Learning for Portfolio Optimization

James Zhang; Ravi Panguluri; Sashwat Venkatesh; Sumit Nawathe

REVIEW 5 major objections 5 minor 14 references

Multimodal Deep Reinforcement Learning for Portfolio Optimization

T0 review · 5 major / 5 minor · reviewed 2026-08-11 · deepseek-v4-flash

Challenge this review Re-run · record.json Download PDF Read on arXiv ↗ Review JSON

Pith's one-line read A deep RL trader using SEC and news sentiment beats most S&P100 benchmarks in backtests.

desk verdict A careful EIIE re-implementation with sentiment channels whose own tables contradict the paper's headline claim that combining data sources improves performance. read the letter →

arxiv 2412.17293 v1 pith:FNPBHP6D submitted 2024-12-23 q-fin.PM

Sumit Nawathe , Ravi Panguluri , James Zhang , Sashwat Venkatesh This is my paper · ORCID

classification q-fin.PM MSC 91G1068T07

keywords reinforcementlearningportfoliooptimizationmultimodaldatasentimentanalysisSECfilingsnewsheadlinesEIIEpolicyDifferentialSharpeRatio

verification ladder T0 review T1 audit T2 compute T3 formal

The pith

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

The reading

This paper tries to show that feeding a deep reinforcement learning portfolio agent both price data and sentiment signals from SEC filings and news headlines improves trading performance on S&P100 stocks. Its central claim is that the best configuration — a CNN-based EIIE policy trained on the combined price-plus-sentiment state with a simple profit reward — posts the highest net profit, Sharpe ratio, and Sortino ratio among all trained agents and reference benchmarks in the 2018–2019 test period, with only Equal Buy-and-Hold doing better. The authors also argue that the profit reward is far easier for these agents to learn than the Differential Sharpe Ratio, and that compact CNN/RNN policies overfit less than a full-tensor MLP. A sympathetic reader would care because the result suggests a practical recipe: regular alternative data, a compact policy, and a simple reward can outperform established online portfolio selection methods on a large-cap universe.

What carries the argument

The central object is the state tensor $S=(S^*,\delta)$ that adds alternative-data channels to historical prices: for each asset and each day in the lookback window, the tensor stores high, low, and close prices together with a SEC sentiment score (Loughran–McDonald word proportions) and a news sentiment score from FinBERT mapped through the $\tanh(\text{positive}/\text{negative}/\text{neutral})$ embedding of Equation (1). The policy is the Ensemble of Identical Independent Evaluators (EIIE) from the reference framework, where one CNN or RNN feature extractor scores each asset independently and a softmax converts scores into portfolio weights; the MLP baseline applies the same tensor to a single fully connected network. Rewards are either period profit or the Differential Sharpe Ratio, and transaction costs enter through the fixed-point update factor $\mu_t$. Sparse sentiment dates are forward-filled by exponential decay $y=a(1-\gamma)^t$ with $\gamma\approx0.8$.

What would settle it

Retrain the best configuration (CNN EIIE, combined data, profit reward) with the news and SEC channels replaced by constant zeros, and separately with the same channels filled with random noise matched to the observed sentiment distribution; if test-period net profit, Sharpe ratio, and Sortino ratio stay within noise of the reported 0.166, 0.554, and 0.665, then the sentiment channels are not carrying the claimed signal.

Watch

Extended reading notes

Core claim

The paper's central discovery, stated on its own terms, is that the best-performing model across all three experiment sets is the CNN EIIE model trained on the combined dataset (price, SEC sentiment, and news sentiment) with the Profit reward. In the 2018–2019 out-of-sample test this agent reports a net profit of 0.166, Sharpe ratio of 0.554, Sortino ratio of 0.665, and max drawdown of 0.195, which beats the S&P500 index as well as the OLMAR and WMAMR benchmarks; only the Equal Buy-and-Hold portfolio has higher net profit and risk-adjusted ratios. The paper interprets this as evidence that multimodal company sentiment enriches the state representation, that SEC data in particular is regular and reliable enough to help learning, and that the choice of reward function matters more than model complexity once alternative data is present.

Load-bearing premise

The load-bearing premise is that exponential-decay forward-filling turns sparse SEC filings and news headlines into a faithful daily sentiment signal, because most of the sentiment values the agent sees are imputed rather than measured.

Editorial extensions

If this is right

A practical takeaway is that adding regularly reported alternative data (SEC sentiment) to a compact EIIE policy improves returns under both profit and Differential Sharpe rewards.
The profit reward is a more reliable training signal than the Differential Sharpe Ratio for these RL agents, so practitioners can expect better out-of-sample results from the simpler reward.
CNN and RNN EIIE policies are preferable to a full-tensor MLP because their smaller parameter counts reduce overfitting on the price-plus-sentiment state.
Combining news with SEC data gives the strongest configuration, while news alone is too sparse and irregular to help, especially under the harder reward.
All trained agents in the comparison outperform the OLMAR and WMAMR online portfolio benchmarks over the test period.

Reading between the lines

Editorial extensions of the paper, not claims the author makes directly.

Beyond the paper's claims: the reported gains could be an artifact of the forward-filling rule, since the median ticker has actual news on only 905 trading dates; retraining with unfilled or shuffled sentiment would isolate the true signal.
The sentiment embedding in Equation (1) is an ad hoc ratio that ignores the absolute level of neutral probability; a calibrated alternative such as the raw FinBERT positive-minus-negative probability would test how much the embedding form matters.
The conclusions are drawn from one 2018–2019 test window with 1% transaction costs; extending the same architecture to other periods, universes, or cost levels would show whether the recipe generalizes.
A testable extension is to use the same state tensor with a learned attention or gating layer over the sentiment channels; if the model learns to ignore sentiment on days after long gaps, that would corroborate the paper's interpretation that data regularity drives the improvement.

Share X Bluesky LinkedIn Reddit HN

Signed reviews

No signed human review yet.

Editorial analysis

A structured set of objections, weighed in public.

Desk editor's note, referee report, and a circularity audit.

Desk Editor's Note

read the letter

You should know two things up front. First, the paper does real work: it re-implements the EIIE architecture, adds SEC-filing and news-headline sentiment as extra state channels, and introduces a specific sentiment embedding function (their Eq. 1). The authors are also transparent about data gaps, especially the sparse news coverage. Second, the central claim—that the combined-data CNN with the profit reward is the best model—is directly contradicted by their own tables. Table 7 shows the price-only CNN achieving net profit 0.1667 and Sharpe 0.5615, while Table 11 shows the combined CNN at 0.1661 and 0.5536. The combined model loses. The comparison table that supposedly supports the headline omits the price-only CNN entirely, which looks like cherry-picking the table rather than the model.

The paper is a legitimate extension of an established line of work, not a breakthrough. The sentiment embedding is novel to this paper, and the empirical setup is reasonable in scope: S&P100 stocks, 2010-2020, multiple reward functions and policy architectures. But the extension is incremental, and the claimed edge is small even if it existed. The best RL policy still underperforms a trivial equal-weight buy-and-hold (0.1811 net profit, 0.5952 Sharpe). There are no error bars, no multiple-seed runs, and the best model is selected on test-period metrics, which inflates the risk of overfitting to the test window.

The biggest soft spot is the mismatch between the prose and the numbers. Section 4.4 says the combined-data CNN with profit reward is the best across all experiments; Section 4.5 repeats that claim and adds that it has the highest Sharpe and Sortino. Both statements are false for the price-only CNN, which beats it on all three metrics. That is not a subtle issue. It is the load-bearing result of the paper, and it fails on the paper's own evidence. The news data issue is secondary: with only 84 of 100 tickers covered and a median of 905 reporting dates, the imputed sentiment tensor is mostly a product of the decay rule, and the paper acknowledges this. But even if the data were perfect, the headline claim would still be unsupported.

Who should read this? Someone working on RL portfolio allocation might look at it as a cautionary example of how to report baselines and why single-seed results are not enough. It is not a result to build on. My recommendation: do not send this to peer review in its current form. The internal contradiction alone warrants a reject or a major-revision request before any referee spends time on it. The authors should fix the comparison, add error bars, and decide whether the multimodal claim survives contact with their own tables.

Referee Report

5 major / 5 minor

Summary. The paper proposes a deep reinforcement learning framework for S&P100 portfolio optimization that augments price history with sentiment features derived from SEC filings and news headlines, using EIIE-style CNN, RNN, and MLP policies under profit and Differential Sharpe rewards. The authors report backtests over a 2018–2019 test period and claim that the CNN EIIE model trained on combined price+SEC+news data with a profit reward is the best-performing model across all experiments, and that it exceeds OLMAR and WMAMR benchmarks. The paper also describes data processing, an exponential-decay forward-filling rule for sparse sentiment observations, and a transaction-cost-aware environment.

Significance. If the central claim were supported, the paper would provide a moderately useful incremental demonstration that multimodal sentiment channels can be integrated into EIIE-style RL portfolio policies. The manuscript has several strengths: it carefully describes the data pipeline, implements a nontrivial transaction-cost model from prior work, and explicitly acknowledges data sparsity limitations. However, the reported evidence does not support the headline claim. The paper's own tables show that a price-only CNN with profit reward outperforms the combined-data CNN on net profit, Sharpe, and Sortino ratios, and that the simple Equal Buy-and-Hold benchmark beats every trained RL policy. The selected comparison table omits the price-only configuration that contradicts the claim. Without multiple seeds or error bars, the small apparent differences among RL variants cannot be distinguished from noise. The forward-filling rule for news sentiment, which supplies the majority of the multimodal state entries for most tickers, is a load-bearing assumption that is not validated. These issues undermine the paper's central contribution as presented.

major comments (5)

[§4.4, Table 11 vs. Table 7] The claim in §4.4 that 'our best-performing model across all three sets of experiments is the CNN EIIE model on the combined dataset using Profit reward' is contradicted by the paper's own numbers. Table 7 reports that the price-only CNN with Profit reward achieves net profit 0.166717, Sharpe ratio 0.561471, and Sortino ratio 0.676669, whereas Table 11 reports for the SEC+News CNN with Profit reward net profit 0.166081, Sharpe ratio 0.553644, and Sortino ratio 0.664706. The price-only model outperforms the combined model on all three metrics, so the stated 'best-performing model' is not the best according to the reported results.
[§4.5, Table 12] Table 12, which is used to support the headline comparison of 'best strategies,' omits the price-only CNN with Profit reward from Table 7. Including that row would show that the selected SEC+News CNN is not the best trained policy, and the text's statement that the SEC+News CNN Profit policy has the highest net profit, Sharpe, and Sortino excluding Equal Buy-and-Hold is factually incorrect given the paper's own Table 7 values. The comparison must include all configurations, especially those that outperform the configuration being highlighted.
[§4.2–§4.5, Tables 5–12] Every trained RL policy reported in Tables 6–11 has lower net profit and Sharpe ratio than the simple Equal Buy-and-Hold benchmark, which achieves net profit 0.181138 and Sharpe 0.595225. The abstract's claim that the approach 'delivers superior portfolio performance' and 'outperform standard benchmarks' is therefore not supported; the trained agents only outperform OLMAR and WMAMR, which are themselves weak in this test period. The benchmark comparison should be presented in full, and the conclusion should be revised to acknowledge that the method does not beat the equal-weight buy-and-hold baseline.
[§2.2.3 and Eq. (2)] The multimodal state tensor is largely constructed by exponential-decay forward-filling of sparse sentiment observations. News coverage exists for only 84 of 100 tickers, and the median ticker has reports on only 905 trading dates out of the roughly 2,500 trading days in the full period. Eq. (2) with a tuned parameter γ≈0.8 imputes sentiment scores for the vast majority of state entries, and the paper itself notes in §2.2.3 that 'coverage is still incomplete.' Because the imputation rule determines most of the sentiment channel, the comparisons between price-only, SEC-only, and combined configurations reflect properties of the filling rule as much as the information content of the underlying news data. The paper should provide sensitivity analysis over the imputation rule (e.g., different decay factors, no imputation) before attributing performance differences to the data sources.
[§4, all tables] No error bars, confidence intervals, or multiple-seed results are reported for any RL policy. Differences between configurations—for example, SEC+News CNN (0.166081) versus SEC-only CNN (0.162534) under Profit reward—are small and may be within run-to-run variance, especially given that policy-gradient training is stochastic. The paper should report mean and standard deviation over at least several independent training runs, or otherwise provide statistical evidence that the claimed improvement is not noise.

minor comments (5)

[Table 11 caption] The caption of Table 11 says 'Strategies with Combined Data (DiffSharpe Reward)' but the surrounding text in §4.4 states that the table reports results for the Profit reward. The caption should be corrected to 'Profit Reward' to match the content.
[§2.2.3, Table 3] The statistic 'Minimum Observations 1' is ambiguous: it could mean one reporting date or one headline. Please clarify whether the count is of reporting dates or individual headlines.
[Reference [10]] The reference contains a typo: 'finaincial' should be 'financial.'
[§4.3, Table 9] The text states 'The CNN and RNN policies using the Differential Sharpe ratio are among the strongest contenders,' but Table 9 shows Profit-reward results for SEC data. The intended reference is likely Table 8; please correct the cross-reference.
[§3.6] The notation for the EIIE functions is inconsistent: the text uses fα, gβ, and γ, but also refers to fθ1 and trainable parameters α, β, and γ without defining θ1. Please harmonize the notation.

Circularity Check

0 steps flagged · score 0.0 of 10

No significant circularity: the RL pipeline, reward functions, and sentiment embeddings are externally defined, and no prediction reduces to a fitted input or self-citation.

full rationale

The paper's derivation chain is self-contained in the circularity sense. The state tensor is assembled from CRSP prices, Loughran-McDonald SEC sentiment proportions, and FinBERT news sentiment via Equation (1); the reward functions (profit and Differential Sharpe Ratio) are defined externally, and the EIIE policy architecture is adopted from prior work [3,4]. The only tuned quantity is the exponential-decay hyperparameter gamma in Equation (2), which is conventional hyperparameter fitting and not a prediction target. No load-bearing claim is justified solely by a self-citation, and no uniqueness theorem is imported from the authors' own prior work. The paper itself flags a data limitation in Section 2.2.3 ('our coverage is still incomplete'), which is a soundness caveat rather than a circular step. A separate, non-circular correctness concern remains: Section 4.4's claim that the combined-data CNN is 'best-performing across all three sets of experiments' is contradicted by the paper's own Table 7, where the price-only CNN with Profit reward has higher net profit (0.166717), Sharpe (0.561471), and Sortino (0.676669) than the combined-data CNN in Table 11 (0.166081, 0.553644, 0.664706), and every trained policy underperforms Equal Buy-and-Hold (0.181138, 0.595225). This is an evaluation-selection or reporting issue, not circularity.

Assumptions & free parameters 3 free parameters · 4 assumptions · 0 invented entities

The central empirical claim rests on several modeling assumptions: sentiment scores derived from FinBERT and Loughran-McDonald dictionary accurately capture informative signals; exponential decay forward-filling reconstructs missing daily sentiment; the EIIE policy class and profit reward are appropriate; and the test period is representative. There are no invented causal entities. The main free parameter is the decay gamma, tuned to about 0.8, and model selection is done on test metrics.

free parameters (3)

sentiment decay factor gamma = ~0.8
Controls exponential decay for forward-filled sentiment scores for both SEC and news data (Equation 2); tuned as a hyperparameter in training (Sections 2.1.1, 2.2.2).
transaction cost rate = 1%
Applied to all strategies (Section 4); chosen by hand, materially affects comparison between rebalancing RL agents and buy-and-hold benchmarks.
state lookback window H = not reported
The state tensor includes a historical window of 'the past few weeks' (Section 3.1), but the exact number of days is not stated; an important architectural choice.

assumptions (4)

domain assumption The S&P100 trading environment can be modeled as a deterministic MDP whose transitions are fully known from historical data.
Section 3.3 sets T to a delta function, ignoring market impact, liquidity, and stochastic market dynamics.
domain assumption FinBERT probabilities and Loughran-McDonald word proportions are valid sentiment signals for news and SEC filings.
Sections 2.1.1 and 2.2.2 treat these as ground-truth features without validation against price impact or downstream prediction.
ad hoc to paper Exponential decay with gamma=0.8 reconstructs unobserved daily sentiment between reports.
The authors acknowledge incomplete news coverage (84/100 tickers, Section 2.2.3); the forward-filling formula (Equation 2) is an ad hoc imputation on which the multimodal state depends.
domain assumption The EIIE policy class from [4] and the state-augmentation scheme from [3] transfer unchanged to S&P100 equities and to additional sentiment channels.
Section 3.6 adopts these architectures wholesale; no ablation shows they are appropriate for this data.

how reviews work

0 comments

Cite this review

Pith. "Pith review of Multimodal Deep Reinforcement Learning for Portfolio Optimization." pith.science (2026). https://pith.science/paper/FNPBHP6D

@misc{pith2026241217293,
  author       = {Pith},
  title        = {Pith review of: Multimodal Deep Reinforcement Learning for Portfolio Optimization},
  year         = {2026},
  howpublished = {\url{https://pith.science/paper/FNPBHP6D}},
  note         = {Machine review of arXiv:2412.17293}
}

read the original abstract

We propose a reinforcement learning (RL) framework that leverages multimodal data including historical stock prices, sentiment analysis, and topic embeddings from news articles, to optimize trading strategies for SP100 stocks. Building upon recent advancements in financial reinforcement learning, we aim to enhance the state space representation by integrating financial sentiment data from SEC filings and news headlines and refining the reward function to better align with portfolio performance metrics. Our methodology includes deep reinforcement learning with state tensors comprising price data, sentiment scores, and news embeddings, processed through advanced feature extraction models like CNNs and RNNs. By benchmarking against traditional portfolio optimization techniques and advanced strategies, we demonstrate the efficacy of our approach in delivering superior portfolio performance. Empirical results showcase the potential of our agent to outperform standard benchmarks, especially when utilizing combined data sources under profit-based reward functions.

Figures

Figures reproduced from arXiv: 2412.17293 by the authors.

**Figure 4.** Frequency distribution of our novel news sentiment scores. [PITH_FULL_IMAGE:figures/full_fig_p006_4.png] view at source ↗

**Figure 13.** Performance of benchmark strategies on test period [PITH_FULL_IMAGE:figures/full_fig_p020_13.png] view at source ↗

Figures from the paper (8 more)

**Figure 14.** Figure 14: Performance of strategies using historical price data with Differential Sharpe reward [PITH_FULL_IMAGE:figures/full_fig_p020_14.png]

**Figure 15.** Figure 15: Performance of strategies using historical price data with Profit reward [PITH_FULL_IMAGE:figures/full_fig_p021_15.png]

**Figure 16.** Figure 16: Performance of strategies using price and SEC data with Differential Sharpe reward [PITH_FULL_IMAGE:figures/full_fig_p021_16.png]

**Figure 17.** Figure 17: Performance of strategies using price and SEC data with Profit reward [PITH_FULL_IMAGE:figures/full_fig_p022_17.png]

**Figure 18.** Figure 18: Performance of strategies using price, SEC, and news data with Differential Sharpe reward [PITH_FULL_IMAGE:figures/full_fig_p022_18.png]

**Figure 19.** Figure 19: Performance of strategies using all price, SEC, and news data with Profit reward [PITH_FULL_IMAGE:figures/full_fig_p023_19.png]

**Figure 20.** Figure 20: Comparison of Best Strategies 23 [PITH_FULL_IMAGE:figures/full_fig_p023_20.png]

**Figure 21.** Figure 21: Comparison of Best Strategies (with rescaled y-axis) [PITH_FULL_IMAGE:figures/full_fig_p024_21.png]

Discussion (0). Continue with ORCID to comment.

Reference graph

Works this paper leans on

14 extracted references · 11 canonical work pages

[1]

Deep Reinforce- ment Learning for Optimal Portfolio Allocation: A Comparative Study with Mean-Variance Optimization

Srijan Sood, Kassiani Papasotiriou, Marius Vaiciulis, and Tucker Balch. Deep Reinforce- ment Learning for Optimal Portfolio Allocation: A Comparative Study with Mean-Variance Optimization

work page
[2]

Deep reinforcement learning for stock portfolio optimization by connecting with modern portfolio theory

Junkyu Jang and NohYoon Seong. Deep reinforcement learning for stock portfolio optimization by connecting with modern portfolio theory. Expert Systems with Applications, 218:119556, May 2023

work page 2023
[3]

Reinforcement-Learning based Portfolio Management with Augmented Asset Movement Pre- diction States, February 2020

Yunan Ye, Hengzhi Pei, Boxin Wang, Pin-Yu Chen, Yada Zhu, Jun Xiao, and Bo Li. Reinforcement-Learning based Portfolio Management with Augmented Asset Movement Pre- diction States, February 2020. arXiv:2002.05780 [cs, q-fin, stat]

arXiv 2020
[4]

A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem, July 2017

Zhengyao Jiang, Dixing Xu, and Jinjun Liang. A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem, July 2017. arXiv:1706.10059 [cs, q-fin] version: 2

arXiv 2017
[5]

Qiang Song, Anqi Liu, and Steve Y . Yang. Stock portfolio selection using learning-to-rank algorithms with news sentiment. Neurocomputing, 264:20–28, November 2017

work page 2017
[6]

MAPS: Multi-agent Rein- forcement Learning-based Portfolio Management System

Jinho Lee, Raehyun Kim, Seok-Won Yi, and Jaewoo Kang. MAPS: Multi-agent Rein- forcement Learning-based Portfolio Management System. In Proceedings of the Twenty- Ninth International Joint Conference on Artificial Intelligence, pages 4520–4526, July 2020. arXiv:2007.05402 [cs]

arXiv 2020
[7]

CRSP daily stocks, 2010-2024

Wharton Data Research Services. CRSP daily stocks, 2010-2024

work page 2010
[8]

SEC.gov | how to read a 10-k/10-q

work page

Show all 14 references

[9]

SEC.gov | EDGAR | company filings
[10]

Daily finaincial news for 6000+ stocks, 2020

Miguel Aenille. Daily finaincial news for 6000+ stocks, 2020. Re- treived 2024-05-05 from https://www.kaggle.com/datasets/miguelaenlle/ massive-stock-news-analysis-db-for-nlpbacktests/data

2020
[11]

Moody and Lizhong Wu

J. Moody and Lizhong Wu. Optimization of trading systems and portfolios. In Proceedings of the IEEE/IAFE 1997 Computational Intelligence for Financial Engineering (CIFEr), pages 300–307

1997
[12]

Reinforcement learning for trading systems and portfolios: Immediate vs future rewards

John Moody, Matthew Saffell, Yuansong Liao, and Lizhong Wu. Reinforcement learning for trading systems and portfolios: Immediate vs future rewards. In Apostolos-Paul N. Refenes, Andrew N. Burgess, and John E. Moody, editors, Decision Technologies for Computational Finance: Pro...
[13]

Bin Li and Steven C. H. Hoi. On-line portfolio selection with moving average reversion
[14]

∞X t=0 γtRt s0 = s # Qπ(s, a) = E π

Li Gao and Weiguo Zhang. Weighted moving average passive aggressive algorithm for online portfolio selection. In 2013 5th International Conference on Intelligent Human-Machine Systems and Cybernetics, volume 1, pages 327–330. 16 Appendix A Reinforcement Learning Overview The r...

2013

Pith tools

Reviewed August 11, 2026 · model on record in the stance chip above.

[1] [1]

Deep Reinforce- ment Learning for Optimal Portfolio Allocation: A Comparative Study with Mean-Variance Optimization

Srijan Sood, Kassiani Papasotiriou, Marius Vaiciulis, and Tucker Balch. Deep Reinforce- ment Learning for Optimal Portfolio Allocation: A Comparative Study with Mean-Variance Optimization

work page

[2] [2]

Deep reinforcement learning for stock portfolio optimization by connecting with modern portfolio theory

Junkyu Jang and NohYoon Seong. Deep reinforcement learning for stock portfolio optimization by connecting with modern portfolio theory. Expert Systems with Applications, 218:119556, May 2023

work page 2023

[3] [3]

Reinforcement-Learning based Portfolio Management with Augmented Asset Movement Pre- diction States, February 2020

Yunan Ye, Hengzhi Pei, Boxin Wang, Pin-Yu Chen, Yada Zhu, Jun Xiao, and Bo Li. Reinforcement-Learning based Portfolio Management with Augmented Asset Movement Pre- diction States, February 2020. arXiv:2002.05780 [cs, q-fin, stat]

arXiv 2020

[4] [4]

A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem, July 2017

Zhengyao Jiang, Dixing Xu, and Jinjun Liang. A Deep Reinforcement Learning Framework for the Financial Portfolio Management Problem, July 2017. arXiv:1706.10059 [cs, q-fin] version: 2

arXiv 2017

[5] [5]

Qiang Song, Anqi Liu, and Steve Y . Yang. Stock portfolio selection using learning-to-rank algorithms with news sentiment. Neurocomputing, 264:20–28, November 2017

work page 2017

[6] [6]

MAPS: Multi-agent Rein- forcement Learning-based Portfolio Management System

Jinho Lee, Raehyun Kim, Seok-Won Yi, and Jaewoo Kang. MAPS: Multi-agent Rein- forcement Learning-based Portfolio Management System. In Proceedings of the Twenty- Ninth International Joint Conference on Artificial Intelligence, pages 4520–4526, July 2020. arXiv:2007.05402 [cs]

arXiv 2020

[7] [7]

CRSP daily stocks, 2010-2024

Wharton Data Research Services. CRSP daily stocks, 2010-2024

work page 2010

[8] [8]

SEC.gov | how to read a 10-k/10-q

work page