Our only URLs are

All other sites are scams – especially be wary of:

benumbs.cards & bennumb.cards & bennumbs.cards & benumb.cc & many more…

(it can be hard to notice the S and extra N if not careful.) 

Welcome to the real deal. 

Please bookmark this link — the other sites have simply copy/pasted our html and don’t actually have any cards to sell. 

They can be easy to fall for if you aren’t cautious!

Influence of 30 extra halvings

I appear to recollect each halving being approached with some trepidation as to how the miners would adapt. Traditionally, value goes up and extra hash energy will get turned on. Can this occur one other 30 instances?

In my easy understanding, every halving will double the price of manufacturing. For many commodities, the worth will attain equilibrium close to the associated fee or manufacturing. Assuming 30 extra halvings, assuming a present price of about $15,000 per BTC (which I believe is a low estimate), this places the bitcoin price of manufacturing at ($15000 x 2^30) or over $16T per BTC in 2140.

This appears form of absurd. Particularly because it assumes the problem and hash fee keep fixed to what they’re at present in 2022. Given the trajectory of hash fee, it must be protected to imagine the problem will likely be a lot greater in 2140 than in 2022. This is able to imply the price of manufacturing is even greater than $16T, and the market value would must be greater than $16T/BTC to help this. That will put the market cap at 21M x $16T or $336 Quintillion. Once more, an absurd quantity as that is in t*odays {dollars}* (inflation would make this even greater).

One other chance is that the hash fee reaches a most after which begins to drop off. If the problem is simply too excessive, miners should drop out of the race and scale back the problem (and price) to be below the market value. This is able to result in centralization, or at the least a discount within the variety of entities controlling the mining hash fee.

I perceive that sooner or later, a lot of the block rewards will likely be in transaction charges, however how does this impression the general economics of price of manufacturing? The charges are at present working round 1-2% of the block reward, although when the community is busy it has spiked to 30-40%. From what I can inform the charges will keep comparatively the identical in absolute BTC earned, however the coinbase reward will get lowered with the halvings. This looks like it could put a cap on price of mining a block thats one thing like 100x what it’s at present.

So this could put price of blocks round $1.5M, and since that is for six.5 BTC at present… that might cap it round $9.75M cap to BTC? Extra real looking however unsure the maths holds up…

5 thoughts on “Influence of 30 extra halvings”

  1. I think you’re extrapolating too much. 2140 is over 100 years away. who knows how things will change with bitcoin and with power generation as a whole. Even today we’re getting cheaper and cheaper energy whether it’s from nuclear, renewables or fossil fuels.

    you can’t know the future too far, so we need to think about the immediate issues, let the next generations figure that all out

    Reply
  2. Money can change within 50 years for we all know. Bitcoin is a solid back contestant to fiat, and is coded deflationary as a fundamental basis. Whether we use it past 2100 or later isn’t the important part, but the absolute credibility, scalability, decentralization, and deflationary coding make it the best alternative to fist and gold.

    Scandals will yield regulations, regulations will attract retail and VC. trust the process

    Reply
  3. The beauty is the difficulty adjustment roots out all the miners with expensive energy during the bear market. I have strong confidence in 120 years (30 cycles) BTC mining companies will be only using very cheap stranded energy, if not way way before then.

    Reply
  4. I believe your mistake is the belief that cost of production will double with each halving. I’m faily sure the difficulty of mining changes with the computing power used, so if mining becomes unprofitable, the difficulty and profitability will reach an equilibrium as you were saying. I remember reading a description similar to this on this subreddit within the last week.

    Reply
  5. The hash power doesn’t control the price. The price controls the hash power. It’s far more likely that at the halving, the hash power will drop quite a bit. That is, unless the price also goes up at the same time.

    Reply

Leave a Comment