Two guesses:

Guess one, they made a big transaction with most of their hot wallet, partially paying out to someone who withdrew, with the rest going back to the hot wallet. This transaction is in the queue but not finalized yet. The network won't accept any other spends of that money. Because that would be a double spend. So until that transaction is finalized (or it gets to old) that bitcoin is 'stuck'.

Or alternatively their hot wallet is empty and they have an outstanding transaction to refill it from their cold wallet. (They have the cash, but they have to wait until they can get it from their 'vault', and there is a traffic jam on the way to the vault).

I am not sure the first guess actually still works this way, I vaguely recall there are now mechanisms for one transaction to 'override' another non-confirmed transaction.

> there are now mechanisms for one transaction to 'override' another non-confirmed transaction.

There's at least two, "child pays for parent" and replace-by-fee. It's pretty hard to get regular transactions really "stuck", these days.

Ah ofcourse child pays for parent essentially always works. If I recall correctly replace-by-fee requires wallet support perhaps even at the time of creating the original transaction. But child pays for parent is just universal because it doesn't need to replace the transaction in the mempool, it just pays the miners to move it out of the mempool.

Excellent hypotheses, it should be possible to see these transactions if they are out there. That's why it's a public blockchain.

What queue? Is there a queue in the bitcoin network?

I thought it was a question of paying a high enough transaction fee and you would get your transaction done faster / within 10 minutes.

> What queue? Is there a queue in the bitcoin network?

Yes, that's what the mempool is.

>I thought it was a question of paying a high enough transaction fee and you would get your transaction done faster / within 10 minutes.

that's broadly true, but if transaction volume spikes then your initial fee estimate might be too low. there are ways around this (RBF, CPFP), but it's unknown whether their wallet software handles it.

How would this be an excuse for Binance to block outgoing transactions?

Unless somehow all of Binance's bitcoin was stuck in the mempool ...

They just described a situation where most of their readily available bitcoin (exchanges do not keep more than a minimum of their available cryptocurrencies in a programmatrically available wallet) could be stuck in the mempool.

It wouldn't need to be all their bitcoin, just their hot-wallets. Moving bitcoin out of cold wallets should generally be a slow process.

It's a queue, but not a first-come first serve one. The problem I imagine is that, once you set your transaction fee, and it turns out to be too low, it is hard to retroactively pay more fee. Just because the system doesn't allow two spends of the same output in the mempool at the same time, and canceling a transaction is hard.

You are right about your first guess.

Childs can pay fees for their parent now.

You know how the internet is a series of tubes? Well someone clogged the bitcoin tube over at Binance.

It all makes sense now.

Bitcoin is off ~20% against USD in the last two days, ~13% in last day.

That much movement is bound to drive transaction and redemption volumes way up at any exchange.

1 guess: Transaction fees spiked and they didn’t pay enough for their transaction.

Doubt it's that. They ran out of reserve for their fractional reserve. Crypto on exchange wallets has already been dwindling the past few months.