Latest / The Day Trading Podcast with Fexingo: Active Trading, Charts, and Short-Term Market Strategy / How Day Traders Can Outsmart AI Trading Agents
Transcript
- Lucas: A CNBC story yesterday — July 28 — said startups and brokers are building AI agents designed to trade 24/7. Not just executing algorithms, but agents that can adapt, learn, and place trades while you sleep. Luna: Yeah, I saw that. It's getting traction fast. A few of those firms already have live beta accounts. Lucas: Right. And the natural reaction from a lot of day traders I talk to is — 'Does this mean the edge is gone? How do I compete with something that never blinks?' Luna: It's a fair panic. But I think the answer isn't 'you can't.' It's 'you have to change what you're looking for.' Lucas: Exactly. Look at the current backdrop. The VIX is at 18.40 — not screaming fear, but not complacent either. The Russell 2000 is down 0.4 percent over five days. We've got a market that's grinding, not trending. And in that environment, AI agents are going to feast on obvious patterns — but they still have blind spots. Luna: What are the blind spots? Give me three. Lucas: First: context. An AI agent can read a price move, but it can't read the nuance behind a headline. Take the Visa news — they're cutting 7 percent of staff, partly because AI is reshaping work. That's a long-term structural story. An agent might see the initial dip and short it. A human knows that Visa is streamlining, which could be bullish margins over time. Luna: So you're saying the human can hold a thesis through volatility in a way an agent can't — at least not yet. Lucas: Exactly. Second: catalyst timing. AI agents are great at mean reversion and momentum in liquid names. But they struggle with event-driven plays where the timing is unpredictable — earnings pre-announcements, regulatory surprises, or merger arbitrage that depends on human interpretation of court rulings. Luna: That's a place where a focused day trader with a specialty can still have a real edge. Lucas: Third: capital discipline. Most retail traders blow up because they don't size correctly. But the AI agents being built right now? They're funded by venture capital — they have deep pockets. They can absorb drawdowns. A retail trader with a $50,000 account cannot afford to trade like an agent. That means you have to be more selective. Luna: Selective about what, exactly? Fewer setups per day? Lucas: Fewer, but higher-conviction. If an agent is taking 50 trades a day, it's going to be right on 55 percent of them. That's fine for them. You need to take 5 trades with 70 percent win rate. You can't chase every wiggle. Luna: That brings up a related point — the Russell divergence. Small caps are lagging. The Russell is down 0.4 percent while the S&P is basically flat. That's a signal that money isn't rotating into risk. AI agents might see that and short the Russell. But a human can ask: is this a fundamental slowdown or just a tax-loss harvesting quirk? Lucas: Exactly. And that kind of question is your edge. The AI doesn't know it's late July — it doesn't know about seasonal patterns in small caps. You do. Luna: You keep coming back to context. It's really the theme of this whole conversation. Lucas: It is. And honestly, this is the kind of conversation that makes me glad we do this show every day. A couple of dollars a month — for listeners who find value in these breakdowns — that's genuinely what keeps these episodes coming. If you've gotten something out of it, buy me a coffee dot com slash fexingo. It's a small ask, but it makes a real difference. Luna: Yeah, it's true. And we don't have ads, so listener support is the engine. Just a thought. Lucas: So back to the AI agents. The other thing to watch is the VVIX — it's at 98.51, up 5.6 percent over five days. That's not extreme, but it's creeping up. That tells me options market makers are pricing in more tail risk. And that's exactly the kind of environment where human judgment on tail risk can pay off, because agents tend to ignore low-probability events. Luna: So you're saying in a market where the VIX is moderate but the VVIX is rising, the real opportunity might be in the wings? Lucas: Exactly. Put spreads, broken-wing butterflies — structures that profit if nothing happens but also have a tail if volatility spikes. An agent won't touch those because they're too complex for the strategy they're trained on. Luna: What about the headline on Visa cutting 7 percent? That's a specific catalyst a human can trade. Lucas: Great example. Visa's stock gapped down initially, then recovered. An AI might have shorted the gap and gotten squeezed. A human who read the full story — AI streamlining, long-term margin expansion — could have bought the dip. That's the difference. Luna: So the playbook is: use your humanity — your ability to read nuance — and don't try to beat the bots at their own game. Lucas: Exactly. Trade like a human. Use filters agents can't replicate: relationship networks, regulatory knowledge, seasonality, earnings call tone. That's where the next decade of edge lives. Luna: And maybe keep your trade counts low and your conviction high. Lucas: That's the formula. As AI gets better, the boring, patient human with a thesis might actually become more valuable, not less.