Historically, miners have neglected transaction fees – and with good reason!
In the past, the microscopic fees meant almost nothing in comparison to the massive block reward. During 2009-2013, fees were essentially $0. The subsidy was 50 BTC per block. But now, the block subsidy is 3.125 (down -94%). It will continue to fall (at ~16% per year, compounded), until, finally, it reaches zero.
This is an op-ed by LayerTwo Labs CEO Paul Sztorc and does not represent the opinion of Blockspace Media.
Painful though this fall is, it has been offset by the skyrocketing price of Bitcoin — up roughly 1,000,000x over the past 16 years. But this, too, is unsustainable. The sober reality is that the price of Bitcoin cannot go much higher from here. Another 200x would take it to $25 million per coin, at which point Bitcoin would be more valuable than all money in the world, combined. (We should certainly take that as our goal!)
But: that’s the ceiling. Even if we attain that lofty goal, the entire 200x ROI it provides will be wiped out (for miners) after 28 years. Even at fifty million USD per coin, in 2080 this translates to $9,536 per block (and that’s in future 2080 dollars – probably enough to buy one ham sandwich).
So – like it or not – the age of transaction fees has finally arrived!
Most future bitcoin mining revenue will come from merge mined L2s
To understand this new age properly, we must appreciate two things: merged mining and layer 2s (L2s).
Merged mining was invented by Satoshi in 2010, and it allows miners to find blocks on multiple networks at once. Historically, Bitcoin miners used it to mine, for example, Namecoin. These merged-mined Altcoins have increased revenues by only a tiny amount, less than 1/100th of 1%.
But that story changes, with the merged mined Bitcoin L2. This is an L2 which pays transaction fees to bitcoin miners. In contrast, non-mined L2s (such as the Lightning Network, FediMint, ARK, and Liquid) do not pay their transaction fees to miners.
A Bitcoin block can only contain 2000-3000 transactions. At $0.10 per txn, the txn fee revenue is limited to $200-$300, per block. In contrast, the L2s have unlimited throughput. I myself recently tested some, achieving a total transaction rate of 15.6 million L2 transactions (per L1 block). At the same fee-rate of $0.10 per txn, that would be $1,560,000 in revenue (plus the $300 from on-chain L1 fees, and the block subsidy of $350k).
This is why miners must shift their focus. L2s are where 99.99%+ of the money will be made — not the block subsidy (which is dwindling down to nothing), nor the on-chain blockspace (which is capped at 4 MB). Billions of USD in annual transaction fee revenue is at stake. If miners do nothing, the L2 space will probably be captured by the non-mined L2s — that revenue will go to “Lightning Service Providers” (or Fediment multisig holders, etc).
Large ASIC-manufacturers (namely: Bitmain, MicroBT, and Canaan), should care most of all. In the future when transaction fees are the bulk of mining rewards, a newly produced ASIC is only worth the transaction fee revenues it expects to bring in. In a world of static fees, the “pie” (so to speak) is fixed at a small size. As more ASICs are sold, the “pie” is cut into smaller and smaller pieces – each ASIC must be worth less and less. However, when transaction fees grow, the pie grows larger for everyone. Everyone can pull in more money, each block – even if everyone’s percentage of total hashrate is diminishing.
Bitcoin miners, please rescue us from bad L2s!
Mined L2s are also better for Bitcoin as a whole. When mining revenues are high, hashrate is high as well. The network has ample proof of work, enough to deter 51%-attacks and double-spend attacks. It aligns miners’ incentives with the interests of real users who transact on the network every day. When people use the real Bitcoin network to transact, they are more likely to care about the network’s health. They are more likely to run a node, and to care about issues affecting the network.
In contrast, the 10-year experiment with non-mined L2s seems (to me) to have been a failure. Lightning (the crown jewel) received over $100 million in funding, and near-unlimited praise from our community. Tens of thousands of elite developer-hours, if not more. The result is a big disappointment – see for yourself what lightning developers are saying these days.
Even if a non-mined L2 ever succeeds, it would set the stage for future conflict. Miners will fight the L2s over those fees (I presented on this topic at OP Next). That’s dangerous for Bitcoin.
ASIC manufacturers should take the lead
To support mined L2s, I created the “Bip300 CUSF activator”. This is a free piece of open source software that miners can run at any time. (They can also abandon it at any time). If >50% hashrate runs the activator, then new mined L2s will be on equal footing with the Lightning Network. These L2 can launch and attract users – and their transaction fees.
How might this happen, in practice?
It is hard to contact each individual miner – there are too many. That’s good for decentralization, but in this case it creates a collective action problem.
Would pools help? After all, there are only 15 or so. Pools are easier to get in touch with. Pools have a brand – and accountability.
Unfortunately, pools are sometimes punished for doing the right thing. Last April, Foundry emailed members about the upcoming “epic sat” ordinal. However, by then, the word “ordinal” had become associated with the (unrelated) Bitcoin Magazine 2023 ordinals controversy, leading to angry emails from members. Foundry’s leadership was distressed by this, and, so I was told, adopted a policy of “never having an opinion on anything ever again.” Ultimately, ViaBTC mined this epic sat and sold it for 33.3 BTC, netting their members (in effect) 10 extra block rewards. Foundry’s members missed out! But a genius can only do so much, if he has a fool for a master.
So, I repeat: the large ASIC manufacturers should take responsibility for supporting mined L2s. For them the opportunity is highest. They should research this issue, and then use their clout to educate their end-customers. Finally, those end-customers can convey their opinion to pools.
