Perpetuals.com Powers BayesShield(TM) AI with NVIDIA Hardware to Block 92% of Losing Retail Trades in Major Backtest Victory

Rhea-AI Summary

Perpetuals.com (Nasdaq:PDC) announced BayesShield™ AI completed a pilot backtest filtering 92% of losing retail trades in BTC-USD perpetual futures over a 12-month retail dataset. The company is scaling production using NVIDIA hardware and training on >11.7 billion retail trades and >1 million user histories, targeting a broader commercial rollout in H2 2026.

The platform will integrate across the firm's unified stack and a CySEC-authorized MiFID II multilateral trading facility, aiming to protect retail traders and support product issuers.

Positive

• Backtest filtered 92% of losing retail trades
• Training dataset: >11.7 billion retail trades
• Training dataset: >1 million unique user histories
• Scaling on NVIDIA hardware for production
• Planned commercial rollout in H2 2026

Negative

• Result based on backtesting of historical data, not live trading proof
• Retail derivatives segment faces increasing global regulatory scrutiny

News Market Reaction

-10.20% 2.7x vol

10 alerts

-10.20% News Effect

+37.9% Peak Tracked

-26.1% Trough Tracked

-$2M Valuation Impact

$14M Market Cap

2.7x Rel. Volume

On the day this news was published, PDC declined 10.20%, reflecting a significant negative market reaction. Argus tracked a peak move of +37.9% during that session. Argus tracked a trough of -26.1% from its starting point during tracking. Our momentum scanner triggered 10 alerts that day, indicating notable trading interest and price volatility. This price movement removed approximately $2M from the company's valuation, bringing the market cap to $14M at that time. Trading volume was elevated at 2.7x the daily average, suggesting increased selling activity.

Data tracked by StockTitan Argus on the day of publication.

Key Figures

Losing trades blocked: 92% Backtest period: 12 months Retail trades dataset: Over 11.7 billion trades +5 more

8 metrics

Losing trades blocked 92% Portion of losing retail BTC-USD-PERP trades filtered in backtest

Backtest period 12 months Historical retail BTC perpetual futures data analyzed

Retail trades dataset Over 11.7 billion trades Individual retail trades across multiple years of activity

User trading histories More than 1,000,000 users Unique user trading histories used for behavioral patterns

Price move -10.52% PDC 24h price change prior to this news

52-week range high $7.96 52-week high vs current price $4.51

52-week range low $3.776 52-week low vs current price $4.51

Market cap $15,374,873 Market capitalization prior to the news

Market Reality Check

Price: $3.96 Vol: Volume 70,007 is below th...

low vol

$3.96 Last Close

Volume Volume 70,007 is below the 20-day average of 294,018 (relative volume 0.24). low

Technical Price at $4.51 is trading below the 200-day MA $5.89 and 43.34% under the 52-week high.

Peers on Argus

No peers with momentum data today; the -10.52% move in PDC appears stock-specifi...

No peers with momentum data today; the -10.52% move in PDC appears stock-specific rather than part of a sector rotation.

Previous AI Reports

1 past event · Latest: Jan 22 (Positive)

Same Type Pattern 1 events

Date	Event	Sentiment	Move	Catalyst
Jan 22	AI platform coverage	Positive	+4.7%	Media feature on AI-powered derivatives and trading infrastructure.

Pattern Detected

Limited AI-tag history: the prior AI headline saw a positive 4.68% next-day move, unlike today’s decline.

Recent Company History

This announcement extends Perpetuals.com’s AI narrative, following a prior AI-focused feature on Jan 22, 2026 that coincided with a 4.68% gain. That earlier coverage highlighted AI-powered derivatives and blockchain-based settlement. More broadly, January news included the acquisition and rebrand to Perpetuals.com and a launch of quantum-resilient custody vaults, both accompanied by positive price reactions. Today’s BayesShield™ AI pilot results build on these themes of AI risk analytics and infrastructure expansion.

Historical Comparison

AI

+4.7 %

Average Historical Move

Historical Analysis

Past AI-tagged news saw an average 4.68% gain, while today’s AI pilot update precedes a -10.52% move, marking a sharp divergence.

Typical Pattern

AI news has progressed from general platform coverage to detailed BayesShield™ pilot performance and dataset scale.

Market Pulse Summary

The stock dropped -10.2% in the session following this news. A negative reaction despite strong pilo...

Analysis

The stock dropped -10.2% in the session following this news. A negative reaction despite strong pilot metrics fits a divergence from past AI news, which averaged a 4.68% gain versus today’s -10.52% move. The stock traded well below its 200-day MA $5.89 and 43.34% under its 52-week high, suggesting existing weakness. Investors may have focused on execution risk, commercialization timelines, or broader crypto market pressures when interpreting the BayesShield™ announcement.

Key Terms

perpetual futures, blockchain infrastructure, machine learning, derivatives, +4 more

8 terms

perpetual futures financial

"in the BTC-USD-PERP perpetual futures market, based on backtesting"

Perpetual futures are contracts that let investors bet on an asset’s price without a set expiration date, combining the ability to use borrowed funds with a recurring small payment or receipt that keeps the contract price close to the actual market price. They matter because they offer continuous, high-leverage exposure and deep liquidity—useful for quick trading and price discovery—but also raise the risk of forced exits and amplified losses, so they change how investors manage risk.

blockchain infrastructure technical

"combining blockchain infrastructure and artificial intelligence to transform"

Blockchain infrastructure is the underlying technology — the networks, software, and data storage — that records and moves digital ownership and transactions in a secure, tamper-resistant way. Like roads and pipes for a city, it determines how fast, cheap and safe digital activity runs; investors watch it because its performance, costs, security and regulatory posture directly affect a project’s usefulness, adoption potential and long‑term value.

machine learning technical

"validated the Company's proprietary machine learning models by analyzing"

Machine learning is a set of computer programs that learn patterns from large amounts of data and improve their predictions or decisions over time, like a recipe that gets better each time it’s adjusted based on taste tests. For investors it matters because these systems can speed up analysis, spot trends or risks humans might miss, automate routine work, and potentially create competitive advantages or cost savings that affect a company’s performance.

derivatives financial

"retail losses in leveraged derivatives markets are predictable and"

Derivatives are financial contracts whose value depends on the price or performance of another asset, such as a stock, bond, commodity, currency or interest rate. Investors use them to hedge against risk, to speculate on future price moves, or to gain exposure without owning the asset — like buying insurance or placing a leveraged bet — so they can both protect portfolios and magnify gains or losses, affecting risk and market liquidity.

options writers financial

"help product issuers-such as options writers, market makers, and"

Options writers are investors or traders who create and sell options contracts, taking on a legal obligation to buy or sell a stock at a set price if the buyer chooses to exercise the contract. They receive an upfront payment (premium) for taking that risk, similar to an insurance company collecting a premium in exchange for covering a potential claim. For investors this matters because writing options can generate steady income but also caps upside or exposes the writer to significant losses, so it directly affects portfolio return and risk.

market makers financial

"issuers-such as options writers, market makers, and derivatives platforms"

Market makers are firms or traders who always stand ready to buy and sell a stock by posting continuous bid and ask prices, acting like a store that both sells and buys the same product so customers can trade anytime. They matter to investors because they keep trades flowing and determine the bid-ask spread — which affects how quickly orders execute and how much investors pay or receive when buying or selling shares.

cross-chain infrastructure technical

"Ledgera cross-chain infrastructure, PerpetualPay.net, and tokenization services"

Cross-chain infrastructure is the technology that lets different blockchains communicate and transfer assets or data between each other, like a highway interchange or translator connecting separate networks. For investors, it matters because it can increase liquidity and use cases by making tokens and information accessible across platforms, but it also introduces technical and security risks that can affect asset value and market confidence.

multilateral trading facility regulatory

"CySEC-authorized MiFID II-licensed Multilateral Trading Facility, with a broader"

A multilateral trading facility (MTF) is an organized electronic venue where multiple buyers and sellers can trade financial instruments like stocks, bonds or derivatives, run by a firm other than a national stock exchange. Think of it as an alternative marketplace or online auction house that matches orders from many participants; it matters to investors because MTFs can offer different pricing, faster execution, lower costs or access to trading hours and instruments that traditional exchanges may not provide.

AI-generated analysis. Not financial advice.

TOKYO, JP / ACCESS Newswire / February 3, 2026 / Perpetuals.com Ltd. (Nasdaq:PDC) ("Perpetuals.com" or the "Company"), a regulated financial technology company combining blockchain infrastructure and artificial intelligence to transform digital asset trading, today announced the successful completion of a BayesShield™ AI pilot program(the "BayesShield™ AI Pilot Program"). The pilot has demonstrated the system's ability to filter out 92% of losing retail trades in the BTC-USD-PERP perpetual futures market, based on backtesting against a full year of actual retail trading data.

The BayesShield™ AI Pilot Program validated the Company's proprietary machine learning models by analyzing historical retail trading behavior in the Bitcoin perpetual futures market. When applied to 12 months of real retail trade data, the AI system successfully identified and would have filtered 92% of trades that ultimately resulted in losses for retail participants.

"These pilot results exceed our expectations and validate years of research into retail trading behavior," said Patrick Gruhn ("Mr. Gruhn"), the CEO and the founder of Perpetuals.com. "The data show that the vast majority of retail losses in leveraged derivatives markets are predictable and preventable. BayesShield™ represents a paradigm shift in how trading platforms can protect their users while maintaining market integrity."

Amid the current downturn in cryptocurrency prices, which has tested trader resilience across the board with sharp price corrections, the BayesShield™ AI Pilot Program demonstrates exceptional real-world utility. For example, its predictive intelligence identifies and blocks the majority of losing trades before execution, providing critical protection in turbulent markets while reinforcing crypto's fundamental robustness and long-term growth potential.

The BayesShield™ AI Pilot Program analyzes the specific profit and loss profiles of individual traders, identifying patterns in collective market sentiment to calculate win/loss probability in real-time. The system is designed to help product issuers-such as options writers, market makers, and derivatives platforms-optimize hedging strategies while providing enhanced protection for retail participants.

"The retail derivatives market has historically operated on a model where the majority of participants lose money," continued Mr.Gruhn. "BayesShield™ fundamentally changes this dynamic. By predicting which trades are likely to fail before they execute, we can create trading environments that are both more ethical and more sustainable for all participants."

Building on the pilot's success, Perpetuals.com has commenced full-scale development of the BayesShield™ AI Pilot Platform. The production system is being trained on NVIDIA hardware on one of the largest retail trading datasets ever assembled:

• Over 11.7 billion individual retail trades spanning multiple years of market activity

• More than 1 million unique user trading histories providing statistically significant behavioral patterns

• Cryptocurrency spot markets including Bitcoin, Ethereum, and major altcoins

• Cryptocurrency derivatives, including perpetual futures, options, and leveraged products

• Traditional stock markets, enabling cross-asset pattern recognition and model validation

Protecting Retail Traders Through Predictive Intelligence

The announcement coincides with the successful completion of the Company's technical integration following its business combination with EarlyWorks Co., Ltd. This integration consolidated technology stacks, operational infrastructure, and data systems into a unified platform. The redesigned website at perpetuals.com now serves as the central hub for all products and services, including the Kronos X^®exchange suite, Perpetuals Vault custody solutions, Ledgera cross-chain infrastructure, PerpetualPay.net, and tokenization services, reflecting the Company's expanded capabilities and unified brand identity.

The BayesShield™ AI Pilot Program positions Perpetuals.com to address systemic inefficiencies and ethical concerns in retail derivatives trading-a segment that has faced increasing regulatory scrutiny globally due to high retail loss rates.

The Company plans to deploy the BayesShield™ AI Pilot Program across its regulated trading infrastructure, including to a third-party, CySEC-authorized MiFID II-licensed Multilateral Trading Facility, with a broader commercial rollout planned for the second half of the year. Institutional partners, trading platforms, and product issuers interested in early access or integration discussions can contact the Company through perpetuals.com.

About Perpetuals.com Ltd.

Perpetuals.com (NASDAQ:PDC) is a financial technology company that combines blockchain infrastructure and artificial intelligence to transform digital asset trading. The Company develops and operates Kronos X^®, a proprietary, multi-asset exchange platform and blockchain-based settlement solution fully compliant with European regulations, including MiCA, DORA, and EMIR. The Company provides financial market infrastructure as a service from Equinix FR2 in Frankfurt, Germany, alongside Eurex and Xetra, enabling clients to operate 24/7 trading of crypto spot, derivatives, tokenized securities, and structured products.

The Perpetuals.com team pioneered regulated tokenized financial products, including Pre-IPO contracts for Coinbase, Airbnb, and Robinhood-as reported by Forbes-as well as tokenized stocks traded on major exchanges. Building on machine-learning analysis of millions of retail trade transactions, the Company has developed AI-powered risk intelligence designed to analyze trading patterns in real-time.

For more information, visit perpetuals.com.

Trademark Information: Kronos X^® is a registered trademark in the European Union under filing number 019097099 and a pending trademark registration in the United States. Perpetuals.com™ is the subject of a pending trademark application in the European Union under filing number 019186468. Ledgera™ is pending trademark registration in the United States and Europe. PerpetualsPay.Net™ is a registered trademark in the European Union and the United Kingdom. BayesShield™ is a pending trademark registration in the European Union and the United States.

Forward-Looking Statements: This press release contains forward-looking statements as defined within the meaning of the Private Securities Litigation Reform Act of 1995, including, but not limited to, statements set forth in the Company's filings with the Securities and Exchange Commission. Words such as "expect", "will", "positions", "advancing", "aligning", and other similar expressions may indicate forward-looking statements, though not all forward-looking statements contain such words. These statements reflect the Company's current view with respect to future events, are subject to risks and uncertainties that could cause actual results to differ materially, including regulatory approvals, completion of the announced transaction, market conditions, and risks detailed in the Company's SEC filings, and are necessarily based upon a number of estimates and assumptions that, while considered reasonable by the Company, are inherently subject to significant business, economic, competitive, political and social uncertainties, and contingencies. Should one or more of these risks or uncertainties materialize, or should the assumptions set out by the Company underlying those forward-looking statements prove incorrect, actual results may vary materially from those described herein. Individuals are cautioned that forward-looking statements are not guarantees of future performance and accordingly investors are cautioned not to put undue reliance on forward-looking statements due to the inherent uncertainty therein. These forward-looking statements are made as of the date of this press release and the Company does not intend, and does not assume any obligation, to update these forward-looking statements, except as required by law.

Media Contact

Perpetuals.com Ltd.
Email: press@perpetuals.com
Website: www.perpetuals.com

Investor Relations Contact

ir@perpetuals.com

SOURCE: Perpetuals.com

View the original press release on ACCESS Newswire

FAQ

What did Perpetuals.com (PDC) report about BayesShield™ on February 3, 2026?

Perpetuals.com reported a BayesShield™ pilot backtest that filtered 92% of losing retail trades in BTC-USD perpetuals. According to the company, the result derives from a 12-month backtest on real retail trading data and validates their ML models.

How large is the dataset Perpetuals.com is using to train BayesShield™ for production?

Perpetuals.com is training BayesShield™ on a dataset of over 11.7 billion retail trades and more than 1 million user histories. According to the company, this spans crypto spot, derivatives, and traditional markets for cross-asset validation.

When does Perpetuals.com plan a broader commercial rollout of BayesShield™ (PDC)?

The company plans a broader commercial rollout in the second half of 2026. According to Perpetuals.com, initial deployment targets its regulated trading infrastructure and a CySEC-authorized MiFID II multilateral trading facility.

What technology is Perpetuals.com using to scale BayesShield™ for production?

Perpetuals.com is training and scaling BayesShield™ using NVIDIA hardware for production environments. According to the company, NVIDIA systems support model training on one of the largest assembled retail trading datasets.

How will BayesShield™ be deployed across Perpetuals.com's products and partners (PDC)?

Perpetuals.com will integrate BayesShield™ into its unified platform, including the Kronos X exchange suite and custody services. According to the company, deployments include internal infrastructure and third-party regulated trading venues.

Does Perpetuals.com say BayesShield™ results reflect live trading performance?

The announced results reflect a backtest on historical retail trades, not demonstrated live performance. According to Perpetuals.com, the pilot validated models using past retail behavior and supports further production development.