Most bitcoin investors never move their coins off the exchange and thus don't have a distinct address on the blockchain, so your sampling technique is probably just as flawed as time of app download, if not more so.
I'm curious: What is your source for this statistic?
For comparison: There were about 40M non-zero addresses earlier this year. See eg https://news.bitcoin.com/number-of-addresses-holding-btc-tap... [The current number is on Glassnode: https://studio.glassnode.com/metrics?%3Bm=blockchain.UtxoLos...]
FTX is one of the larger exchanges, and FTX bankruptcy headlines say they "may have 1 million creditors". See eg reuters.com/technology/ftx-officials-contact-with-us-regulators-filing-2022-11-15/
Edit: I'm trying to figure out if someone has actual stats or whether the statement is just intuitively guessing. There are so many bitcoin stats out there, it's hard to keep track of it all.
An exchange, when depositing or withdrawing to a user or another exchange, uses a ton of address, sometimes leaving behind very small amounts. Same with users. I'm sure 75% of non-zero address have very, very little BTC in them.
>I'm curious: What is your source for this statistic?
They have no reason to.
Unless the investor is specifically working on some HFT arbitrage scheme between exchanges, then they don't even need to move crypto out of the exchange.
Also given the complexity of rolling your own wallet, it's just a risk that's not worth it when KYC exchanges are relatively stable.
Apart from having insider knowledge in how cryptocurrency exchanges operate and their data, how could you possibly know this?
One address on the blockchain can belong to one or more users, just like one user can have one or more addresses on the blockchain, so being able to compare feels impossible even if you had insider knowledge/data.
Bitcoin on a centralized exchange cannot be traded for goods and services until explicitly withdrawn, unlike most brokerage accounts where cash is cash and many will even give you a debit card so you can spend directly from the account. This strongly decouples bitcoin and most other cryptos from the total money supply.
When you're on Binance and you send them one BTC as a block chain transaction, you are receiving one Binance!BTC that does not exist outside of that walled garden.
Just look at FTX - some 1.8B in BTC liabilities but zero BTC assets
This splitting hairs over "true" ownership ignores the fact that the original investor analysis done by GP here premised on chain transactions is still semantically wrong.
that still leaves us with "most".
everyone knows the theoretical issues and incompleteness of this statistics collection - since there is no single public or verifiable stat to use - its pretty new to be a person that chooses assume its more likely that people are having distinct addresses on chain, than assume its more likely they're on an exchange
kind of funny how we go full circle here, but okay. hope that helps your reasoning. this article has no way of making its claims either, so there's really no point of appealing to either authority or bolster any opinion.
The issue of one's "private" data then becoming part of a public dataset could be alleviated with deidentification, plus a better acceptance that when companies accumulate data about you, it's generally going to get exploited in some way or another.
Every cryptocurrency exchange would have to go bankrupt before you could make sweeping statements like that if so.
Is consistently proven to be a myth.
If I buy some shares in a company that does things with [X], I think it makes sense to say that by doing so I'm "investing in [X]", even if I don't e.g. have the necessary license needed to directly own any [X].
"You aren't a real real estate investor if you buy the brooklyn bridge from a guy on a streetcorner in Harlem" is true because the claim on the bridge was never credible, not because you never physically took possession of the bridge itself. You could argue that the exchange's IOU was never credible, but that's different from saying that any BTC-based IOU is made of lies.