Reckless – Chapter 10: Bitcoin’s Risk Free Rate

Chapter 10 of the book Reckless: The Story Of Cryptocurrency Interest Rates is published below. The full book is available on Amazon. The book was written before the bankruptcy of FTX and therefore does not include coverage of this event. However, the book does provide useful commentary in the run up to the failure of FTX, which provides context for the eventual calamity.

The traditional pre-cryptocurrency system has a concept of the risk free rate of return, the return you get from holding government bonds. This interest rate should in theory be lower than all other rates in the economy, which have higher risks. Nobody should lend to a company or individual at a rate lower than the risk free rate. There is a similar idea in Bitcoin, in that you can earn risk free money in some protocols, which could be a floor on interest rates in the cryptocurrency-system. However, these are not totally risk free, they are in theory only counterparty risk free. You can still lose money in other ways, for example by getting hacked.

The Lightning Network Rate

The first and most obvious example of a possible risk free rate in Bitcoin, is providing liquidity in the lightning network. Assuming one runs their own lightning node, there is no custodial relationship with a third party and one can earn lightning network routing fee income. This income is therefore counterparty risk free. However, one needs to have their private key on an exposed online machine and therefore there are risks. For example, the risk that your machine gets hacked and the funds get stolen. There is also the risk that your machine goes offline for an extended period and as a result, you could lose money.

The lightning network is a network of interconnected peer-to-peer payment channels. Instead of using the blockchain to make payments, senders and receivers need to find a path with sufficient liquidity along these non-custodial channels. Only in the event of a bad actor should the Bitcoin blockchain be used as a dispute resolution service. Lightning has several advantages over on-chain Bitcoin payments, namely: lower fees, greater privacy and speed.

One of the main problems with the lightning network is finding a path with enough liquidity to make payments. This makes lightning potentially unsuitable for very large payments. On the other hand, since those who provide liquidity receive routing fees, an equilibrium fee level should be obtained, where there is sufficient liquidity, at a price.

To provide liquidity, all one needs to do is provide capital (in the form of Bitcoin) and run a lightning node. Some users have reported earning an annual yield of up to 1%. Other more expert users have been able to earn even higher rates of return. These are very much variable interest rates, not stable fixed income type returns. If the lightning network becomes popular, one can imagine the kind of economics which could apply. For instance, if other forms of Bitcoin interest rates increase, for example the basis trade premium, lightning network fees may need to increase to compensate investors more for choosing to provide liquidity on the lightning network. Therefore, in a bull market, where futures trade at a large premium over the spot market, lightning fees could be structurally higher.

However, there is no one lightning network interest rate, available for all liquidity providers. Most lightning node implementations have features in place, which will open channels for users and automatically provide liquidity and earn fees. However, the returns from using these are far lower than what is available to the experts who actively manage and rebalance their channels, to maximise the yield. This involves analysing the lightning network, assessing the payment flows and constantly balancing one’s channels. Alex Bosworth is a lightning network developer who is perhaps considered as the leading expert in this field. He disclosed that in the first half of 2021 he made US$30,000 in lightning network fees.

Due to how challenging earning a decent yield is, perhaps this rate should not be considered an interest rate. Again, perhaps Proudhon’s 1849 argument that interest is a “reward for idleness” is not applicable here. At least so far in the lightning network, earning a decent yield requires hard work.

Lightning Channel Marketplaces

Lightning channel marketplaces started to emerge in 2021 and 2022. The idea is that merchants accepting lightning payments will need liquidity and can pay liquidity providers to open channels with them. Marketplaces have therefore emerged, where merchants and liquidity providers come together and exchange cash for liquidity.

Lightning Labs, a key lightning network infrastructure company, developed a service called Lightning Pool. While a startup called Amboss also provides a relatively similar service called Magma. Both these marketplaces are non-custodial. Lightning Pool is designed to aggregate demand and supply for liquidity into a pool and there is an auction process which determines the price of liquidity. Magma is a slightly newer service and at the time of writing is just under one year old. In contrast to Lightning Pool, Magma has more of a peer-to-peer type structure, where entities who desire liquidity can choose the individual liquidity provider they want. The advantage of the Magma structure is that it is easier to select who provides the liquidity. Some liquidity providers are more reliable than others and therefore being able to more easily choose who provides the liquidity has clear benefits. On the other hand, the potential advantages of Lightning Pool are that it could offer better liquidity and it can be more automated.

Magma provides a marketplace which displays the offers (Size and interest rate) and ranks liquidity providers with a reputation score, based on their reliability. Once the liquidity provider is selected, they are paid a fee and open a lightning channel with the buyer. Amboss also publishes average annual yield information on their website. This is the annualised cost to a merchant of buying liquidity as a proportion of the liquidity they receive. At the time of writing, Amboss’ website indicates the average yield based on this metric is 3.27%, based on 46 Bitcoin of deployed liquidity in the history of Magma. Based on data provided by the Amboss team, historically average weekly interest rates peaked at 15% and were reasonably volatile during the first 30 weeks of the service. Since then, rates have settled down and are flattish at around 2.5% to 3.0%.

When purchasing liquidity on Magma the buyer can select the duration they desire and therefore a yield curve can be constructed, based on these different durations. The longest duration available is around six months. Based on historical data, the interest rate declines the longer the duration. This is the opposite of what one observes in a traditional bond market, where longer duration instruments typically have higher yields. The reason for this is that the liquidity provider could withdraw funds from the channel early and therefore the provision of liquidity is not guaranteed. If they do this their reputation will be damaged, but they may still earn the interest income. This risk is reflected in the price. Therefore, although one needs to pay higher absolute fees for longer duration liquidity, annualised yields decline the longer the duration of the deal is.

Due to the length of the duration here, the income in these marketplaces could be more passive than using the lightning network directly to earn routing fees. Therefore, it is possible the market for lightning channels could represent a non-custodial risk free Bitcoin rate of return. However, even this market requires some active management and it should not really be classified as passive income. The market size here is currently very small, Magma has done around 46 Bitcoin of volume and Lightning Pool is around twice that. This is not especially significant in the grand scheme of things, however, it is of course possible that usage of the lightning network grows and these rates become more economically significant.

CoinJoin

CoinJoin is a method of improving privacy when using Bitcoin. More concisely, when spending Bitcoin one cooperates with other users and combines outputs from multiple senders into one transaction. That way it may be much more challenging for anyone analysing the blockchain to determine the flow of funds. If one wants to use this method today, there are essentially two choices: JoinMarket or Wasabi Wallet. Each of these options is a piece of software with a Bitcoin wallet and separate peer-to-peer network like functionality, such that the wallet can cooperate with other user’s wallets, to combine the spending. The implementation we will talk about here is JoinMarket, as a counterparty risk free yield is available. JoinMarket has something called a “yield generator bot”, providing another potential risk free rate. Wasabi Wallet does not appear to have an equivalent feature.

If one wants to use CoinJoin to anonymise one’s coins and achieve stronger privacy, you need to find other peers willing to mix coins with you. However, there may not always be matching demand from somebody else. Different users may have different time preferences with respect to how quickly they want to achieve strong privacy. A user in a rush can either wait for a potentially considerable amount of time or pay somebody else a fee to incentivise them to mix with you. These fees are somewhat analogous to lightning routing fees, in lightning you are paying for somebody else’s liquidity, the equivalent in CoinJoin is that you may also need to pay for somebody else’s liquidity, such that you can mix your coins with them. There is therefore an opportunity for the other side of the privacy trade, the maker, to earn a yield. The JoinMarket Wiki describes the yield generator bot as follows:

Being a market maker allows holders of bitcoin to collect fees. With this, the Yield Generator script is used to earn an income from long-term held bitcoin. The investment is very low risk as the software only signs transactions that are valid and pay operators the correct amount in coinjoin fees. Although the coins must be held on an online hot wallet. The investment has no commitment as bitcoins can be withdrawn at any time. It also improves the privacy of the held bitcoins as well as privacy and fungibility in the entire ecosystem, which makes bitcoin as a currency more useful and thus increases its value.

According to bitcoinkpis.com it appears as if around 2,000 Bitcoin per week flowed through JoinMarket in 2022, although the volume is very volatile, week to week. At current Bitcoin prices this represents around US$2 billion per year in volume. This volume does appear to be quite significant and it has been growing. At the same time, unlike almost every other metric in the ecosystem, it does not appear to be driven by the Bitcoin price, perhaps demonstrating more fundamental non-speculative growth.

One should look at the US$2 billion per annum number with caution however, this may be the same money flowing through JoinMarket multiple times. This is still very much a niche activity and it may not be especially economically significant. In late 2022, there are perhaps around 4,000 Bitcoin at any one time, waiting to participate in on JoinMarket. At the same time, large pools of capital, such as institutional money may not be able to participate in providing privacy liquidity due to regulatory concerns.

As for the magnitudes of the yields one can potentially earn, these are quite low. One can expect to earn an annualised rate of between 1% to 2%, although this variable rate is very volatile. You can earn an annual rate of up to 4% in some periods when demand to mix is strong. These rates do not appear particularly recklessly high and may be quite moderate, for an asset like Bitcoin, which has low inflation and a known final supply cap. JoinMarket could therefore be an attractive option for long term holders looking for counterparty risk free returns. On the other hand, it is technically quite complicated to set up and it does expose the coins to the risk of hacking, as it requires the private keys to be in an online wallet.

Just like with the lightning network, it is possible that usage of the JoinMarket system becomes influenced by the economic cycle. It is possible that in the future, if interest rates rise, privacy becomes more expensive and if interest rates decline, yield hungry privacy liquidity providers may make privacy cheaper. Although the regulatory uncertainty of these privacy systems could dampen the impact here, because capital from certain sources cannot easily flow into privacy systems.

These JoinMarket yields do genuinely appear to be a type of passive income, in that once you have your JoinMarket client set up properly, you can passively leave it on for months on end and earn a yield, without active management and involvement. This is somewhat different from lightning, which requires more active work. Therefore, JoinMarket perhaps offers more of a genuine candidate than lightning for Bitcoin’s risk free rate. Being a maker on JoinMarket is more of a financial type activity than an operating business. However, it would need to grow considerably from the current levels to be significant. For now, both the lightning network and JoinMarket are somewhat segregated from the more speculative, trading driven rates in the ecosystem, which are much more significant. How this may change over time depends on one’s view of the cryptocurrency space. It is possible that the space becomes less speculative and trader driven over time and this privacy related quasi-interest rate gains in significance. For now though, this seems a little farfetched.