For decades, the investment industry has been built on a simple promise: with enough expertise, discipline, and information, skilled professionals can consistently beat the market. That promise has shaped everything from active mutual funds to hedge fund strategies, even as data continues to show that sustained outperformance is the exception rather than the rule.
In 2025, that tension has only intensified. Despite record levels of data availability, computing power, and analytical sophistication, most active managers still fail to outperform broad benchmarks. Against this backdrop, Eldad Tamir, CEO of FINQ, is advancing a more radical proposition: that the next evolution of investing will not be human-assisted AI, but AI-led investing, already expressed through ETFs such as AIUP and AINT.
The performance gap that won’t close
The gap between belief and reality in active investing remains stubborn. According to the latest S&P Dow Jones SPIVA report, 79% of active large-cap U.S. equity funds underperformed the S&P 500 in 2025, one of the weakest years on record for active managers in two decades of data. Over longer horizons, the challenge compounds: roughly 88% of large-cap funds underperform over 15 years, underscoring how persistent the issue has been.
Even short-term improvements in market dispersion have done little to reverse the structural trend. While there are occasional cycles where stock pickers narrow the gap, the long-term trajectory remains largely unchanged: most active strategies fail to justify their fees after costs.
“The whole idea of INDEX investing is based on the merits that humans just cannot process all relevant data in an efficient way, and therefore cannot beat the index in a consistent way,” Tamir says. “Well, that is no longer true. With FINQAI and its relative continuous ranking, we can always find what stocks are top-ranked and what stocks should be sold short or left out in order to do better than the indexes.”
Index investing’s original assumption is breaking
Index funds were built on a simple premise: markets are too complex for humans to consistently process all available information, so broad exposure is the most rational approach. That logic has driven trillions into passive strategies.
But the informational environment has changed. In a typical trading day today, markets digest earnings updates, macroeconomic releases, alternative data, and sentiment signals in real time, far beyond what any individual portfolio manager can track manually.
“A portfolio manager may follow a few dozen companies closely, but it’s very difficult to continuously analyze hundreds of companies across many different data sources at the same time,” Tamir explains. “The AI evaluates financial statements, analyst estimates, news, reports, and public sentiment for all 500 companies in the S&P 500 daily.”
This is where the debate begins to shift. If the original justification for passive investing was information overload, AI introduces a competing thesis: that overload is no longer a human constraint.
AIUP and AINT: two expressions of the same intelligence
FINQ’s approach is already being tested in live market structures through two ETFs built on the same underlying AI ranking system, but expressed differently depending on investor objectives.
AIUP is structured as a concentrated long-only portfolio, designed for investors who still want directional exposure to U.S. equities while replacing discretionary stock selection with systematic AI ranking.
AINT takes a different approach, applying a long-short, market-neutral framework using the same ranking logic, going long higher-ranked companies and short lower-ranked ones.
“The goal was to show that the AI framework is not tied to a single market view or strategy,” Tamir says. “A long-only strategy like AIUP is designed for investors seeking exposure to U.S. equities with systematic stock selection. A market-neutral strategy like AINT uses the same rankings but expresses them differently—going long the higher-ranked companies and short the lower-ranked ones.”
The implication is subtle but important: the innovation is not the ETF wrapper; it is the decision engine underneath it.
From human bias to systematic execution
Even when human managers have access to similar data, execution remains inconsistent. Behavioral finance research has long shown that emotional responses, fear during drawdowns, and overconfidence during rallies can significantly impact returns over time.
Tamir’s critique is direct. “I believe people are bad at making cold, logical decisions,” he says. “They add feelings such as fear and greed. They easily fall into inherited conceptions, and their ‘computing power’ for heavy lifting in online data processing is lousy.”
In contrast, AI systems apply the same rules across all market environments, regardless of volatility or narrative pressure.
“Humans react to greed, fear, headlines, or short-term narratives during crises,” Tamir says. “The AI continues to collect market data, evaluate all companies, and apply the same analytical process regardless of whether markets are calm or under stress.”
Redefining what it means to beat the market
The debate is often framed as active versus passive investing. But the rise of AI introduces a third category: systematic, continuously learning decision engines that do not rely on prediction or intuition, but on scalable, repeatable ranking systems.
That shift matters because even passive investing is no longer purely “do nothing.” It is a rules-based system that has already replaced human discretion. AI, in Tamir’s view, is the next step in that evolution.
“Financial markets generate enormous amounts of data, and technology is simply better suited to analyze that information and apply consistent decision frameworks,” he says. “Human portfolio managers are inevitably influenced by fear, greed, narratives, and incentives. AI systems can process far more information and make decisions systematically without those biases.”
The question, then, is no longer whether AI can outperform humans occasionally. It is whether human-driven decision-making can remain competitive in a system where machines update, rank, and rebalance continuously across thousands of signals.
For Tamir, the direction of travel is already clear. “We are just in the initial stage of the immense opportunity that AI can bring to this market.”