About the Bitcoin Observatory

One person, open data, and numbers you can check yourself.

The Bitcoin Observatory measures Bitcoin’s price against its power-law trend and the floor beneath it. Alexander Mooij builds it and runs it. The data and the formula are published, and so are the times the model fell short, so you don’t have to take anyone’s word for it.

Who builds it

Alexander Mooij

I was a stubborn student.

I first came across Bitcoin in 2013. I thought it was interesting and shrugged it off. It crossed my path again in 2015, in 2017 and in 2019, and by then my fiat brain could only see a useless, pointless piece of software.

Money itself I did take seriously, because I know what it feels like to be broke while prices keep rising. In 2015 I started reading the usual books: Rich Dad Poor Dad, Your Money or Your Life, J.L. Collins, Mr. Money Mustache and many other great writers in the financial independence community.

They taught me that what sets you free is building assets and a balance sheet, not just a bigger income, and taking full responsibility for my financial future felt deeply productive. But one question never came up there: why do I need to invest at all? Why can’t I just save?

Saifedean Ammous answered it. When I read The Bitcoin Standard in 2021, it finally clicked. The world’s currencies are built so that more and more debt can be issued in them, because that debt pays for the governments behind them. Saving in that money is a slow loss. Hard money is the way out, and Bitcoin is the hardest money there is. That was the insight.

Lyn Alden later picked up the thread in Broken Money. Once you see how broken money is, and that you can opt out, you understand what Bitcoin is.

In 2022 I bought a miner as my way in and built it a soundproof box in the shed, with just enough airflow to keep it cool. Mining turned out to be hard: the bitcoin price, the hash rate, the energy price and the price of the machine itself all moved every day. When I heard Saifedean say that all miners need to get wrecked, I knew exactly what he meant.

In 2023 my family and I started travelling full time. In 2025 I found the power law and went deep. In 2026 I started building the Observatory, so anyone can check what I found.

Why trust it

Don’t trust it. Check it.

The core numbers come from one public dataset and one formula. Here is where to look.

5,970daily closes

The same data, the same line

Download every daily close since July 2010, copy the Python into your own notebook, and get the same fit (R² 0.961) and the same floor.

Reproduce the power law
October 2026

A check every month

The formula is tested again every month on fresh prices: did anything close below the floor, and where would a refit on all the data land? The result is published, good or bad.

The formula, month by month
0.546out-of-sample R²

The misses are published too

Fitted on 2010–2020 data alone, the model explains 2020–2026 with an R² of 0.546. Fitted with hindsight on everything, it scores 0.961. A floor drawn in real time also held up less well. Both results are written up next to the claims that did hold.

Read the formal verification
22research papers

Work shown, changes dated

The tools rest on published papers, each with its method and limits written out. New papers, tools and changes are logged with a date.

See the changelog

The fine print

What this site is, and what it isn’t

An expectation, not a promise

The power law is fitted to the past. It can break, and the floor has been crossed before. Read where it could fail before you plan on it.

Not financial advice

I’m not a registered financial adviser. The tools show what the model implies for your numbers. What you do with that is your decision.

How it’s paid for

The tools, papers and ebook are free. Donations and Observatory Pro, a paid subscription, help pay for it.

Built on others’ work

The power law isn’t mine

Giovanni Santostasi, Harold Christopher Burger, Stephen Perrenod and others found and tested it, and arrived at nearly the same exponent independently. The trend line here uses Santostasi’s published parameters from bitcoinpower.law. On top of it, this site builds a floor to plan on, tools to plan with, and tests anyone can repeat.

Who found what: the literature review

The model in two lines

trend = 10−16.493 × days5.688

floor = 0.42 × trend

Days are counted from the genesis block, 3 January 2009. Today the trend is $160,959 and the floor $67,603 (Nov 21, 2026).

How it’s fitted: Methodology

Still not convinced? Good.

Take the data, run the code and try to break the floor yourself. If you find a mistake, email me.

Check it yourself