1/ The problem with Proof of Stake: once more, with feeling.
Here’s a comparison of some economic properties of Proof of Stake (PoS) coins and Proof of Work (PoW) coins. Many PoW/PoS comparisons focus on tech aspects, costs of 51% attacks etc; this thread is more about economics.

2/ PoW: miners expend a lot of electricity to find random numbers that allow them to chose the next block and claim the block reward. To pay for electricity (and other expenses) miners need to sell these rewards (sometimes they sell all the rewards and go bankrupt anyway).

3/ With a PoW coin, every marginal price increase also results in an increase in mining reward (in fiat terms); this brings in more miners, and the miner’s collective expenses rise as well, so coins representing a higher value are dumped on the market.

4/ Thus the amount of investments that are required to sustain the new higher price increases as well (these dumped coins have to be bought by someone).

What strong arms have conquered, strong arms must defend.

5/ The fact that a coin’s price rises despite persistent, increasing selling pressure also shows that there is a persistent demand for it. It tells investors “look here, there are many investors like you who are interested in this”.

6/ PoS coins work differently: to oversimplify, some of the coin holders (who have to own more coins than a certain threshold) vote on what the next block is, with their voting power being proportional to the amount of coins they own.

It doesn’t cost more than running a regular computer, and if everything is well implemented (and the voters all follow the rules) no one risks any funds. It is in a way "greener" than PoW because it consumes no more electricity than a regular computer would.

The important distinction to make here is that mining (PoW) is very expensive and is, over the long run, a zero margin business; whereas staking (PoS) is pretty much the opposite: a perpetual risk-free dividend - a rent, basically.

Note that whenever I talk about PoS in this thread, I sort of include “federated” models in PoS as well, since they share with actual PoS coins the fact that they enable block production for free, with risk-free rewards for block producers.

10/ Because in PoS block production is free, stakers don’t have any expenses to cover, and so they don’t have to sell their rewards; this makes it tempting for stakers to not sell their staking rewards, and compound them instead

11/ thus increasing their share of ownership for free, at the expense of users who just use the coin to transact, particularly if there’ a minimum amount of coin - 32 in Ether, IIRC - that one must have to be able to stake.

12/ Thus a PoS coin doesn’t face the same headwind that a PoW coin does; after a price increase, investors don’t have to face any increased selling pressure from the miners.

13/ Since the cost of a 51% attack on a PoS coin would increase with the price of the coin every price increase will lead, theoretically, to an *immediate* increase in security (when the price increase by X%, the cost of a 51% attack immediately rises by X% a well).

14/ This price/security increase happens at *no cost* for users or stakers, and it’s immediate, every price increase feeding a positive feedback loop: higher price -> higher security -> higher price. This immediate feedback loop makes PoS coins cheap to bootstrap.

15/ Whereas with ith PoW coins, the market has to keep buying back all the miners’ expenses - the coins they dump - just to sustain the price at a given level.
And the market has to keep buying more to sustain an increasing price.

16/ Investors and users have to pay for the coin existing and being secure, and they have to pay for every security increase!
Now you must be thinking… wait, isn’t Proof of Stake awesome? Free security, risk-free gains for stakers, lasting price increases… What’s not to like?

17/ Well, it’ more complicated than that. In the context of competing permisionless, peer to peer systems (as cryptocurrencies are), it’s not just what your system can do that will decide of its success or its failure; it’s also what other “system builders” do and build.

18/ Let me explain: what good is your system if it’s so cheap at launching and bootstraping that anyone can do the same? What good is it if it’s so cheap to run that what it provides is available for free elsewhere?

19/ Because there’s a lot of money to be made (those sweet staking rewards) at no cost, every one will want to jump in and create one’s own Proof of Stake coin or token. I’m not just talking about nerds like you and I, but also companies and countries.

20/ Why would you work to enrich some lucky early insiders who aren’t doing any work nor taking any risk?
A PoS coin’s would be easy to replicate: just fork a coin’s codebase, put some techno-bable on a $15 wordpress website, and voila!

21/ You can try the same thing with a PoW coin, but once launched it has to compete for mining power, and the miners will dump their rewards onto users - every price increase will have to be digested by the market.

22/ It will take time for it price to increase, every price increase making it more tempting for miners to sell rewards that they know they will have to sell at some point - because over the long run mining is a zero margin business.

23/ These price increases have to face not just the ever-increasing miners’ selling pressure, but also the fact that there are already more established PoW coins that have been precisely through this, and have been able to prove that they could attract long-term oriented hodlers.

24/ Why does attracting long-term oriented hodlers matter for a coin’s prospects? Coin holders get to have a say in their chain’s “governance” by selling forks which they are wise enough to see would make their coin inferior as money.
Hodlers aren’t easily forked!

25/ By huddling together, these hodlers have an impact on what their currency is and can be; a cryptocurrency isn’t just a protocol, it’ also a community; a currency which attracts people who understand the mechanics of currency competition has better chances of winning it.

26/ By having to cilmb its way out of an energy hole, PoW coins get to show the committment of their community, who is continually buying out its miners; that bootstrapping and price increase is harder to provoke or replicate than is the case for PoS/federated coins.

27/ With PoS, you don’t even need these pesky, expensive miners. Just give yourself a big premine, spin up a node on a free heroku plan, and start staking! Not only do you get free money, you’re getting free dividends as well! Hey, I really should do the same…

28/ The fact that a PoS coin has a big market cap doesn’t mean much about its merits relative to other PoS coins: the fact that it has been almost free to bootstrap will deprive it of the opportunity to signal that it has value in the eyes of committed, long term holders.

29/ Perhaps that PoS coin has a big market cap because it has the best marketing, or got some mysterious “whales” supporting its price, or had sweet deals with Microsoft or some Russian bank, or is secretly funded by the NSA. Or it cheated! It cheated!
Or all of the above.

30/ Who knows really.
I heard Tron has struck a great deal with Japan’s space exploration agency! That’s it, I’m FOMOing into Tron! (*who* is Tron btw?)
Or perhaps its only stakers are the same 2 guys who founded it and hold 30% of it, and they just don’t want to sell. For now.

31/ So the cheap-to-boostrap nature of PoS coins, combined with the potential risk-free staking gains, make it tempting for those who would want to use a chain or invest in one to instead build their own.

32/ This is not theoretical, this is starting to happen: exhibit A being Binance’s own BNB token, which started on Ethereum and will soon move to Binance’ own chain, which, you guessed it, uses Proof of Stake!

33/ Also, a specific point that remains to be seen: I would expect that as Blockchains become more interoperable through decentralised exchanges and atomic swaps, it will become ever easier to move value between chains…

34/ … which would make it harder for each PoS chain to claim an app-store like network effect (but IMO blockchain interoperability would not undermine the emergence of *monetary* network effects).

35/ In PoW coins: there are no risk-free gains for early insiders nor from miner - mining is an absolutely brutal business (I have a lot of respect for miners), while holding a PoW coin like bitcoin doesn’t provide you with any risk-free dividends.

36/ (well, except for that crazy wave of airdrop in 2017 but that’s something else and not related to PoW; actually, airdrops are much worse in PoS world, precisely for the reasons explained above)

37/ The resulting equilibrium will, IMO, be an ever-growing kaleidoscope of Proof of Stake coins, forks of coins, ICOs, STOs and the like, each controlled by its own clique of premine investors (doubling as database admins to change their chain’s history whenever it suits them).

38/ These assets would have attributes opposite to those of a money: no fair distribution, no liquidity, nothing to distinguish them from the competition; hard to see how any of them could acquire the critical mass and be the kind of neutral ground that an economy needs as money.

39/ These many PoS blockchains would end up being priced based on the revenue they’re able to distribute to stakers rather than on their ability to become money; they would be priced like hosting or cloud providers.

40/ And the price of the kind of security that a PoS blockchain can provide (ie its transaction fees, ie its staking revenue) would end up converging to its production cost: nothing.
(note that in PoS I still include federated and various flavours of "proprietary" chains)

41/ One could imagine a reductio ad absurdum version of this where there are so many equivalent, interoperable PoS chains that it is the chains (or rather whatever industry consortium or research group behind them) that pay for developers to come and build their dapps on them!

42/ As a way of attracting potential clout, “developer mindshare”, future talent/product acquisition, as a PR move, because one bureaucrat found it cool, because that company is led by control freaks, whatever... it’s free anyway.

43/ (or just good old "not invented here" syndrome)

44/ “We at Soy Corp have created the Soy Chain, where developers can build and run their dapps for free (see conditions) as long as they’re integrated with our range of products; developers will also be flown to our headquarters once a year to taste our delicious soy cakes”

45/ PoS has been compared to attempts at achieving perpetual motion. It’s a very apt comparison, because it’s tempting to believe that perpetual motion is possible (“wait, I know it’s impossible, but what if I put a magnet here and a spring there… in a vacuum chamber…”).

46/ Free security and risk-free staking gains feeding into a permanently rising price sound great, and the various Proof of Stake designs I’ve seen make it very tempting to believe it’s possible, but it seems the universe always has a way of preventing free lunches.

47/ In fact, it’s the property of PoS that makes it so appealing (the fact that it doesn’t consume more energy than running a regular computer, thus making it “greener” than PoW) that is also its greatest weakness: it dooms PoS coins to being cheap copycats of each other.

48/ There’s still a lot to cover on this topic but that’s all for today.

The deeper I look into the workings of cryptocurrencies, the more singular Bitcoin appears to me, like a freakish rogue wave slowly rolling through an otherwise flat ocean.

The end!

paging @saifedean @theonevortex @MartyBent @pierre_rochard @bitstein @MrHodl @adam3us @NickSzabo4 @MustStopMurad @BitcoinErrorLog @_jonasschnelli_ @alpacasw @WorldCryptoNet @MadBitcoins @catoshimeowmoto @kyletorpey (a tweetstorm where I compare the economics of PoW and PoS)

also paging @notgrubles @real_vijay @francispouliot_ @hodl_american @giacomozucco @hodl_tarantula @Bitcoinandboyz @WhalePanda @udiWertheimer and let's page some ethereans too! @VitalikButerin @terencechain @IslandHunting @RyanSAdams