Paper: arXiv 2309.09094

Authors: S. M. Masrur Ahmed

Abstract

Backtest is a way of financial risk evaluation which helps to analyze how our trading algorithm would work in markets with past time frame. The high volatility situation has always been a critical situation which creates challenges for algorithmic traders. The paper investigates different models of sizing in financial trading and backtest to high volatility situations to understand how sizing models can lower the models of VaR during crisis events. Hence it tries to show that how crisis events with high volatility can be controlled using short and long positional size. The paper also investigates stocks with AR, ARIMA, LSTM, GARCH with ETF data.

Complexity vs Empirical Score

  • Math Complexity: 6.0/10
  • Empirical Rigor: 6.0/10 (capped from 6.5: transaction costs not modelled; no out-of-sample test; see penalty flags)
  • Quadrant: Holy Grail — high math complexity, high empirical rigor (effective empirical rigor capped from 6.5)

Why this score: The paper combines several advanced mathematical models (Kalman Filter, GARCH, ARIMA, LSTM) and statistical tests (Kolmogorov-Smirnov) with an explicit backtesting framework, using specific financial data (ETFs) and risk metrics (VaR).

Research Flowchart

  flowchart TD
  A["Research Goal<br>Assess Position Sizing Strategies<br>in High Volatility Markets"] --> B["Data & Models<br>Stocks/ETF Data + GARCH, ARIMA, LSTM"]
  B --> C["Methodology<br>Backtesting & VaR Simulation"]
  C --> D{"Computational Process"}
  D --> E["Short Position Sizing"]
  D --> F["Long Position Sizing"]
  E & F --> G["Risk Mitigation Analysis"]
  G --> H["Key Findings<br>Optimized sizing reduces VaR<br>during crisis events"]

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 developing sizing models for financial trading and backtesting them in high‑volatility situations to lower Value‑at‑Risk during crisis events.

Source sentence
The primary aim of this study is to develop sizing models for financial trading and backtest to high volatility situations to understand how sizing models can lower the VaR during crisis events on stock and ETF market.

Assumes 2

  • The Variance‑Covariance VaR method assumes returns are normally distributed. model
    Source sentence
    the Variance‑Covariance method has a major drawback for VaR estimation as the model assumes normal distribution for returns on investment.
  • Kalman filter model assumes a constant regression coefficient can represent systematic risk. model
    Source sentence
    If constant regression coefficient does not work then the best possible alternative model will be the Kalman filter model for estimating systematical risks.

Data

  • Yahoo Finance SPY dataset · Yahoo Finance benchmark public
    1950-01-01 to 2020-02-28 · SPDR S&P 500 (SPY)
    Source sentence
    For the research purposes of this paper, the dataset was collected from Yahoo Finance, covering the period from January 1, 1950, to February 28, 2020. The benchmark for the dataset was SPDR S&P 500 which as known as SPY.

Costs

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

Finds 3

  • During high‑volatility periods, short positions tend to generate higher returns than long positions. performance
    Source sentence
    During periods of high volatility, it becomes evident that shorting stocks tends to yield higher returns.
  • The Kalman filter methodology performed poorly under high volatility, relying heavily on long positions and failing to meet predefined parameters. negative
    Source sentence
    This approach did not perform well when confronted with high levels of volatility. The Kalman filter method appeared to rely heavily on long positions, and it struggled to meet predefined parameters.
  • The Variance test delivered the most promising returns among backtesting strategies, though it exhibited higher volatility and depended mainly on long positions. performance
    Source sentence
    In contrast, among all the backtesting strategies employed, the Variance test demonstrated the most promising results in terms of returns. It displayed resilience even in high‑volatility scenarios. However, it’s worth noting that the Variance test exhibited higher volatility compared to other methods. This strategy pre

Checked by others

No other paper in the archive has been found to support or contradict these findings (compared with 15 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.