B I T C O I N
The Lightning Network is Growing Rapidly? Rather not.
How the hype around Bitcoins new technology is just hype and nothing more
Sometimes it is said that the lighting network on Bitcoin is experiencing rapid growth. That’s not really true. The statistics rather show that the growth of the Lightning network has stagnated for almost three months. What’s going on?
The Lightning network is considered Bitcoin’s future: a payment network that solves all the shortcomings of on chain transactions on Bitcoin, bringing infinite scalability, greater privacy, and instantly finalized payments, all of which skeptics often miss about Bitcoin. Lightning has been hailed as this final solution to many problems for three to four years; since early 2018, the off chain network has spread to Bitcoin.
One would think that Lightning would grow rapidly with the urgency and importance of the company. Occasionally this impression is also created. But when we look at the Lightning statistics, we see exactly the opposite — a network that has apparently been stagnating for almost three months.
Charts, Charts, Charts
We start with probably the most important and unambiguous charts:
Here we see the number of channels (yellow) and the total capacity (green) of the Lightning network on the 6-month chart. As you can see, the network jumped to about 43,000 channels and a capacity of about 1060 Bitcoin in mid-March. Since then it has stagnated, with a rather declining trend. There has been no growth in either the number of channels or the capacity of the network in the last three months. The chart of P2SH.info is confirmed by the charts on bitcoinvisuals.com, for example.
After all, there is one value that continues to grow: the number of Lightning-Nodes. However, it is also noticeable here that the growth has slowed considerably since mid-March. While the number of Lightning-Nodes rose from about 1,900 in December 2018 to about 4,000 on March 21 — doubled in about three months — only about 250 new nodes were added in the following three months, which does not even correspond to a growth of 10 percent per quarter.
The necessary consequence of the two charts is that both the capacity and connectivity of the nodes have decreased: More nodes provide fewer channels.
This chart is not easy to understand. The orange line shows the average number of channels a node has — 15.8. The blue curve shows the number of nodes of the “90% percentile”. I am not sure enough how to interpret this to speculate here. Probably it means 10 percent of the nodes with the most channels. If you have good hints, please share them in the comments. The green curve, the 50% percentile, represents the median and has dropped from 5 to 4 in recent weeks. The extreme difference between average and median indicates that the number of channels among the nodes is very unequally distributed: a few nodes are very well connected, most of them rather loose.
In this constellation, the “density” of the network necessarily decreases. In the concrete case, density means the ratio of possible and actual channels. Of course, a node can’t create every possible channel; but the ratio of it probably represents the density indicator.
A no less interesting statistic is that of the “cut nodes”: The nodes that connect certain areas in the network and whose shutdown would cut a path between these areas.
The blue columns show the absolute amount of cut nodes, the orange curve its the share of all nodes. This shows that the pure number of cut nodes is constantly increasing, which is not surprising since the number of new nodes in the network is also increasing. Until the end of March one could observe that although the absolute number of cut nodes increased, their relative share of all nodes decreased. Since the end of March, however, the absolute number has accelerated, and the relative number has also risen from around 5.7 percent to 7.5 percent at present.
It is difficult to understand exactly what this means: Is the sphere of the “influential” becoming decentralized since nodes are important for scouting? Or is the network losing its inner density so that it is less resilient to the failures of individual nodes? Rather than speculating on such questions of statistical network theory, we speculate on the background of this tendency.
Another important data source is some comments from LNBIG on Reddit. LNBIG is the most important liquidity provider in the lighting network. It operates about 20 nodes with numerous channels and bitcoins. If you wish, you can ask for liquidity on the LNBIG website. The great leap in a capacity that the Lightning network made at the beginning of March — from just under 800 to a good 1000 bitcoins — is mainly due to LNBIG.
The operator admits that LNBIG is the source of the current decline in capacity and channels because it has closed many channels. “But I would also like to say that the now-closed channels didn’t work at all … it’s difficult to understand if you have to open a channel to another node or if it works without it. I would say that there will be more quality than quantity from my side now.” To be honest, LNBIG would like to close more channels, “but I don’t want to do that to keep the LN community from getting into trouble”.
The situation is paradoxical: “When you open a lot of channels, everyone complains that you’re taking over the network. If you close the channels — then there are complaints.” LNBIG would be “very happy” if other players opened more channels and provided the network with more liquidity. “But that’s almost the strangest thing to happen.” After all, LNBIG’s share of the network is declining. He has about 473 Bitcoin in his channels, which is almost half of the liquidity in Lightning.
Another very interesting comment concerns his earnings. What does the largest liquidity provider in the Lightning network gain by forwarding transactions? “I have about 200 to 300 transactions a day, rarely 600. I get 5,000 to 10,000 Satoshi a day in fees, about 0.4 to 0.8 dollar cents, which is a maximum of 20 dollars a month.” Opening and closing channels have cost him more than $1,000 in on chain fees so far. “Therefore, I do not earn anything so far. I’m doing this more to support the network.”
Enormous media attention and a dilemma
Let’s get this straight: Both the number of channels and the capacity of the lighting network is stagnating or declining; the largest liquidity provider is processing just 200 to 300 payments a day, suggesting that there are barely more than 600 lighting payments a day in total; and there seems to be no way of earning by providing liquidity to the network at the moment.
This contrasts quite strongly with the media attention the Lightning network enjoys. On our blog there have been 6–7 articles about Lightning since the beginning of March, among the colleagues of BTC-Echo you can find more than 40 articles under the search word “Lightning” since the beginning of March, of which at least 20–30 are explicitly about Lightning. With the English-language Bitcoin magazines I stopped counting after 30 hits with Google in this period, with Coindesk there are 30 hits with the search for Lightning since at the beginning of March. On average there are at least 2, rather 3 articles about Lightning per week in the big Bitcoin media.
How can all this be explained? On the one hand, to become a bit of a meta one could justify the gap between media coverage and actual growth with the fact that the Lightning network is still the only really present glimmer of hope for the Bitcoin media. After all, Bitcoin on the chain, i.e. without Lightning, has little chance of growing seriously. This weakness can easily be ignored if you focus on bringing good news to the Lightning network.
On the other hand, one could try to explain the weak performance of Lightning without questioning the concept in general. So Lightning remains a (theoretically) perfect solution for the scalability problems of blockchains. The software is currently still beta, and building a new network is hard tedious work. As long as you can’t pay much with Lightning — there is no killer app yet — there are few reasons for users to use the off chain network; and as long as there are few users sending transactions via Lightning, there are few motives for liquidity providers to distribute bitcoins in channels.
This is a kind of dilemma: the users need liquidity providers, and the liquidity providers need users. If there is neither one nor the other, it is difficult to get started. However, this relationship can change quickly. All you need is a strong, convincing Lightning application, such as the ability to deposit Bitcoins in real time at stock exchanges or Bitcoin ATMs, in order to boost both user numbers and liquidity. Once the avalanche starts rolling, everything else could happen by itself.
It should also be mentioned that the available statistics do not cover the entire lighting network. There are “private” nodes that do not route transactions, but merely maintain a one-sided channel through which users can pay. One example is wallets that can only send Bitcoins via Lightning but cannot receive them. They should be missing in these statistics. It could, therefore, be that the actual number of channels is higher than shown here. But how it behaves with it concretely, I cannot say, unfortunately.
Finally, it should be mentioned that Lightning — in its current state — is not the most user-friendly method to pay with Bitcoin and to receive payments. It works well if all users are prepared and you only send small amounts of less than one Euro. But as soon as you try to receive money independently from payment service providers, it becomes tricky and you will inevitably have to deal with liquidity complications. If payment is then even greater than 20 euros, the Lightning network will, in my experience, be extremely unreliable. None of this is suitable for attracting new users or developing business models that involve a large number of people forming a large number of transactions.
That’s why you should be fair and honest at this point to admit that Lightning is an experiment after all. There is still work to be done to make it as reliable and user-friendly as Bitcoin (or other cryptocurrencies). And just as it is not possible to guarantee that Lightning will be a success, it is also not possible to declare Lightning a flop based on the current numbers. It still takes time.