You handwave through your main point. Why isn’t the security model viable?
Fee revenue can stay constant in Bitcoin terms (~25M sats/block) and totally pay for the security of the system. Today that means a reward of ~6.5 Bitcoin/block. In 5 halvings, that means ~44M sats/block.
So all Bitcoin has to do to meet today’s security budget is hit the purchasing power equivalent of ~$450k today… by 2040.
I somehow think it’ll do far better than that.
In the 50s the subsidy will go away. It will get alot smaller before then, especially in the 20s and 30s. That means that the order of the blocks is controlled by hash roughly equivalent to the total daily fees. If you control the hash and are not a good actor you can peform deep reorgs which will be disruptive to users and markets. Under the current security model it wont cost that much to do, and its effect on the market cap could be high. So we have to consider what to do about that. Is that less hand wavy?
Yes, less hand wavy, and it comports with what I thought you were saying. Thanks.
But I’m not sure it overcomes my point. In 5 halvings (~2040), the subsidy is ~19M sats. Holding the fee at 25M sats, that fully covers the current security budget at a PPP of $450k/₿ in today’s dollars.
In 10 halvings (~2064), the subsidy will be 610k sats/block. A few of 25M sats/block would need to have a PPP of ~$761k/₿ in today’s dollars for the same budget as today.
At the limit (0 subsidy, in 2141), the same budget is maintained (for a fee of 25M sats/block) at a PPP of $780k/₿.
That seems a rather bearish PPP price point for 2040, 2064, and 2141.
If the PPP rises faster than this (which pretty much everyone in Bitcoin believes), then the security budget increases relative to today. Also, if total block fees go up, the security budget increases. Of course, missing these to the downside means the budget goes down, relative.
Drivechains are argued to increase the net block fee, but they’re not the only thing which can do that. And many argue that they do it in a potentially negative manner. Other techniques will also increase the net fee (or reduce the per-effective-tx fee, such as LN channels). As adoption increases, that will also bid up the fee, but PPP increases should act to push it down. So I feel 0.25₿ (25M sats/block) is pretty reasonable, long term.
So, I’m not yet convinced that the security mode is as fragile as you are suggesting.
