# The 90% nobody builds for

*Trading is mostly psychology. So why does every tool optimize the part that matters least?*

There's an old line that trading is 90% psychology, 9% data, and 1% luck. I believe it. My worst decisions with a stock never came from missing a number — they came from fear, from FOMO, from a headline that hijacked my afternoon and the itch to just *do something*.

Which is the part most tools get backwards: they sharpen the 9% and quietly wreck the 90%. Blinking red arrows, alerts engineered to spike your pulse, "🔥 trending now," a feed built to keep you reacting. All of it acts on your psychology — usually to make you trade more, not to help you decide better.

I didn't want that anywhere near my 90%. I was building Pip for my own portfolio, and what I wanted was the opposite: not something that manufactures urgency, but a calm, reliable, personalized co-pilot that reads back the reality that actually matters to me — each stock measured against the reason I bought it — and then gets out of the way.

*(Honest aside, up front: Pip is not advice. It never tells you what to do, and nothing here is a prediction. It's a sourced account of what happened and why it might matter to the reason you gave for owning a stock. The call is always yours.)*

This piece is about the calm half — what Pip actually reads, and the one signal that matters more than the rest. *(For the full map of what moves a stock — the market itself, earnings expectations, and more — start with [the plain-English guide](/learn/what-moves-a-stock-you-own).)* The harder half — making sure the AI doing the reading doesn't make things up — is [the next one](/learn/how-we-keep-the-ai-honest).

## The five signals — and why each one

No single feed tells you what's going on with a company. Each has a blind spot the others cover:

- **Price & market moves.** The tripwire. A move tells you *something* happened — but price alone can't say *what*, or whether it's the company or just the whole market having a rough day. So Pip treats price as a question, never an answer.
- **Company news.** The human-readable "what." Useful, but noisy, and often written to spin. Good for context, weak as ground truth.
- **SEC filings (8-K, 10-Q).** The primary source — material events in the company's own words, on the record. The closest thing to ground truth an outsider gets. *(How to read them yourself: [the filings guide](/learn/filings-and-events-that-move-a-stock).)*
- **Insider transactions.** What the people closest to the company are doing with their own shares — filed, dated, public. Facts, not motives. *(What they do and don't tell you: [the insider guide](/learn/what-insider-trades-signal).)*
- **Earnings, as filed.** The scheduled moments, reported the way the company files them — not adjusted against anyone's expectations.

Read together, they turn "the stock is down 4%" into "down 4%, here's the 8-K it filed this morning, and here's the link."

But even all five, read on a single morning, only give you a snapshot — and a snapshot can quietly mislead you. Which is why the signal that matters most isn't really in that list at all.

## Zoom out: the trend is usually the real story

Look at an exponential curve up close and it looks like a straight line — steady, unremarkable, nothing to see. Zoom out, and you realize it was bending the whole time.

Stocks do this constantly. A run of small, unalarming days is a straight line at the daily scale and a steep curve at the quarterly one. The move that matters is often the one too gradual to notice up close — which means a briefing that only ever looks at *today* is looking through the exact lens that hides it.

So Pip doesn't just read the pixel; it reads the curve. Every briefing places today's move inside the trajectory it belongs to — a blip in a flat stretch, the continuation of a months-long climb, or the first crack in something that had been steady. The shape is the context that makes a single day mean anything.

Two honesty notes, because trend is the signal it's easiest to abuse:

- **When the trend is the story and nothing explains it, Pip says so.** Sometimes a stock has moved for weeks with no filing, no news, no clean reason on the record. Pip flags that as exactly what it is — an unexplained move worth a look — instead of inventing a cause to sound smart.
- **Pip shows you the shape; it does not draw it forward.** That a curve has been bending is a fact about the past. Where it goes next is not — and Pip won't pretend otherwise. The trend is there so you're not fooled about *what's actually happening now*, never to tell you what happens next.

That's the difference between trend as insight and trend as a crystal ball. Pip only does the first.

## The catch

All of this — five signals, the real trend, each measured against your own reasons — is only worth anything if you can trust the thing doing the reading. And "an AI that reads your stocks" *should* make you suspicious, because language models are fluent, confident, and perfectly happy to make things up.

That's the harder half of the problem, and the half I care about most. It's the next piece: **[The AI that shows you the source behind every line →](/learn/how-we-keep-the-ai-honest)**

If you want a calm morning read on the stocks you own — built for your 90%, not against it — [Pip is opening to a small first group](https://pipthebot.com/#join).

Source: https://pipthebot.com/learn/the-90-percent-nobody-builds-for/

Pip is for research and education only — not investment advice, and not a recommendation to buy, sell, or hold any security. Pip reflects your own stated reasons against public information and never tells you what to do. You decide.