Nobody Panic-Sells Because They're Bad at Math

Every trading forum has the same thread, over and over: someone sells at the bottom, watches the price recover a week later, and posts something like "I knew better, I just couldn't watch it anymore." The replies are always some version of "should've done the math." But that's usually not what happened. Most of these people did the math. They had the number in front of them the whole time. They just couldn't sit with what the number was making them feel. That gap is basically what I spend my time studying. I'm in a finance master's program right now, focused on behavioral finance, and the variable I keep coming back to is math anxiety, specifically how it's different from math ability . Those two things get treated as the same problem in almost every fintech product I've looked at, and I don't think they are. The standard fix in fintech, when users make bad decisions, is always some version of: simplify the interface, add a chart, explain the jargon, reduce friction. It's a reasonable instinct, and for a real chunk of users, it works. But there's research now separating math anxiety from raw numeracy skill, and the anxiety part predicts bad financial behavior almost independently of whether someone can actually do the calculation. People don't avoid checking their portfolio because they can't read a percentage. They avoid it because checking makes them feel something they'd rather not feel. A cleaner dashboard doesn't fix that. It just makes the uncomfortable number easier to read, which isn't the same as making it easier to look at. Here's what actually gets under my skin a bit. The industry clearly already knows emotion drives behavior more than information does. You can see it in the design of pretty much every trading app: confetti animations on a buy, login streaks, leaderboards, a little haptic buzz when a trade goes through. That's not accidental. Someone on a design team understood exactly this dynamic and used it to drive engagement, not decision quality. It works, in the sense that people open the app more and trade more. Whether they trade better is a separate question nobody on a growth team gets measured on. And the older behavioral finance research on overtrading is pretty consistent: more frequent trading tends to correlate with worse returns, not better ones. Gamification didn't reduce the anxiety. It just gave people a faster, more addictive loop to act on it. There's no clean framework for fixing this yet, but a few directions seem better than what exists now. Graduated exposure seems underused, letting people make small, real-money decisions with real but low stakes before they're dropped into a decision that actually matters. Not a demo account with fake numbers, since fake stakes don't produce the same emotional response as real ones. Showing a range instead of a single precise number might help too. A single number implies a certainty nobody making the model actually has. A range at least admits the uncertainty is real, which is more honest than most dashboards currently are. And the least exciting fix, probably the most important one: measuring product success by whether people's decisions get better over time, not by how often they open the app. That's a genuinely hard thing to build and measure. It's also the only metric that matters if the actual goal is helping people make sound decisions, rather than just keeping them logged in.