Let’s assume that Proof-of-Stake does turn out to be possible and finds usage in a cryptocurrency. When this happens, the days of Proof-of-Work are counted. That is because Proof-of-Stake is inherently less expensive for the community to finance.

The security provided by both Proof-of-Work and Proof-of-Stake originates in the limited supply of capital to all actors. The only thing between a bad actor and a 51% attack is the capital needed to stake the necessary amount of currency or buy the necessary amount of mining hardware. Both Proof-of-Stake and Proof-of-Work are thus essentially proofs of capital.

And since mining and staking are open to everybody, capital will flow in and out of these activities until a risk-appropriate return on investment is reached, maybe 5%. Therefore, this value can be assumed to be the same for Proof-of-Stake and Proof-of-Work. What’s different, though, is the cost incurred by the currency’s users.

Let’s follow a staker (Proof-of-Stake-er) through one investment cycle.  Let’s say that he wants to invest $1M for one year.

He stakes $1M for one year and at the end he collects his, say, 5% return on investment and gets his capital back. The necessary $50,000 are paid for by the users of the cryptocurrency either through inflation (generation of new coins) or transaction fees.

Next, we will compare this to a miner (Proof-of-Work-er) who is making the same investment. He buys mining hardware for $1M and runs his mining operation for one year. At the end of the year, he sells his mining hardware, gets his $1M back and now wants to see 5% RoI.

The problem is that he has spent a lot of money on electricity, salaries, rent and other costs. So, to come out at 5%, he needs to receive much more than just $50,000. And  thus the users of the cryptocurrency will also need to pony up for electricity costs and all the other overhead, in order to achieve the same RoI for miners and thereby the same amount of capital protection as they would have had with Proof-of-Stake.

The users of a Proof-of-Work based currency like Bitcoin need to accept greater inflation and/or greater transaction fees. Which ultimately will make the currency less useful as a way to store and transact money.

After writing this post, I discovered this wiki article which lays out my points with a bit more mathematical rigidity: https://github.com/ethereum/wiki/wiki/Proof-of-Stake-FAQs#what-about-capital-lockup-costs