An independent record of finance YouTube

What finance YouTubers predicted, and what actually happened

Every prediction is recorded word for word, dated, and linked to the exact second of the video, then checked against real prices and published economic data. Automatically, video by video, with nothing picked by hand.

2,354of 3,238 eligible videos read so far1,962predictions recorded766results checked against real data
What was said
Coin Bureau

@coinbureau

“For altcoins, though,   a scaling strategy makes more sense. Most are  sitting near the bottom of a multi-year range,   and history shows that a bounce is more likely  than a total collapse before the bull run ends.”
Watch at 18:15Fear vs. Greed - The Ultimate Crypto Indicator!!
PredictionETHwill riseChecked at fixed windows, from 30 days to 5 years

Recorded as altcoins, a word for a group: this call is judged on ETH, the stand-in the method assigns to it.

What happenedafter 1 year
Incorrect

Entry price (first close after the video)

$3,168

Nov 15, 2025

Price at close

$1,835

Nov 15, 2026

Return

−42.1%

S&P 500 +14.5%

ETH closed at $3,168 on Nov 15, 2025. On Nov 15, 2026 it closed at $1,835, a return of −42.1%. SPY returned +14.5% over the same window. The call was that it would rise.

30 days

Incorrect

90 days

Incorrect

180 days

Incorrect

1 year

Incorrect

1 of 5

The predictions whose results arrived most recently, newest in front — whatever they say.

Callsight (getcallsight.com) is an independent record of what finance YouTubers predicted and what then happened. As of Nov 18, 2026, it has read 2,354 of 3,238 eligible videos from 4 tracked channels, recorded 1,962 predictions word for word and checked 766 results against real prices and published economic data.

Up / down calls

90 days to Sep 9, 2026

Leaderboard

Judged after 1 year. Ranked by score (hit rate above the base rate, return against a benchmark and calibration), not by hit rate alone. How the score works

Full leaderboard
#ChannelScoreHit rate
1Graham Stephan5.1M subscribersScore70Hit rate60.6%104 callsOn the 84 of 104 calls that have a base rate: 57.1% correct, base rate 47.2%
2Andrei Jikh3.3M subscribersScore65Hit rate58.8%51 callsOn the 49 of 51 calls that have a base rate: 57.1% correct, base rate 41.7%
3Joseph Carlson511K subscribersScore56Hit rate70.4%71 callsOn the 69 of 71 calls that have a base rate: 69.6% correct, base rate 60.5%
4Coin Bureau2.7M subscribersScore38Hit rate46.6%178 callsOn the 160 of 178 calls that have a base rate: 46.9% correct, base rate 49.2%

Leaderboard of the financial channels tracked here. Their calls are checked under the same published method, and every rate is shown with the number of checked calls.

Recently called instruments

The instruments with the newest calls, newest first, and how those calls turned out.

All instruments

The economy

Calls on house prices, interest rates, inflation and recessions, checked against published US economic data (via FRED).

All topics

Just settled

Predictions whose result arrived most recently, in the order it arrived.

Latest calls

How it works

  1. 1

    We read the videos, not a selection

    Videos are read automatically, none picked by hand, so the memorable calls count no more than the forgotten ones. Older videos are still being worked through; the figures at the top of this page say how far it has got.

  2. 2

    We record each prediction

    Each call is kept word for word, with the date and a link to the second it was said. A paraphrase is never shown as a quotation.

  3. 3

    We check it against reality

    Code, not opinion, compares each call with real prices and economic data at 30, 90, 180 and 365 days, then at 3 and 5 years, and the method is published in full.

Consensus signals

A signal needs 2 channels with a proven record to agree on one instrument, with no channel that has a proven record calling it the other way.

How signals work

Blog

Questions

Is this investment advice?

No. This is a record of what people said in public and what happened next. Nothing here tells you to buy or sell anything.

How is a prediction judged correct?

By fixed rules applied the same way to everyone. A price call is measured from the first closing price after the video went public; an economic call is measured on the series it is about, as first published. Every rule is on the methodology page.

What is the base rate?

How often the called move happened anyway in comparable past windows. A hit rate only means something beside it: a call that comes true most of the time anyway says little.

Why are some predictions not scored?

When a creator is reading out someone else's forecast, such as a bank's target, it is shown with the source's name but not counted for or against them. Vague statements with no instrument or no direction are counted separately and never scored.

Can a creator dispute a result?

Yes. Anyone can report a problem with any prediction from its page. The report reaches the operator, and a call taken off the lists keeps its page, with the date and the reason.