Mikhail Nikulin
Sep 5, 2018 · 2 min read

Agree, the scenario with the back-and-forth transfers has even more sense to keep a channel opened for both parties. But then hub would need to decide either to close a channel or to keep it opened even if the channel balance is in favor of hub (so the hub’s money is blocked in the dead-end channel). Closing the channel has more sense for hub but it kills the benefits for the client. In contrast keeping the channel opened leads to the intrest rate fee model which I’ve mentioned in the previous comment.

Even despite of that if there is a hub willing to keep channels opened it would need to wait for the “salary” transfers back to users during a whole month. Until that the significant fraction of the hub’s capital might be locked. Having more and more users hub whould need to lock more capital. This will dry out the channels with the main part of the LN. So to be able to route payments hub will apparently need to find more capital or to close some of the biggest dead-end channels to unlock its money. The smaller channels might be kept opened at the same time.

In addition the hub that aggressively grows the number of single users will face the significant number of small channels opened. The capital distributed across these small dead-end channels will be sliced into many tiny pieces. Closing these channels will trigger lots of tiny inputs on the hub’s hotwallet caused by the UTXO model which might take a lot of fees to aggregate that sort of dust.

    Mikhail Nikulin

    Written by

    Lykke Exchange co-founder