I've been thinking about the nearly-0 fee environment we've been in for awhile now and am curious about others' opinions. Satoshi said, to paraphrase, "There will either be lots of transactions in the future or none." Well, we're 17 years in and there are basically none to drive a fee market. In a future where the block subsidy has become negligible and second layer solutions have been so successful they keep most transactions off-chain, do you think BTC security could be maintained by people willing to mine at a loss, for whatever reason? For instance, my Bitaxe costs me money to run, albeit a small amount, but I run it anyway because it's fun and interesting. I don't know that I've seen much serious addressing of a sustained ~0 fee environment in the future. I know Peter Todd has proposed tail inflation forever but I don't think this would be seriously considered and I'm interested in other scenarios, if any exist.
pull down to refresh
I wouldn't call a few million, or even a few hundred K sats, nearly-0. That's still rather material in future epochs considering fee rates are already at the lower bound.
There's thousands of transactions nearly every block.
Second layer creates more chain transactions, not less, that's the jevon's effect. Lightning node runners invoke the chain more than they ever would have in its absence: #1437045
Lightning nodes are micro-institutions on a multi-institutional ledger.
Mining cannot be "profitable" over a long time horizon, it's not meant to be. Market dynamics mean profit will always be arbitraged out. Break-even mining, waste-heat generation (space heaters etc), ideological mining, altruism, or as a institutional security COST are keys to decentralization.
For-profit mining is an anomalous condition created not only by the subsidy, but real world transient activity like arbitrage on stranded energy (nat gas wells) and reduced data center footprints for legacy compute (AI is finally squeezing this out), the treasury company fad (publicly traded miners), and demand by institutions for provenance coins (new coin base coins without a chainalysis history that institutional lawyers see as a liability)
That'd render Bitcoin worthless as it undermines it's core premise of infinity / 21M... won't happen.
It's important to think of the chain as institutional technology, institutions have barely arrived. It's not for coffees, unlocking paywalled videos, or payments between APIs. It's for settlement akin to ACH/Wire transfers.
Replacing ACH/Wire transfers alone would exceed its throughput potential, but supply is the more material constraint. There's literally not enough coin to distribute, so a few hundred million institutions at best will have access to the chain.
That's true. I guess what I meant is that despite this, there is no demand driving fees above nearly rock bottom.
These are exactly the kinds of ideas I was looking for with my post Thank you! I was trying to think of scenarios where mining is not profitable.
Interesting idea that it was never meant to be profitable. I mean in most industries, profit gets arbitraged down, but not out completely because otherwise people would stop. What makes mining materially different in this sense? Exactly those things such as altruistic mining, waste-heat, etc?
Bitcoin the industry is different than Bitcoin the network or Bitcoin the money
Since blocks are not a product or service, there's no sales pipeline, nor a moat a producer has to protect margin. It's completely arbitrage-based.
Given the perfect arbitrage for lack of moat or sales pipeline, the non-monetary reasons to mine drive that arb negative. Altruism, waste heat, ideology, provenance premium all add competition to production costs.
That's not to say miners will always operate at a loss, just the reward will have to come from somewhere other than arb we associate with mining today. The industry must and will change.
Focusing on industrial scale and not just plebs running space heaters, the best example is provenance coins being worth more than "used" coins to institutions. If it cost 100k to mine 95k worth of coin, but an institution will pay 105k for the provenance coin, the miner is still up 5k/coin despite mining at a loss.
If a greenhouse spends 100k to mine 95k worth of coin, but saves 10k on their heating bill, they're still up 5k/coin despite mining at a loss.
We can then scale these principles to the pleb with space/water heaters, or to national banks of sovereign countries that don't want foreign countries in complete control of the ledger they use for the world reserve currency. They may mine at substantial loss simply to protect their holdings because they assign it an insurance value beyond the mine value of the coin.
I think it's healthy that fees are at the lower bound, I want to see less industrial mining, more decentralized mining... if fee's were higher it'd favor at-scale mining who can better arbitrage it.
I always think what if blocks become so full fees make mining a really lucrative business. Plus a million sats in today’s purchasing power can be vastly different 50 to 60 years from now.
If this protocol is going to become the world’s financial operating system I think your opinion on mining not being profitable is a bit bearish on the economics of the entire protocol
There's no economic rationale for it to be profitable over the long term, since there's no moat to prevent competition and the resultant race to the bottom.
If fee rates spiked, that would just bring more miners online, thereby increasing overall hash rate and driving the cost of mining higher until it's unprofitable.
It's all just math, there no branding or sales channels, no regulatory capture. Anyone can arb out the premium.
Access to ASICS and energy have been the source of any margin, but mentioned AI has squeezed the energy/footprint arbs, and ASICS have become increasingly commoditized over time... especially with AI also increasing the number of fabs.
The waste-heat/ideological/security dividends take otherwise an perfect break-even arb opportunity and drive it negative.
Thinking mining has to be profitable is bearish, that assumes Bitcoin only exists as an arb play, repackaging capex and marking it up to gain fiat. My outlying takes are because i'm unequalled in bullishness.
I'm not a miner, but my understanding is that something like a BitAxe gamma is 100x less powerful than a typical industrial miner. So I would think the hash rate would be quite a bit lower than it is now if we were only relying on hobbyist miners.
Maybe there is a world where such miners get included in heatpunk applications (space heaters, hot water heaters, anyone working on a bitcoin miner oven?) and there are such devices are so widespread that a significant amount of hash power is generated this way, but it's hard for me to imagine that actually happening.
I suppose the important thing for bitcoin security is the relative hashpower -- if there aren't very many industrial miners out there, then perhaps the hobbyists provide enough security. But in that scenario, I wonder what has happened to all the ASICs...
I don't think existing electronics like working at oven temperatures, and the miners would need to reach a higher temperature in order to create a gradient driving heat towards the oven.
Well, I guess we'll have to settle for beef jerky dryers...
I would rather mine at a small loss that gets more than offset by Bitcoin's increase in purchasing power over time than have tail inflation. However, I do not know if there are enough people willing to do that to secure the network.
I could also envision a situation where like insurance for your house people purchase hashrate as insurance for their stack.
incentives oppose hardforks that make the economic fundamentals worse. there are already plenty inflationary altcoins, anyone who wants to use them for payments already is, and anyone who tried holding them as investments has probably cut losses and returned to harder currency.
kudos for mining at a loss.
Right, that's why I don't think hardforking to tail inflation is feasible. Have you thought about how this hypothetical future might work in light of that as a non-starter?
there have been at least two hardforks[1] that created altcoins, both extending the block size. Both collapsed in value, the second one much more rapidly than the first, and neither of them even touched inflation. I suspect a fork adding inflation would get sold off in a similar way.
wikipedia lists four, although two of them were forked from the first fork, and only two are direct forks of Bitcoin. ↩
This gets at the hardest part of Bitcoin's long-term design. The subsidy halving schedule is deterministic—by around 2040-2050 you're essentially at zero inflation, and miners need fees or they leave. We're betting on two things working: (1) Lightning and other L2s actually scale to handle most economic activity, keeping onchain transactions rare and therefore valuable, and (2) people value settlement security enough to pay for it. The fee market only works if both happen. The risk is real—if transactions remain cheap and plentiful, there's no fee pressure to secure the base layer. I've spent years modeling different accumulation and security scenarios, and what I found is that tracking actual transaction patterns and fee evolution matters more than price speculation for understanding where this heads. The next 10 years of L2 adoption data will basically answer your question.
pool.blvrg.net