Paper: arXiv 2405.01604

Authors: Ashish Anil Pawar, Vishnureddy Prashant Muskawar, Ritesh Tiku

Abstract

Algorithmic trading or Financial robots have been conquering the stock markets with their ability to fathom complex statistical trading strategies. But with the recent development of deep learning technologies, these strategies are becoming impotent. The DQN and A2C models have previously outperformed eminent humans in game-playing and robotics. In our work, we propose a reinforced portfolio manager offering assistance in the allocation of weights to assets. The environment proffers the manager the freedom to go long and even short on the assets. The weight allocation advisements are restricted to the choice of portfolio assets and tested empirically to knock benchmark indices. The manager performs financial transactions in a postulated liquid market without any transaction charges. This work provides the conclusion that the proposed portfolio manager with actions centered on weight allocations can surpass the risk-adjusted returns of conventional portfolio managers.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 3.0/10
  • Quadrant: Lab Rats — high math complexity, limited empirical testing

Why this score: The paper employs advanced DRL architectures (DQN, A2C, DDPG, LSTMs) and discusses MDP formulations, representing significant mathematical complexity. However, the empirical testing is described in a ‘postulated liquid market without transaction costs’ and lacks details on code, specific datasets, or statistical robustness, making it not backtest-ready.

Research Flowchart

  flowchart TD
  A["Research Goal:<br>Develop DRL Portfolio Manager<br>to outperform benchmarks"] --> B["Methodology:<br>Deep Reinforcement Learning<br>DQN & A2C Models"]
  B --> C["Data/Inputs:<br>Historical Market Data<br>Asset Prices & Volatility"]
  C --> D["Computational Process:<br>Environment Simulation<br>Weight Allocation & Short Selling"]
  D --> E["Training:<br>Learning Trading Policies<br>Optimization without Transaction Costs"]
  E --> F["Outcome:<br>Superior Risk-Adjusted Returns<br>vs. Conventional Managers"]
  F --> G["Conclusion:<br>DRL effective for Portfolio Optimization"]

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 reinforced portfolio manager that allocates weights to assets, allowing long and short positions.

Source sentence
In our work, we propose a reinforced portfolio manager offering assistance in the allocation of weights to assets.

Assumes 3

  • Transactions occur in a liquid market with no transaction charges. market
    Source sentence
    The manager performs financial transactions in a postulated liquid market without any transaction charges.
  • The agent must avoid large changes between consecutive portfolio vectors to minimise transaction costs. execution
    Source sentence
    To minimise transaction costs, the portfolio management agent needs to prevent itself from significant changes between consecutive portfolio vectors.
  • The trading problem is modeled as a Markov Decision Process (MDP). model
    Source sentence
    If we take the process to be an MDP, we can assume the problem to be of maximising nature.

Data

  • Crypto-currency market back‑test data crypto-onchain unknown
    Source sentence
    The profitability of the model exceeds all surveyed conventional portfolio selection approaches, as shown in the paper by the results of three back-test studies in a crypto-currency market over different times.
  • 28 equity assets from Yahoo Finance · Yahoo Finance market-prices public
    28 equity assets
    Source sentence
    The data for 28 assets in the portfolio was extracted from Ya…
  • Stock price dataset (2009‑2015 training, 2016‑2017 validation, 2018 test) market-prices unknown
    2009‑2018
    Source sentence
    The dataset is partitioned into a training set that covers the 2009-2015 period, the 2016-2017 validation set and the 2018 stock data test set, respectively.

Recognised sources: Yahoo Finance

Costs

Not modeled A trading or execution result is reported without costs.

Source sentence
The manager performs financial transactions in a postulated liquid market without any transaction charges.

Finds 3

  • The model’s profitability exceeds all surveyed conventional portfolio selection approaches in crypto‑currency back‑tests. performance
    Source sentence
    The profitability of the model exceeds all surveyed conventional portfolio selection approaches, as shown in the paper by the results of three back-test studies in a crypto-currency market over different times.
  • The DDPG‑based model outperformed the Dow Jones Industrial Average with a 25.87% annualized return on test data. performance
    Source sentence
    the Dow Jones Industrial Average by lucratively producing a 25.87 percent annualized return on the test data.
  • The proposed portfolio manager can surpass the risk‑adjusted returns of conventional portfolio managers. performance
    Source sentence
    the proposed portfolio manager with actions centered on weight allocations can surpass the risk-adjusted returns of conventional portfolio managers.

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 14 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive (1 by independent authors)

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.