Hacker News Reader

Comment by runeks | Hacker News Reader

> With the illiquidity of the btc => usd conversion, they’ll have to pay the merchant before selling the btc, which means they’ll have to take a huge margin to cover potential slippage.

Bitcoin sell liquidity is currently $162MM USD on Bitfinex at 0.5% slippage. Buy liquidity is $543MM USD (at 0.5% slippage). For comparison, the sell liquidity on Coinbase Pro at the same slippage is only $4.5MM USD.

My site https://cryptomarketdepth.com/ tracks the liquidity of various cryptocurrencies over time (for the exchanges: Binance, Bitfinex, Bitstamp, Bittrex, Coinbase Pro).

View on HN

The problem isn't volume, it is latency. It can take 30m for a btc transaction to settle. That's a nontrivial amount of volatility to exposure yourself to.

Could PayPal make this the customer's problem, by only allowing them to exchange or spend settled funds?

I am probably way out of my depth here. But if Paypal or Square have a large enough market/volume of transactions couldn't they act as their own market maker, settle immediately internally? Kind of like the whole storm over Citadel GME. And then also not have to pay the BTC fees?

That'd require them to hold your Bitcoin for you, so they could just move some from one person to another. Isn't that counter to the Bitcoin philosophy?

I believe that’s what they do currently. I made the mistake of buying some LTC recently on PayPal. Turns out you can’t actually send it to any other LTC address ... your options are only “buy” and “sell”.

Ideally everything would be self hosted p2p btcpayserver like technology. But mass adoption requires some compromises. Not everyone wants to self host and self custody, and if you’re willing to compromise a little on that you can have massive UX gains, especially if you’re looking to swap to fiat on either end anyways. As long as you retain the ability to self custody your currency should you desire, it’s not even that big of a compromise IMO.

So it does exactly what paypal already does only with more steps and volatility...

AH crypto don't ever change or do you're a fucking waste.

Can't they sell before confirmation? Probably on the order of 99% of transactions that propagate to the whole network with a sufficient fee end up being confirmed, and that only takes a few seconds to happen.

I would go further to say that for a large portion of merchant transactions, confirmations are a non-issue. It is not trivial to pull off a double spend and no one is going to do it for a cup of coffee. The miniscule number of times that it might happen could be written off as the cost of doing business, the same way the big credit card companies quietly eat all the credit card fraud that happens.

Also, and I could be mistaken, my understanding is the double spend can only happen when the person trying to double spend controls the private/public key to the coins. Normally when you deposit crypto in an exchange (or this hypothetical Paypal wallet) you lose that access and your balance is just a row in a database while "your" coins are swept into a large hot/cold wallet run by the service you are depositing into.

This is correct. Customer double spend risk is not what’s ring discussed here. The risk is PayPal losing money on the spread. A user buys a pair of shoes with bitcoin, PayPal pays the shop owner fiat and takes ownership of the bitcoin. Now PayPal is on the hook for selling that bitcoin if they want to keep their cash flow in fiat. The price of bitcoin can move wildly in the time it takes for a Bitcoin transaction to settle. In reality PayPal is a huge company and can sell to many brokers who will take on the risk for PayPal.

But Bitpay and others provide instant settlement and manage the volatility on their side. Merchants don't have to risk anything.

Probably by being an authorized participant in GBTC - is the premium still negative? Wonder how long they can keep their business going that way.

Then use the lightning network, it's really what should be used for smaller payments like this anyways. You don't pay for coffee with a wire transfer, for the same reason you shouldn't be using layer 1 for small everyday purchases like this.

But you need dollars on the other side. For this to work without PayPal hosting the BTC you need the following:

1. You send BTC to PayPal 2. PayPal sends equivalent dollars to the merchant.

But what is "equivalent"? What is the price of BTC at any given moment? It fluctuates a lot and you can't "lock-in" a price without waiting minutes (as opposed to milliseconds in other financial systems). So between the time that you sent dollars to the merchant and when you sold the coins you are exposed to risk of holding BTC.

Lightning does not help with this.

That's a fair point. In order to use lightning, PayPal would either need to hold some BTC in reserves on layer 2, or work with an exchange that accepts lightning transactions.

Lightning requires the bitcoin network to verify the transaction though. It's like using a personal check - until the bank clears it you're at risk.

That's not true at all. You can perform lots of transactions over lightning, and settle with a single on chain transaction. The money clears instantly, and is not at risk, even if you keep the funds on layer 2.

No, the previous comment was correct, in that an open lightning channel balance is inherently at risk, if you're wallet software isn't monitoring that the lightning nodes you're dealing with follow the protocol. If a node broadcasts an older HTLC transaction to bitcoin's main chain and your software isn't watching, you may loose some/all of your channel balance.

That is correct, but the GP said:

> It's like using a personal check - until the bank clears it you're at risk.

Which is 100% not true, it's not like a personal check. You are only at risk if:

* your counterparty decides to try and cheat you (despite knowing they will lose even more money if they attempt to screw you and you notice, so some game theory is involved)

* you are not watching onchain for a period longer than the monitor period (typically several days)

* you have not elected to have someone else monitor on your behalf (in exchange for a small cut of the "profits" if your counterparty tries to cheat you)

My point being - it's nowhere near the same as a personal check.

That's just like a personal check. The same methods you cheat someone with a scam check applies to the IOUs on Lightning.

* Bad actors exist and will do this with checks or lightning, game theory applies to both

* It doesn't matter if you're watching the chain, you've already transacted with the buyer

* 3P Escrow is not unique to Lightning and is sometimes used with checks too

> Bad actors exist and will do this with checks or lightning, game theory applies to both

It does, but with lightning, as long as you are watching the chain periodically, you are guaranteed to not lose money. The same cannot be said for checks.

> It doesn't matter if you're watching the chain, you've already transacted with the buyer

It does matter if you watch the chain - this is how you determine if your counterparty is trying to close out the channel with an out of date state (at which point you can submit an onchain transaction that corrects the problem + collects a penalty from your counterparty).

> 3P Escrow is not unique to Lightning and is sometimes used with checks too

Watchtowers are not anything like escrow. It's simply allowing a third party to submit penalty transaction iff your counterparty attempts to submit an out of date state. They do not take custody of your funds.

People watch for bad checks as well. Plenty of cash registers used to have example bad checks laminated on them, back when the modern financial system was still using IOUs like that.

You can make a lightning transaction, drain your bitcoin wallet before it verifies, and no amount of watching the chain will get the seller back whatever they traded you for that transaction. Just because you identify the fraudulent transaction before it finally verifies doesn't mean that the fraud wasn't committed. Same with bad checks.

I thought you were talking about 3P escrow, because watchtowers are even worse than escrow in fighting fraud, for reasons outlined above.

> You can make a lightning transaction, drain your bitcoin wallet before it verifies, and no amount of watching the chain will get the seller back whatever they traded you for that transaction. Just because you identify the fraudulent transaction before it finally verifies doesn't mean that the fraud wasn't committed. Same with bad checks.

No you cannot, again, that's not how it works. You commit funds to a channel on chain before you can spend funds in a channel. I don't think you understand how lightning works.

Yes, you commit funds before the initial transaction, but loop allows you to move them to another wallet before the channel closes.

Wait, for real?

Lightning is even dumber than I thought - and I already thought it was DOA.

It pretty much is - Lightning plateaued out and has been known to have empty channels.

https://bitcoinvisuals.com/ln-channels

If I’m interpreting this right, it means the bitcoin market on bitfinex is a bit over 3x more sensitive to sells than to buys? And the possible sell volume on coinbase before slippage is tiny.

That is correct. But this is only for March 30. On the day before the buy/sell liquidity were roughly equal.

You can click on a point in the time series graph to see this.

I am curious how in the world does Bitfinex offer USD liquidity these days?

I mean lets say you trade USD/USDT on Bitfinex.

How do you get USD (not USDT) out of Bitfinex?

A Regular bank wire withdrawal is sent out from Bitfinex within 5-10 business days and credited to your account following processing by the receiving party.

2. An Express bank wire withdrawal is sent out from Bitfinex within 1 business day and incurs a 1% processing fee.

Is that USD or USDT?

The fiat on-ramp is probably a pretty big blocker for many new adopters. If paypal fixes that, it could have a large effect.

It’s USD, not USDT.

Reply on news.ycombinator.com