Below is a collection of data on some of the top LN routing nodes in September. On the Y-axis is their BTC-denominated routing volume, and above each bar is a normalized ratio of routing volume divided by public capacity.

Interestingly, these routing nodes are all routing more than their public capacity each month.

I wonder if this trend can be extrapolated to the entire Lightning Network (and what kind of discounting would be required to eliminate double-counting)?

With some additional data, we might be able to back into a rough estimate for Lightning Network payment volume.

For example, if we assume every node is routing an average of 1x their public capacity each month, and apply a 50% discount for double-counting payments, we’d be left with 2,500+ BTC of Lightning payments/month (5,000*0.5).

That’s $50M of payments/month, or $600M/year.

No idea if my discount assumptions above are even reasonable, just saying that this information could help estimate a range for overall Lightning payment volume.

Hoping someone can chime in with a better framework for estimating double-counting payments…