Programmable incentives that allow independent trading agents to earn only when portfolios rise will create a fairer market for retail customers, explains Naja.
By Saad Naja|Edited by Betsy FarberUpdated May 28, 2026, 2:29 p.m. Published May 28, 2026, 2:13 p.m. 4 min readMake preferred on (Pakin Songmor/Getty Images)In just a few weeks, Anthropic introduced new financial agents, Circle initiated nanopayments, MoonPay rolled out a debit card for agents, and Gemini launched agentic trading, marking the arrival of the agentic finance era. Despite the novelty of these products, the fundamental business model remains unchanged. Exchanges and brokerages profit when customers trade frequently, and the data clearly shows how this impacts customer portfolios. Ultimately, the implementation of agentic systems has outpaced changes in incentives.
The troubling incentives exchanges prefer you overlook
This conflict is inherent in the industry. Brokerages and exchanges do not require clients to succeed; rather, they benefit from continuous trading. Crypto exchanges and neobrokers have made trading faster, cheaper, and more addictive. The reality is that banks gain from customer retention, exchanges earn from customer trading, and AI models profit from user prompts. Trustworthy agents, however, exist outside these three categories. An independent agent compensated solely when a customer’s portfolio increases poses a challenge to the existing incentive framework of brokerages and exchanges.
In reality, zero-commission trading incurs hidden costs. In 2025, U.S. market makers disbursed over $4.9 billion for order flow in U.S. equities and options, a rise from about $3.8 billion in 2021 among the twelve largest U.S. brokerages. This principle also applies to the crypto market, where the derivatives volume in Q1 2026 hit roughly $18.6 trillion, accounting for 70% of global crypto trading, with perpetual contracts overtaking spot trading. The economics of exchanges favor rapid trading over thoughtful decision-making.
At its peak, Robinhood derived over 75 percent of its revenue from payment for order flow (PFOF), the unseen foundation of "free" trading, where market makers compensate brokers to direct customer orders. Every broker utilizing this incentive model relies on frequent trading, which can detract from long-term gains.
Advisory services aren’t any better. Robo-advisors charge 0.25 percent annually on assets, regardless of account performance. In contrast, human advisors often charge around 1 percent, based on the principal even during down years. This extraction is intentionally embedded in the model: advisors are compensated even when clients incur losses.
Reducing exchange friction makes poor trades easier to repeat
The harsh reality is that exchanges prefer customers to trade frequently instead of winning. When retail investors incur losses, exchanges still profit. Research from PiP World indicates that 74% to 89% of retail traders lose money. Platforms impose charges at every stage, and an AI-powered exchange might simply expedite a return to losing trades.
The April 14 SEC authorization of FINRA's removal of the Pattern Day Trader rule eliminated the $25,000 minimum-equity barrier. This reduction in friction leads to more trading, generating increased order flow, which ultimately benefits brokers, regardless of the customers' profit and loss (P&L) outcomes.
Introducing AI agents that prioritize customer P&Ls
The solution to this detrimental cycle for retail traders is the agent designed to do what the current exchange structure avoids: trade less, reduce size, wait, and shield customers from their worst instincts. In volatile markets, the best strategy often involves abstaining from poor trades and minimizing exposure before emotions take over. Maintaining discipline in the face of market reactions is challenging for exchanges to promote, as it reduces order flow. An agent that profits by safeguarding customer P&Ls disrupts the prevailing incentive model.
The upcoming battleground is who profits from the agents' order flow
Regulators are tightening the old "free trading" framework. The EU's PFOF ban will come into effect on June 30, 2026, eliminating the revenue stream behind "free" trades for neobrokers in Germany and Austria. Trade Republic, a European savings platform, has already identified an alternative path to secure a BaFin license for internalizing order flow.
While traditional finance scrambles to address these changes, crypto innovators are racing to develop on-chain frameworks for AI agents. In markets characterized by narrow spreads, fragmented liquidity, and rapid execution, agents utilize nanopayment systems such as Circle's protocol. Gas-free trading on perpetual DEX Hyperliquid reduces friction, yet maker-taker fees still exist. The real challenge ahead is not merely about removing friction but determining who profits as agents engage in high-frequency trading across these seamless networks.
Independent programmable agents serve as superior intermediaries
Exchanges and brokers have spent years profiting from customers who trade more while understanding less, absorbing minor costs that often go unnoticed. Any agent developed by an exchange will carry the same incentives as that exchange. Would an exchange willingly create an agent that directs trades through a less expensive competitor? Likely not.
Conversely, an independent agent has a singular focus: to enhance and safeguard the customer’s portfolio, directing trades to where they yield the best results for the client. Programmable incentives embedded in smart contracts align the agent’s compensation with portfolio growth. Customers can track where their money goes, verifying the agent's compensation, timing, and rationale. With independent agents, clients retain more of the value that previously leaked to exchanges through order flow, spread markups, and idle cash interest.
The agent is rewarded for disciplined trading rather than incessant trading. It can trade frequently when signals are strong, reduce exposure when risks rise, and refrain from trading during market noise. The first on-chain agentic platform to demonstrate this alignment will provide retail investors with a fairer counterpart, whose financial interests finally align with theirs.
Artificial IntelligenceNote: The views expressed in this column are those of the author and do not necessarily reflect those of CoinDesk, Inc. or its owners and affiliates.
More For You
A bipartisan bridge to the future: Why the Senate must finish the job on digital Assets
By Ji Hun Kim|Edited by Betsy FarberMay 27, 2026The recent markup of the Clarity Act has shown that there is momentum for regulation. It is crucial for Congress to advance the bill to set rules that this generation needs and a framework for future generations, as urged by Kim.
Latest Crypto News