Blog
Feb 27, 2026
The AI Found the Right Stock. Triangulation With Me and The Arena Found a Better One.
Terminal X’s Agent told me to buy Quanta Services. Forty-four billion dollar backlog, thirteen analyst upgrades in a week, Goldman’s top infrastructure pick. It was the right call. I passed anyway.
This is a story about using two different instances of AI to triangulate a single trade idea. The Agent explores. The Arena Trader executes. When I ran the same question through both and layered in my own judgment, I ended up buying a completely different stock with better risk-reward. It started with a name I’d never heard of.
I was scanning the market at the open on February 26th and noticed a ticker I didn’t recognize trading on unusually high volume: LNT. I had no idea what the company was. So I did the simplest thing possible and asked the Terminal X Agent: what is LNT and why is it trading heavy today?
The Agent is Terminal X’s AI research tool, built for exploration. In Lite mode, it answers market questions in real time, pulling from institutional-grade news, analyst activity, earnings data, and technical signals. It told me LNT was Larsen & Toubro, a major Indian infrastructure conglomerate, and that the volume was being driven by a string of power transmission and substation contracts across the Middle East and India. Trading volume was above 2.3 million shares on the day.
That was interesting on its own, but what caught my attention was the broader pattern. I asked a follow-up: is there a theme here that could surface other good long candidates? The Agent identified critical infrastructure as the thread. Power transmission, grid modernization, and data center energy demand were all driving capital into the same sector. A single aberrant move on a stock I’d never heard of had opened up an entire investment theme.
That’s when I went deeper. I switched the Agent into Deep Research mode, where it spends 60 to 90 seconds reading broker research, public filings, and call transcripts before delivering a structured analysis with specific price targets and cited sources. I asked it: what’s the best stock to buy on this theme?
The Agent came back with Quanta Services (PWR). The case was strong. Record $44 billion backlog. Direct exposure to the AI data center power buildout through the new “Bring Your Own Power” federal policy. Thirteen sell-side upgrades in the same week, Goldman leading at $685.
But PWR was trading at 43x forward earnings after an 18% run in a single month. I spent 12 years trading at Goldman Sachs, and I know this setup: a strong company where the move has already happened. The fundamentals were right. The entry was late.
So I told the Agent: find me the same trade that hasn’t moved yet. It surfaced Primoris Services (PRIM). Record $11.9 billion backlog, 26x forward earnings, and a $6 billion pipeline in natural gas generation feeding data center power demand. Same thesis, roughly 60% of the multiple.
The Arena Trader is a separate instance of AI from the Agent. If you’ve read our previous posts about Terminal X’s trading agents, you know it: the Trader manages a live portfolio with real money, evaluates momentum, order flow, and risk-reward before touching anything, and publishes every rationale in real time. Where the Agent explores the research landscape, the Trader decides whether a setup is actually tradeable right now.
I brought PWR to the Trader first. Without any prompting about my concerns, it passed. RSI above 71, overbought after the surge. Volume running at 66% of average, consolidation rather than fresh buying. The Trader independently confirmed what my gut was telling me.
Then I brought PRIM. The Trader ran its own analysis: RSI at 48, neutral and clean. The 50-day moving average at $147.78 providing structural support. JPMorgan and DA Davidson had both raised price targets to $165-$180. The Trader bought it.
This whole thing started because I saw a stock I didn’t recognize doing something unusual. One question to the Agent led to a theme. The theme led to a top pick. My own experience told me the top pick was priced wrong. A second question found the overlooked alternative. And a completely separate AI confirmed the read and executed the trade.
That’s triangulation: two different instances of AI with different analytical frameworks, plus your own judgment, converging on a single answer. The Agent saw that PWR had the strongest fundamental case. The Trader saw that PRIM had the better entry. I connected the two.
The best trades I made on the Goldman desk worked the same way. A thesis, a second opinion, and a tighter entry than the consensus was watching. Terminal X runs that process in minutes instead of hours, drawing on data that used to require a Bloomberg terminal and a team of analysts.
The AI found the right stock. Triangulation found a better one. Two tools, three rounds, one trade.
by: By Jacob Koenig