Paper: arXiv 2309.15640

Authors: Jakub Michańków, Paweł Sakowski, Robert Ślepaczuk

Abstract

This paper proposes a novel approach to hedging portfolios of risky assets when financial markets are affected by financial turmoils. We introduce a completely novel approach to diversification activity not on the level of single assets but on the level of ensemble algorithmic investment strategies (AIS) built based on the prices of these assets. We employ four types of diverse theoretical models (LSTM - Long Short-Term Memory, ARIMA-GARCH - Autoregressive Integrated Moving Average - Generalized Autoregressive Conditional Heteroskedasticity, momentum, and contrarian) to generate price forecasts, which are then used to produce investment signals in single and complex AIS. In such a way, we are able to verify the diversification potential of different types of investment strategies consisting of various assets (energy commodities, precious metals, cryptocurrencies, or soft commodities) in hedging ensemble AIS built for equity indices (S&P 500 index). Empirical data used in this study cover the period between 2004 and 2022. Our main conclusion is that LSTM-based strategies outperform the other models and that the best diversifier for the AIS built for the S&P 500 index is the AIS built for Bitcoin. Finally, we test the LSTM model for a higher frequency of data (1 hour). We conclude that it outperforms the results obtained using daily data.

Complexity vs Empirical Score

  • Math Complexity: 6.5/10
  • Empirical Rigor: 6.0/10 (capped from 7.0: transaction costs not modelled; see penalty flags)
  • Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 7.0)

Why this score: The paper employs advanced machine learning (LSTM) and time series models (ARIMA-GARCH) with sophisticated ensemble construction, indicating moderate-to-high mathematical complexity, while the study features extensive walk-forward backtesting on multiple asset classes and frequencies from 2004 to 2022, demonstrating high empirical rigor.

Research Flowchart

  flowchart TD
  A["Research Goal: <br>Test Hedging Properties of Algorithmic Strategies"] --> B["Input Data<br>Asset Classes: Equities, Commodities, Crypto<br>Period: 2004-2022"]

  B --> C{"Model Generation<br>Four Algorithmic Types"}
  C --> D["LSTM<br>Deep Learning"]
  C --> E["ARIMA-GARCH<br>Time Series"]
  C --> F["Momentum<br>Price Trend"]
  C --> G["Contrarian<br>Reversion"]

  D & E & F & G --> H["Ensemble AIS Construction<br>Single & Complex Strategies"]
  
  H --> I["Outcome: <br>LSTM outperforms all models<br>Bitcoin is best hedge for S&P 500"]
  B -.->|"Hourly Data Test"| J["High-Frequency Validation<br>LSTM Daily < Hourly Data"]

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 novel approach to hedge portfolios of risky assets by diversifying ensemble algorithmic investment strategies (AIS) rather than single assets, using LSTM, ARIMA‑GARCH, momentum and contrarian models.

Source sentence
This paper proposes a novel approach to hedging portfolios of risky assets when financial markets are affected by financial turmoils. We introduce a completely novel approach to diversification activity not on the level of single assets but on the level of ensemble algorithmic investment strategies (AIS) built based on

Assumes 4

  • Walk‑forward optimization with in‑sample periods of the last n × 365 days (n=1…5) and out‑of‑sample forecasts 1 day ahead. execution
    Source sentence
    walk‑forward optimization: – in‑sample: estimation of the model parameters (LSTM) or optimization of p and q orders (ARIMA‑GARCH) – in the in‑sample period we use the last n × 365 actual days, where n = 1, 2, 3, 4, 5; base case = 3Y
  • The new loss function MADL is appropriate for training the LSTM models. model
    Source sentence
    a new Loss function: MADL, introduced by Michánǩow et al. (2022)
  • Asset returns are generated by the four selected theoretical models (LSTM, ARIMA‑GARCH, momentum, contrarian). model
    Source sentence
    We employ four types of diverse theoretical models (LSTM - Long Short-Term Memory, ARIMA-GARCH - Autoregressive Integrated Moving Average - Generalized Autoregressive Conditional Heteroskedasticity, momentum, and contrarian) to generate price forecasts
  • Trading signals are based on next‑day forecasts and are executed without slippage. execution
    Source sentence
    buy/sell signals definitions based on the next day forecasts

Data

  • Empirical data for assets BTC, GLD, SPX, UNG, ZWF market-prices unknown
    2004-2022 · BTC, gold, S&P 500, energy commodity, soft commodity
    Source sentence
    Empirical data used in this study cover the period between 2004 and 2022.
  • Higher‑frequency (1‑hour) LSTM data market-prices unknown
    Source sentence
    Finally, we test the LSTM model for a higher frequency of data (1 hour). We conclude that it outperforms the results obtained using daily data.

Costs

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

Finds 4

  • LSTM‑based strategies outperform ARIMA‑GARCH, momentum and contrarian models in hedging performance. performance
    Source sentence
    Our main conclusion is that LSTM-based strategies outperform the other models
  • The AIS built on Bitcoin provides the greatest diversification benefit for the S&P 500 ensemble AIS. performance
    Source sentence
    the best diversifier for the AIS built for the S&P 500 index is the AIS built for Bitcoin
  • Using 1‑hour data improves LSTM results compared with daily data. performance
    Source sentence
    We conclude that it outperforms the results obtained using daily data
  • Complex (ensemble) AIS are not more efficient than individual strategies. negative
    Source sentence
    After analyzing the results presented in Table 2 and Table 4 we can assert an unfavorable response to RQ3: Are complex (ensemble) AIS

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 21 related papers). Coverage is partial, so this is neither replication nor refutation.

Cited by 1 paper in the archive

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.