Energy equities have benefited from higher commodity prices and renewed geopolitical strain in the Middle East, creating a market backdrop where some parts of the sector are outperforming the broader index while providing potential downside cushioning. Within that environment, algorithmic and model-driven selection has isolated less obvious beneficiaries rather than the well-known integrated majors.
As part of a limited-time Flash Sale priced at less than $8 a month, premium subscribers received an AI-backed call on Profrac Holding (NASDAQ: ACDC) that produced a +31.9% return in March. The machine-learning engine behind the selection had flagged the company earlier in the month based on a composite signal that combined momentum metrics, volatility-informed entry timing, and a recovery narrative.
The selection process reflected analysis from more than 150 investor-grade models consolidated into a single machine-learning framework. For ACDC, the models highlighted several attributes:
- Momentum, Volatility, and Recovery Upside - The engine selected Profrac Holding for its recent upward price movement, an entry opportunity created by volatility, and evidence of a potential rebound in growth.
- Three-month and year-to-date moves - The stock has risen approximately 34% over the past three months and is up roughly 27% year-to-date.
- Valuation context - Trading near book value with a Price/Book of approximately 1x, the company’s valuation appeared notable relative to its reported scale of about $2 billion in revenue and around $287 million in EBITDA.
- Recovery backdrop - A tightening pressure pumping market plus an expected activity recovery in 2026 were cited by the model as factors that could support a financial turnaround.
Despite the recent run, the models noted that Profrac still sat at about 46% of its 52-week high, implying that the selection could retain upside potential from that perspective. When macro forces aligned with the company-specific signals the model prioritized, the result for subscribers was a significant monthly gain.
Subscribers who missed ACDC’s move were reminded there were other timely picks in the March list. The AI’s March selections also included several energy-related names that posted material monthly returns:
- Par Pacific Holdings Inc (NYSE: PARR) - +24.5% in March alone.
- PBF Energy Inc (NYSE: PBF) - +22.39% in March alone.
- Marathon Petroleum Corp (NYSE: MPC) - +15.5% in March alone.
- HF Sinclair Corp (NYSE: DINO) - +14.3% in March alone.
The gains in March followed a period of strong performance in January and February, when subscribers captured outsized returns on several non-energy names that at the time were included on the AI’s lists but were later removed as the model updated its view. Examples cited included:
- Ultra Clean Holdings Inc (NASDAQ: UCTT) - +139.56% in the first two months of the year before being dropped from the AI list.
- InnovAge Holding Corp (NASDAQ: INNV) - 63%+ in February alone before removal from the AI list.
- Generac Holdings Inc (NYSE: GNRC) - +65.26% in the first two months of the year before being dropped from the AI list.
The platform reports that since its official AI launch in November 2023, its collective picks have materially outperformed the benchmark. The performance figures presented are: since launch, +179.99% versus +60.12% for the S&P 500, representing a +119.88% outperformance; and in 2026 year-to-date, +12.28% versus -0.90% for the S&P 500, a +13.19% outperformance. These results are stated as recorded from the launch of the underlying AI models.
How the selection process operates is described in the service notes. At the start of each month the AI refreshes strategies with up to 20 stock picks. Those choices reflect a blend of more than 150 established financial models that the machine-learning system compiles using over 15 years of global financial data. Stocks may be added, retained, or removed month-to-month as the model reassesses medium-term growth potential.
For tracking and comparability, each strategy applies equal weighting across selected stocks. That equal-weight framework serves as a benchmark to measure the model’s ability to identify opportunities across the portfolio, although subscribers are not required to replicate that weighting exactly.
The service emphasizes that stock picking remains probabilistic in nature. The model’s process includes exiting positions that no longer meet its criteria, and the results cited include several high-profile winners the model identified before subsequent removal. Promotional subscription calls are highlighted for both app and web users as avenues to access the full lists of monthly picks under the Flash Sale arrangement.
At the tail end of the published materials, the platform notes that proprietary ProPicks AI evaluates companies such as Marathon Petroleum Corp (MPC) and thousands of others monthly using over 100 financial metrics, assessing fundamentals, momentum, and valuation. The product is positioned as an objective, data-driven tool that ranks opportunities by current risk-reward profiles and cites past successful recommendations as examples of model outcomes.
Readers are reminded that prices referenced were accurate at the time of publication and that offers for membership can vary by region. The materials present performance histories and model mechanics but do not guarantee future results.