Shock horror: Bitcoin’s wealth distribution is outrageously uneven. The top 2.8 percent of wallet addresses control 95 percent of the supply of bitcoin, according to statistics. At first glance, it looks like bitcoin’s community has a more uneven distribution of wealth than the US–where the top 10 percent of citizens hold more than 75 percent of all wealth–and South Africa, the current champion of inequality. Not so fast.

The statistic in question comes from BitInfoCharts, a website that tracks the distribution of bitcoin across all known Bitcoin addresses. The website calculates the data from publicly available Bitcoin information but the data is a) not really reliable b) doesn’t really tell the whole story, and c) shouldn’t be used to draw such extreme conclusions. Here’s why.

Satoshi Nakamoto’s hoard of bitcoin

When Bitcoin was first created in 2009, its pseudonymous inventor, Satoshi Nakamoto, mined a huge number of bitcoins to get the ball rolling (otherwise no blocks would have been produced). BitMEX estimates that he, she or they mined 700,000 bitcoin, worth $7.4 billion in today’s prices.

But once a bunch of other miners joined the network and it was working as expected, Nakamoto dropped off the network. Since then, not a single bitcoin has been moved from the addresses Nakamoto used, and most believe it’ll probably stay that way. But, these coins—working out to 3.9 percent of the current supply of bitcoins—still exist, and continue to tip the scales of the bitcoin wealth distribution in the wrong direction.

Lost, missing and forgotten coins

Nakamoto’s treasure trove is not the only bitcoin that won’t be moving any time soon. One of the features of the currency is that you can’t delete coins–burning bitcoin simply moves them to an address that nobody can access. But there are just as many, if not more coins stuck in addresses wallet owners have lost access to, be it via losing a password or throwing away a hard drive with $86 million worth of bitcoin stored on it as this guy did.

In fact, Chainalysis estimates that potentially up to 3.8 million bitcoins (including Nakamoto’s) may be gone for good. This accounts for 20 percent of the current supply, and since many of the coins assumed lost were mined in large amounts in the early days, removing them from calculations on wealth inequality starts to make everything look more distributed.

Exchanges hold a lot of coins

The majority of addresses holding bitcoins that aren’t assumed to be lost, are owned by crypto exchanges. These look after coins on behalf of their users—but it doesn’t mean they own the coins. Putting security concerns aside (exchanges are goldmines for hackers), users can withdraw their coins any time they wish.

Trader Sasha Fleyshman points out that seven of the top ten most populated bitcoin wallets are owned by crypto exchanges, and that they contain around 727,000 bitcoins. On its own, this accounts for 4 percent of the supply of bitcoins, put together with the hundreds of other crypto exchanges and numbers shift again.

It’s easy to make Bitcoin addresses

One problem specific to this data is that it focuses on the percentages of addresses, i.e. how many addresses are there and how much is held in each. But, it doesn’t factor in if those addresses are active. It’s really easy to make a Bitcoin address, exchanges often create a different address every time they send and receive money. This results in millions of empty addresses, or ones with barely a satoshi in them.

Join Daily Debrief

Want the best of crypto news straight into your inbox? Sign up to Debrief.

Join Daily Debrief

Or

The data assumes there are millions of bitcoin paupers floating around the internet, but in reality, it’s more like the addresses have simply been discarded.

Bitcoin address doesn’t mean Bitcoin user

Another reason why bitcoin wealth distribution can’t be measured this way is because it’s measuring addresses, not people. The statistics provided by BitInfoCharts is simply a tool for seeing which addresses own how much bitcoin. But, any one person can own a multitude of addresses, with varying supplies of bitcoin.

This actually skews the data the other way, making it look more fairly distributed than it actually is. It’s easy to understand why this data is used; it would be far more complicated to find out who owns which addresses, in order to total them up. But regardless, it proves that this data—while useful for various reasons—is utterly misleading as a metric for measuring the real wealth distribution of bitcoin.