The next time someone posts a loud forecast — "this asset is going to triple," "the market crashes by autumn" — ask them one question. Not "why." Ask: "Would you put your whole portfolio on it right now, with no take-backs?"
Watch the idea evaporate.
The distance between what a person says and what they'd actually stake money on is the entire content of financial noise. 99% of public ideas live in exactly that gap: loud enough to be heard, cheap enough that no one has to answer for them.
The only product with no recall for defects
A forecast is the only product on earth whose maker pays nothing for a defect. A car with a factory flaw gets a recall and a lawsuit. A building that collapses gets a courtroom. A public forecast that doesn't come true gets nothing. Zero. A week later the author posts the next one, to the same audience.
Free to make, costless to be wrong. Under those rules, the feed fills up not with accurate ideas but with loud ones — the selection runs on volume, not on being right.
Here's a number worth holding in your hand. The research firm CXO Advisory collected 6,584 public stock-market forecasts from 68 well-known experts between 2005 and 2012 and graded every one. Average accuracy: around 47%. Worse than flipping a coin. And it gets worse — the spread of accuracy across the experts formed an ordinary bell curve, exactly what you'd get if they were guessing at random.
This doesn't mean they're stupid. It means forecasting the market in public is an activity where knowledge isn't even required. Only a microphone is.
Why we buy it again and again
A built-in perception bug fires here: we read confidence as information, and volume as conviction. Both are backwards.
The most uncomfortable proof came from psychologist Philip Tetlock. Twenty years, 284 experts, more than 82,000 forecasts. The conclusion that made his book famous: the average expert barely beat chance — "worse than a chimpanzee throwing darts at a board." But here's the detail usually left out: the more famous the expert, the more airtime they had, the worse they predicted.
That's not a paradox. It's mechanics. Loudness and accuracy run on opposite incentives. To be right, you have to doubt, wait, admit uncertainty. To be loud, you need the opposite — confidence with no caveats, a forecast with no range, a date and a number. The person who actually knows sounds boring: "probably," "under these conditions," "not sure." The algorithm doesn't surface that. It surfaces the one shouting "100%."
So the loudest voice in the room is structurally the least informed. Not always — but often enough that you can bet on it.
The deeper danger isn't the money you lose
You might think the cost of noise is the loss from bad advice. You followed a loud call, you lost money — painful, but once.
The real cost is subtler and more expensive. By listening to the loud ones, you outsource your conviction to people with nothing at stake — and you never build a single record of your own. You become a permanent spectator to other people's unaccountable guesses, without one line of your own on the scoreboard.
And memory finishes the job. A year from now you'll remember the borrowed forecasts that came true ("I read that this would happen") and erase the ones that flopped. Exactly like the gurus in the CXO study, you grade yourself on a curve you drew yourself. You don't have a rating — you have a feeling that you "basically understand the market." That feeling is backed by nothing and is worth precisely what a stranger's loud forecast is worth. Zero.
What to do about it
Two moves. Both simple, both unpleasant.
First — the portfolio test. On everyone, including yourself. Any idea — someone else's or your own — passes through one filter: would you size a position on it right now? A specific share of your portfolio? If not, it's not an idea. It's a mood wearing the costume of analysis. Most loud calls wouldn't survive a single second if the author had to name the size of their bet.
Second — keep your own permanent scorecard. Every time you hold a real opinion on an asset, write it down: date, claim, your conviction from 1 to 10, and what exactly would prove you wrong. Then don't edit it. A year later, count your real hit rate.
It will hurt. For almost everyone it lands closer to 47% than to the imagined 90%. But it's the first honest number about yourself as an investor you'll ever have. The person who knows their real hit rate stops needing anyone's microphone — and stops being anyone's audience.
Why hot air is always free
Noise is free for one reason: nobody keeps the receipts. A forecast dissolves into the feed faster than it can be checked, so there's no shame attached to it.
Now imagine the opposite. A world where every published idea carries a permanent rating — like in chess. Wrong, and the rating drops. Right, and it climbs. You can't edit or quietly delete a call that aged badly. Every "100%, I guarantee it" is a bet, and it's recorded forever.
In that world, 99% of the voices go silent on day one. Not because they were banned. Because they were never willing to pay to be heard. What's left is the people who actually knew — and now, at last, can prove it.
Those who know don't shout — as long as shouting costs nothing. Make shouting expensive, and the silence around it becomes the most honest signal on the market.