On Fees

On Fees

When sending less than €1,000 of Bitcoin it’s worth paying attention to fees, but keep mind that your payment is competing with transactions that move €100,000 on equal terms. Transactions are charged per byte, not as a percentage of the amount. But willingness to pay is obviously a percentage of the amount.

It’s interesting to note that although the price has increased more than 10x over the past year, transaction amounts in BTC terms haven’t changed much:

Neither have fees as a percentage, they still hover around 0.75%:

 

Why is this? Perhaps it’s because many hodlers became rich and got comfortable moving 10x the amounts around that they were used to. Or perhaps higher value activity entered the ecosystem, pushing lower value out, or users became more efficient.

The use of SegWit addresses allows for 50% cost savings, yet its adoption is stalling around 12%. This is likely because major services haven’t upgraded yet, and that may be partially due to them being distracted by massive user growth due to the price rise. But then, why aren’t people switching over to competitors that already offer SegWit support?

Percentage of transactions that use SegWit according to segwit.party

Is the market for wallets so inefficient that users don’t churn in order to save 50%? It’s understandable that free wallets don’t have advertisement budget to get this message accross, but not all wallets are free. If we were to take Erik Voorhees’ $40 fees claim at face value (which you shouldn’t), then Ledger could advertise their wallet by saying you can earn it back after using it thrice.

In that case, the following affiliate link should make me rich:

Save 50% on Bitcoin Fees Now!

Why aren’t they sold out? Why aren’t there $25 e-courses on saving transaction fees? Why isn’t Youtube full of advice to save on fees, targeted at non-technical folks? Most, but not all, of the content around this topic serves to promote investment offers, an obviously very lucrative market. But at present 500 BTC per day levels, fees are becoming a $2.5 billion market. You’d expect that to attract some entrepeneurs as well. Where are they?

Why are people not flocking to coins with low fees, despite good wallet support and descent liquidity? BCash has about 29,000 transactions per day, Litecoin 77,000 transactions vs Bitcoin’s 220,000. With no mechnism to prevent spamming, because those coins are not at full capacity, those numbers are probably optimistic. Pareto’s principe suggests that there should be far more low value transactions than high value transactions, so if there really was huge demand for low-fee currencies, one would expect these altcoins to have far more, not fewer, transactions.

There’s much room for improved efficiency to take better advantage of the existing block space, but a more pessistimic interpretation of what’s going on is that the majority of existing Bitcoin users don’t care about fees. And if that’s true those fees won’t go down no matter what we try in terms of better coin selection, SegWit, Replace-By-Fee, transaction batching, etc. The people moving these large amounts of money won’t bother using these techniques, thus freeing up space for others, unless perhaps we make it trivial, default and beg or guilt-trip them into using it.

Which brings me to Drivechain, which offers an interesting advantage in this situation that Lightning doesn’t have (yet?). In order to use Lightning you need to open a channel first and you need to close it later. In addition to that, you first need to actually receive those coins from somewhere else. Drivechain on the other hand, is more similar to altcoins in the sense that any user can just generate an address and buy the sidechain-coin directly, very much like how they buy bitcoin. It decouples the usage of layer 2 from the task of moving between layer 1 and 2. That allows economies of scale for moving between layers, i.e. far fewer transactions. That is, if I understand Drivechain correctly; it’s hard to wrap my head around. This interview is a good place to start:

Of course these technologies don’t contradict each other and Lighthning is much closer to actually being deployed. It also benefits high value transactors due to it’s orders of magniture improvement on speed of settlement, giving them a stronger incentive to move off-chain than SegWit did. It’s also possible to unilatterally fund a channel, so exchanges can do for every new customer, allowing them toreceive their first coins directly on a Lightning channel (someone still needs to pay fees though, even if the channel is never closed in practice).