Then it is not the sunk cost of the mining that drives it in your model, but the cost of new mining. (the sunk cost is mere history).
Also, it is not clear that having more or fewer miners on the network drives prices - it seems more likely that more miners reduce transaction/confirmation times, improving service levels and increasing value (to the extent that actual transactions drive value or price).
> Also, it is not clear that having more or fewer miners on the network drives prices
Of course it is. If it costs a miner $10k in electricity, rent etc to mine 1 BTC, they will need to sell at least $10k worth of BTC to cover their cost. If they stop mining, they will have less pressure to sell.
Miners hold over 2mm BTC so hard to see how their actions have no direct impact on price.
My point, is that when electricity costs approach zero, more people will mine which will impact prices.
But it seems that the cost of electricity can drive down prices, not only up.
You said that price of electricity 'backstops' prices, and that miner spending $10K on electricity will need to sell that much BTC. Seems more like an inverse relationship. They need to sell $10K whether the price is $100/BTC or $100K/BTC, so lower BTC value makes more selling. Similarly, higher elec costs increase ready supply of BTC on sale, more supply lower prices.
On the other side, lower elec costs, mean more hoarding by miners, reducing sell-side pressure/supply
This would seem to make elec costs inversely related to BTC dollar value, not supporting it - so an effect, but in the other direction?
Also, it is not clear that having more or fewer miners on the network drives prices - it seems more likely that more miners reduce transaction/confirmation times, improving service levels and increasing value (to the extent that actual transactions drive value or price).