Perpetuals.com Reports Strong Backtest Results for BayesShield AI Trading Engine

August 6th, 2026 3:52 PM
By: Newsworthy Staff

Perpetuals.com's BayesShield AI trading engine shows 380% hypothetical returns in backtest, highlighting the potential of AI-driven trading signals based on successful trader behavior.

Perpetuals.com Reports Strong Backtest Results for BayesShield AI Trading Engine

Perpetuals.com Ltd. (NASDAQ: PDC) has announced the backtest results for version two of its patent-pending BayesShield artificial intelligence engine, which powers its UpsideOnly paper trading platform. The results are significant as they demonstrate the potential of the AI model to generate substantial returns while managing risk, which could have implications for retail traders and the broader fintech industry.

Using the same production code deployed on the live platform, the company replayed approximately 1.07 billion historical order fills over a 375-day period. The backtest generated a hypothetical non-compounded return of 380% with a maximum drawdown of 27% and a return-to-drawdown ratio of 4.4. Starting with $400,000 in simulated capital, the strategy produced hypothetical profits of $1.522 million, ending the test period with $1.922 million. These figures are particularly noteworthy because they were achieved without predicting market prices directly; instead, the AI model identifies consistently successful traders and generates automated trading signals when qualified users independently take similar positions at roughly the same time.

According to the company, the strategy posted positive returns in every calendar quarter, avoided liquidations during the test period, and remained nearly flat during a sharp October 2025 Bitcoin sell-off. This resilience during a volatile period suggests that the AI engine's approach could offer a more stable trading strategy compared to traditional models. However, Perpetuals cautioned that the live trading period remains too limited to draw conclusions about future performance.

The announcement comes as UpsideOnly, launched in May 2026, has grown to more than 225,000 users and recorded $70 billion in simulated notional trading volume across 25 assets. This growth indicates strong market interest in AI-powered trading tools, and the backtest results could further bolster confidence in such platforms. For a detailed look at the full press release, visit this link.

Perpetuals.com Ltd. is a fintech company developing AI-powered trading products and prediction markets, with a global footprint across the United States, Europe, and Asia. Its proprietary trading platform, Kronos X, combines advanced AI and data analysis, trained on billions of trades, to monitor market activity in real time and identify patterns for trading and risk decisions. The platform provides multi-asset coverage with self-clearing blockchain-based settlement and operates with full MiFID II, MiCA, DORA, and EMIR compliance. For more updates on PDC, visit the company's newsroom at https://ibn.fm/PDC.

The implications of these backtest results extend beyond Perpetuals. As AI continues to penetrate financial markets, the ability to replicate successful trader behavior rather than predict market movements could represent a paradigm shift. This approach may reduce the reliance on complex predictive models that often fail during black swan events. Moreover, the strong performance during a Bitcoin sell-off suggests that such AI systems could offer diversification benefits to retail investors.

Nevertheless, it is crucial to approach these results with caution. Backtested performance does not guarantee future results, and the live trading period is still short. The company itself acknowledges this limitation. Investors and traders should consider these findings as part of a broader evaluation of AI-driven trading platforms.

Source Statement

This news article relied primarily on a press release disributed by InvestorBrandNetwork (IBN). You can read the source press release here,

blockchain registration record for the source press release.
;