[disruptalot]

> By collating all such timestamps, Ndinga’s firm can aggregate the cost basis for every bitcoin transaction on record.

If i'm understanding the methodology here, the assertion that it translates to investors is really flawed. You can't assume:

- a transaction appearing on chain constitutes a buy

- that the transaction represents 1 investor

- that 1 investor = 1 address

[ayngg]

The article looks like some kind of ad for 21shares, which isn't even primarily an on chain analytics firm. I assume it is looking at on chain data for liquid btc to see what percentage of that is currently being held at a price lower than when it was last transacted.

[nostrademons]

On-chain data is basically useless when it comes to Bitcoin investors. The vast majority of people who hold Bitcoin for price speculation do so through an exchange like Coinbase or Binance. There the Bitcoin is just held in an exchange's wallet, commingled with millions of other customers, and the exchange holds a private record of each customer's individual balance. The exchange generates a wallet address for a customer only when they want to send crypto to their account.

It's the civil libertarians, criminals, and tech aficionados that hold Bitcoins in unhosted wallets. When there's on-chain activity for these groups, though, it's usually a transaction rather than a trade. They hold Bitcoin for the activities they can do with it, not for any hope of the price rising.

[tomrod]

Sure you can. Whatever those transactions of bitcoin represented to the buyer and seller, they were worth a certain value on a certain exchange at a certain time. That has a different value on a certain exchange now, hence the evaluation of a delta.

Whether the bitcoin holder sees it as a loss of value is beside the point, since purchasing in bitcoin is hard still.

It's like using Euros in the US. Some folks accept them, but generally dollars are preferred.

[kkjjkgjjgg]

Most Bitcoin trades don't even register on the blockchain. And not every transferal on the blockchain is an investment. What a silly methodology.

[tomrod]

And used all the time for Forex, no? Currency transfers need not be tracked to happen either? Yet we can still say how strong the yen is relative to the peso.

[randomhodler84]

It’s a dumb methodology;

I’ll give you an example, I bought at BTC $36k last month and just withdrew it from exchange after all the fiat cleared in from the legacy banking channel today. It hit the chain when BTC was $42k, so these jokers would think the cost was 42. You can’t get an accurate picture from on chain data alone.

[cyberge99]

Another major failure of the methodology is that it fails to take in to account all of permanently lost key/locked up BTC.

[listenallyall]

You can withdraw bitcoins from the blockchain? That would be fascinating new feature if it existed. /s

[randomhodler84]

I added “from exchange” for clarity. I would have thought the ref to legacy banking channel would be enough reason to imply I was interacting with a fiat to cryptocurrency exchange.

[jallen_dot_dev]

Not really the same, because you can look at what currencies are actually being traded for. These researchers are not looking at bitcoin trades.

A better analogy would be like looking at clearing house data, and examining the net number of shares bought or sold at each broker, to determine how many individual investors are in the red. Even though the price changed throughout the day and you're only looking at the price at settlement time. And the net doesn't tell you anything about how many investors actually traded.