In a Contest of AI Agents Beating the Market, the Price Mechanism Wins
Markets Are Efficient But Don't Tell Anyone
During the second week of July 2026 in the world of finance, widely repeated headlines announced this bombshell: JP Morgan AI Agents Beat 60/40 Portfolio! What is most remarkable about this news is the number of outlets that picked it up (Bloomberg, Financial Review, Yahoo Finance, msn, AI Chronicle and many others) which correlates perfectly with the story’s insignificance.
First, the headlines are revealing. Every one of these news agencies has confirmed the purpose of all active managers is to beat the market – or beat the market on a risk adjusted basis. The other half of the headline declares that beating a generic, balanced portfolio of stocks and bonds (60/40 strategies) is somehow a standard of excellence.
For the record, JP Morgan Investment Management is one of the premier asset managers in America. And as a pioneer in the development of low cost retail funds and institutional portfolios, they have earned their place as an industry leader. As reported by Financial Review, it makes sense they would experiment with the kind of AI driven innovation that their clients expect:
Researchers at the bank built an array of AI-powered investing agents that shift between stocks and bonds depending on changing market conditions. In back tests spanning the past two decades, the best-performing system topped a traditional 60/40 portfolio – 60 per cent in stocks and 40 per cent in bonds – by 0.7 of a percentage point a year with lower volatility.
Second, for a model to outperform a passive 60/40 strategy by 0.70% per year – one that may increase stock market exposure above 60% in a back-tested environment that avoids weak markets, is not meaningful. Doing it with less volatility does mean something, but how much less? So, to better understand the investment strategy discipline, Yahoo Finance describes it this way:
Using agents powered by models from OpenAI and Anthropic, the JPMorgan team designed a system that classifies the market into four regimes based on growth and inflation: Goldilocks, reflation, stagflation and risk-off.
“Goldilocks” is an economy with decent growth and little price inflation, “reflation” is considered a higher growth and higher inflation economy, “stagflation” is economic stagnation or recession with higher prices, and “risk off” is an equity market that favors defensive, or lower volatility stocks. Again, these are back tested results, as JP Morgan and the reporters made clear - but hindsight is 20/20.
However, consider what is missing: the only growth and price inflation scenario that matters. The secular trend that began in the 1790s - when Treasury Secretary Andrew Hamilton reinstated sound currency in America - is falling real prices and average income that rose geometrically. And as one of millions of examples, consider the 1980 commodities bet between economist Julian Simon and biologist Paul Ehrlich. Prices fell sharply. Prosperity and population rose sharply.
Yet, nearly every economist, strategist, and analyst at every major and minor firm is beholden to the role of government as money creator. It is not. So, if this story is such a non-event, why the hype? According to Stock Backgrounder:
The financial industry is undergoing a structural shift towards integrating advanced AI into core investment strategies, moving beyond automation to autonomous decision support. This trend reflects broader efforts to leverage computational power for more dynamic and responsive market engagement.
Essentially, that is MBA-speak for “markets are efficient but don’t tell anyone.” To the objective investor, there is nothing more responsive and engaged than the price mechanism of free markets. And reading between the lines of all these news reports, the immense difficulty for money managers to consistently outperform has not changed:
“The AI agent can be set up with a process to be empowered to make decisions under uncertainty, producing outperformance vs a reasonable benchmark,” the strategists wrote in a note on Thursday, describing the work as the firm’s first attempt to build an AI system for identifying market regimes.
Are not markets always uncertain? What constitutes a reasonable benchmark? And what about periods of time when markets do not have “regimes” of consistent volatility and sentiment? What is sentiment? There is a lot of ambiguity here. Furthermore, the regimes created for the AI Agent testing are similar to the “factor-based” investing models of a decade ago that never gained traction for any firm not named Dimensional Funds, but I digress.
In summary, capital markets are extraordinarily complex, the price mechanism is the information superhighway, and capital flows to talent. And for the vast majority of individual investors, true success is not achieved with market outperformance. Success is achieved with the defined goals and objective data that give our lives meaning and purpose.
And for that, we need to know our risk capacity and funding status. In fact, The Moneyball Method was using the artificial intelligence of statistical analysis long before AI became a household word, but for a rational purpose: to integrate market elegance with human goals and aspirations!
