Some person or group that goes by the name of LNBIG and runs close to half of the entire Lightning Network (LN) has stated he makes less than a dollar a day for the privilege of locking millions worth of bitcoins.
“I have 200-300 transactions through all nodes a day, rarely 600. On commissions, I earn 5,000-10,000 sats per day. It’s $0.4-$0.8. It’s $20 [a] month maximum,” he said.
He is incurring losses for the service he provides, not just in locked liquidity, but in having to pay far higher on-chain fees than what he charges for LN transactions/liquidity, so stating:
“Opening of the channels (closing-opening again) I spent, probably, more than one thousand dollars. Therefore, no earnings now.”
The “samaritan” runs 1,800 channels between his nodes with each channel having a capacity of 0.16, translating to 336.38891179 bitcoin as of 18th of August.
That’s about 40% of the entire network capacity of circa ₿800, falling over the past few months as he gets rid of inactive channels.
There were some 1,080 bitcoin on LN as of April. Now it’s at about ₿830 with the decrease mainly, if not solely, due to LNBIG.
“LNBIG is closing unused PUBLIC channels and re-allocating funds to PRIVATE (mobile) channels. This will lead to 1ml.com and other LN explorers to report in drop of liquidity,” supporters say.
LNBIG himself says he faces a dilemma. “When you open a lot of channels – everyone scolds you that you are capturing the network. When you close – there are also concerns.”
This person or entity apparently has channels opened with all LN users, so extending liquidity to allow them to actually be able to use LN.
If you’re not aware, as Adam Back of Blockstream says, LN can be seen as sort of tabs. Your card for some reason suddenly doesn’t work at your local shop, so you tell the shopkeeper you’ll pay him the 100 satoshis tomorrow.
The shopkeeper writes this down: Trustnodes owes 100 sats. Now he goes to the supplier and says this paper entitles you to 100 sats. The supplier then does the same with the farmer.
This 100 sats is already locked in the system, but for it to move we need “free” 100 sats, an amount that is not owed to anyone. Thus this “free” sats sort of is the paper.
LNBIG is in the business of “creating” this paper to facilitate the revolvement of “tabs” so that you can make the IoU payment to the shopkeeper.
In other words, it is 100% reserve banking. To make that full on reserve work, you need for simplicity let’s say 200% collaterals.
So 1 bitcoin is “free” and acts as the “paper” for the 1 locked bitcoin which can’t move until eventual onchain settlement.
This collateralization requirement is the main non-political criticism of LN with Emin Gün Sirer of Cornell stating last year that the Lightning Network is “economically broken.”
“Suppose somebody could receive up to $10,000 from Coinbase, and Coinbase has, I don’t know, ten million users. So ten million times $10,000, erm $10 billion… $100 billion dollars get’s tied up? Is that right?” – he said.
The other problem appears to be that the fees one charges for an LN transaction seem to be far off from covering the cost of on-chain fees.
That’s perhaps because LNBIG is undercharging in this case. At 300 transactions a day, at a cost of 10 cent per transaction, you can get to $900 a month, so just about covering his on-chain fees.
On-chain fees, however, during this period were not far off from 10 cent themself, perhaps even less.
So even in a pure comparison, it may be the case Lightning Network transactions have to be as expensive as on-chain transactions. Then you add the cost of securing these nodes and channels, of running the system, the opportunity costs in regards to potential price movements or simple other investments and use cases, and really you suddenly have to wonder whether LN transactions are actually cheaper or are instead more expensive than simple bitcoin transactions.
For now the network is very small with just $8 million worth of bitcoin in total capacity, but as it grows, conceptually you’d think the costs would grow too.
Certainly the opportunity costs, but that might actually be secondary to pure costs in as far as the more users and the more channels, the more on-chain transactions, thus the higher fees, to the point where for it to be profitable LN transactions might have to cost as much as on-chain transactions or more.
The political criticism in this case is obviously centralization. This one guy can crash LN and can make it unusable. For now that might not matter because pretty much no one uses the Lightning Network, but if it is being used and it suddenly goes down because one guy with much of liquidity closes channels, then this could be reflected on price.
You can say they’d have no interest in doing so, but mischievousness can be less of a concern than potential hacking of the point of failure.
Yet quite interestingly all of the aspects appear to be more of a distant concern because here there seems to be a pretty fundamental question of whether LN transactions can actually be cheaper than on-chain transactions on an apples to apples comparison due to the constrain of collaterals.
Thus even if one ignores opportunity costs and so on, it appears to be a bit unclear as of now whether the settlement maths does actually work out when we account for collateralization.
As it stands, it seems it doesn’t. Whether that will change, remains to be seen.