Nearly 15 years after its creation, no one, apart from some, questions the viability of bitcoin any more. Everyone, even in traditional finance, agrees that the first cryptocurrency in history is here to stay.
This summer, the boss of BlackRock, which is none other than the largest asset manager on the planet ($9 trillion in assets under management), explained all the good he thought of bitcoin , which in 2017 he still considered an asset for "crooks".
"It's an international asset" that could "overtake national currencies", Larry Fink told US television.
The biggest cryptocurrency on the market ($500 billion in capitalisation) is, however, still a long way from becoming part of our everyday lives 👀.
While people invest in bitcoin, few use it as a means of payment to do their shopping or buy products of any kind. And for good reason, apart from the fact that few shops accept bitcoin, its network is not efficient enough. Not 'efficient' enough in the sense that it can't handle too many transactions.
It can only handle 7 transactions per second (that's theoretically more than 40,000 for Visa), and if there are too many, it becomes congested and transaction fees soar. Who would want to pay €10 in fees for a €5 or €6 transaction? Nobody 🤯. Not to mention that you have to wait around 10 minutes for a transaction to be recorded in the blockchain...
It was to solve this problem that Thaddeus Dryja and Joseph Poon came up with the Lightning Network (LN) a few years ago. They presented their solution in 2015 and published the Lightning Network white paper in 2016 (available here).
The Lightning Network has been available since 2018. The idea is simple: rather than recording every transaction on the Bitcoin network, the Lightning Network allows its users to make transactions OUTSIDE the Bitcoin blockchain - these are offchain transactions - by creating parallel payment channels.
These channels then enable instant, low-cost transactions between participants. As transactions take place, the balances in the channel are updated. Participants can adjust the balances based on their transactions.
"Lightning Network is a second layer of Bitcoin that handles all the small transactions that Bitcoin won't," explains Josef Tětek, Bitcoin analyst for Trezor, a Czech start-up specialising in digital asset custody.
When they wish, participants can close their channel. Final balances are recorded on the Bitcoin blockchain. Transactions made outside the channel are consolidated into a single transaction on the Bitcoin blockchain, and that's it 💡 !
Before we go any further, just a few figures about the Lightning Network 🤓.
Today, there are just over 16,000 nodes on the LN network (almost 47,000 nodes for the Bitcoin network).
Nodes are essential to the network because they enable all transactions to be verified and validated. The more computing power (with computers) you mobilise for your node, the more essential it will become. This is why the most widely used nodes in the LN network are those managed by companies like France's Acinq - in which Bpifrance has invested in 2019.
The thousands of nodes in the network feed 68,000 public payment channels. In these payment channels, there are 4,600 locked bitcoins, or roughly $120 million (at the current price of bitcoin), which partly ensure liquidity on the network.
According to estimates, there would be between 10.000 and 15,000 additional payment channels that are not declared, and therefore not public.
That's it for the figures!
So let's get back to how the Lightning Network system works, which therefore seems, at least on paper, perfect! Except that in practice things are a little more complicated 😅.
"Some people have made too many promises about the Lightning Network, saying that it would make it easy to send a few satoshis (1 satoshi = 0.00000001 bitcoin) instantly, when that's not the case today," explains Simon Glâtre, data analyst at Sato, a publicly listed Canadian company specialising in bitcoin mining.
Of course, not everyone shares this opinion (which is to be expected).
For some, such as the French company Acinq, which developed one of the 4 main implementations of the LN (which is called Eclair), the LN has made enormous progress. "Bitcoin and the Lightning Network represent a real paradigm shift, so it's only natural that it should take time to arrive at a system that is easy for everyone to use," explains Pierre-Marie Padiou, co-founder and CEO of Acinq.
In the meantime, to be more effective, the Lightning Network needs to improve on a number of points:
👉 Liquidity
You have to imagine the Lightning Network as a system with balls (the satoshis) and tubes (the payment channels). To make a payment, you have to push marbles into the tubes.
As long as you have marbles in your wallet, you can push marbles and therefore make payments, but it gets complicated when you run out of marbles in your wallet. "You simply can't make any more payments while you're waiting to put money back in," sums up Simon Glâtre of Sato.
To make the system run more smoothly and avoid this kind of problem, several players have entered the market: the Lightning Service Provider (LSP).
These players are an interface between the user and the Lightning Network. "The role of LSPs is to make users forget the fact that they have to manage their liquidity and always have marbles in their wallet," explains Josef Tětek of Trezor.
There are several types of LSP:
- there are those that are only LSPs such as Lightspark (the start-up founded by David Marcus, the former brains behind Facebook's Libra project), Voltage or LNBIG.
- there are the LN wallets that have decided to create their own LSP. This is the case with Phoenix (developed by Acinq), Breez or Bitkit.
There are currently around twenty LSPs on the market, which can supply Lightning Network users with satoshis when they need them. In practical terms, you want to send 100 euros (roughly 420,000 satoshis, at the current rate) to a friend, and a wallet like Breez will automatically provide you with the liquidity to 'push' your satoshis to your friend's wallet.
The only problem is that there is still a question about the volume of liquidity provided by the LSP. Because if the LSP doesn't provide enough liquidity, the user will very quickly be blocked, and a new channel will have to be opened to pay more money, which costs money in transaction fees. This system is clearly anything but optimal!
To improve matters, the 4 different implementations of the Lightning Network (Eclair, C-Lightning, LND, Rust-Lightning) are working on an improvement: 'splicing'.
This new feature would make it possible to have a single payment channel between the user and the LSP, whose capacities could be increased or decreased as required. Splicing is currently being implemented on Eclair and Core Lightning (developed by Blockstream).
Another feature is expected to improve liquidity management: trampoline routing. The idea is to delegate the search for the best path for a transaction to the largest nodes in the network, which should make transactions even smoother. Trampoline routing is already implemented in Phoenix, but only partially. "We are in the process of deploying it", confirms Pierre-Marie Padiou.
👉 Privacy
As we indicated above, there are currently just over 16,000 nodes on the Lightning Network. Every transaction passes through these nodes.
Today, the person initiating a transaction on a payment channel chooses the payment route, and the person receiving the transaction must give the address of his wallet to receive the payment. With this system, however, it is quite "simple" (although everything is relative) to start from the address of a wallet and trace all the transactions of a user.
To improve things, several developers have developed the "blinded paths" functionality. As the name suggests, the aim of a 'blinded path' is to muddy the waters. The idea is that the payment route is split in two: the first part is defined by whoever sends the funds and the second by whoever receives the funds. Pretty clever!
👉 Asynchronous payments
Anyone who has ever made a transaction on the Lightning Network knows this: you can only send money to someone if they are connected online to receive the funds 🙃.
When you physically pay in a shop, it's pretty straightforward as the merchant will be connected, but the situation is more complicated for remote payments. "If I'm not connected with my wallet, no one can send me money," Simon Glâtre points out.
This system is fairly inefficient, and to improve it, a new "asynchronous payment" feature is being created.
The idea of this feature is to allow you to make a payment that will be on hold and will be completed as soon as the other person connects. "Today we can have funds stuck on the network for days because someone hasn't connected. This is a real problem. It's one of the features that will change things the most," explains Simon Glâtre.
👉 Wallets
Last but not least, there's the subject of wallets, which is undoubtedly the most important topic. And why is that? Because it raises the philosophical question of holding bitcoins.
Today, there are two types of wallet on the market: custodial and non-custodial (you are the holder of your funds).
While most of the community would like to hold their bitcoins themselves on a non-custodial wallet, it is clearly custodial wallets that are the easiest to use. The best known of these is the Wallet of Satoshi. "It's fairly simple to use and the advantage of custodial is that you don't have to manage anything apart from your transactions," explains Lounès Ksouri, an engineer at LNMarkets and Lightning Network specialist.
The downside of this system is that the Wallet of Satoshi which holds your bitcoins. "With them, you don't actually own the funds. In reality, you are not using the LN, but a centralised player who is going to do it for you. Wallet of Satoshi has a big node and they have plenty of payment channels," Simon Glâtre points out.
On the non-custodial side, there are plenty of wallets like Phoenix from Acinq, but to use them you need more control. "You have to open your channels yourself," explains Josef Tětek. "You can't want to be sovereign over your cryptos and not manage anything yourself," counters Pierre-Marie Padiou.
In an attempt to find a happy medium, some are arguing for a hybrid system that would be based partly on custodial: federated wallets.
The aim of "federated wallets", like Fedimint, is to take the best of custodial (efficiency in particular) and the best of non-custodial with the protection of privacy. The idea is to designate a custodial body for a limited number of holders.
"One of the obvious applications is the family circle," explains Simon Glâtre. But it could also very well be managed on a larger scale, he stresses. But it's not certain that everyone will become a fan!
