Analyze the consensus mechanism from an economic perspective: PoW is a diesel car and PoS is an electric car

Translator's Note: The original author Eric Wall likens PoS to electric vehicles. He believes that PoS does have some advantages from an economic point of view, but he still believes that Proof-of- Work ( PoW ) is the best way for us to obtain a safety chain A robust approach is also the most reliable way to design cryptocurrencies.

Three months ago, I posted this picture on Twitter with the following text:

When someone says "PoS is not so wasteful."

p1

Diesel generator powers electric cars

This tweet was a gimmick, but it triggered one of the most insightful conversations I've had on Twitter. The framework for thinking about this issue has changed dramatically since the discussion.

The purpose of this article is to highlight some of these insights, as well as my current thoughts. Keep in mind that this article focuses on the economics of Proof-of-Work (PoW) vs. Proof-of-Stake (PoS). For the sake of brevity, I have not discussed other factors.

cause

5

The purpose of this tweet is simply to convey Paul Sztorc's claim that " Nothing is cheaper than proof of work (PoW) ", which is perhaps the most influential piece of public discussion about PoW and PoS economics article.

Paul argues in his article that in a blockchain system, the work we do to get block rewards (and the cost we will bear) will be equal to the compensation paid, regardless of whether the work involves calculations Hash or staking funds. Simply put, if the blockchain pays an average of $ 1,000 per block, PoW miners will spend up to $ 999 to get paid (but not more than this number). Similarly, PoS pledgers lock in the price of their capital That's up to $ 999.

But Paul also believes that not only individual costs are comparable, but social costs are also comparable! In a world of PoW currency system, we burn a lot of energy to ensure the security of this chain, and in a world of PoS chain, we will lock a lot of capital for a long time to ensure the security of this chain . Paul's argument is that preventing capital from entering the broader economically constrained market will inhibit technological progress and economic growth.

This is why this argument is so important, it gives the economic reason that Bitcoin should remain the status quo, and this reason is not only because of the security guarantee of the system!

During my years in the cryptocurrency industry, Paul's perspective broadly outlines my views on the broader economic debate on this topic. In 2016, I quoted Paul in his master's thesis. After discussing it with many other Bitcoin enthusiasts, I think it currently constitutes a mainstream opinion.

Now, let's take a look at the counter-arguments that emerged in this recent Twitter buzz and my own thoughts.

David Schwartz: PoS will have a bigger impact on your money

p2

David Schwartz's contribution is as follows: Yes, pledgers and miners can spend up to $ 999 to get a $ 1,000 block reward. But the security provided in the form of economic guarantees is quite different. Let's consider the penalty for trying a double spend attack in each system:

In the PoS system, paying $ 1,000 every 10 minutes means that the reward paid every year is $ 50 million. If investors find that a 5% return on capital is worth the investment, they will pledge about $ 1 billion in capital. To successfully launch a double-spend attack in PoS, you need to control more than two-thirds of these funds (about $ 670 million) to generate a conflict chain, and there will be penalties in the event of an attack.

The penalty for double-spend attacks on PoW is more difficult to explain, at least in some cases, the penalty is much smaller. For example, suppose a miner has 51% of the computing power, and he has recovered some or all of the initial hardware investment through mining. In this case, the cost of a reorganization attack (such as the last 7 blocks) does not require $ 670 million at all, and the cost of generating these blocks is only about $ 7,000 (7 * 1000). If the reorganization is successful, the attacker will also receive block rewards (regardless of how much these blocks are worth after such an attack).

This is not to say that PoW double spending is feasible, this is just a hypothetical example. Although the analysis ignores many of the practicalities of the attack, it may be misleading, but in a narrow economic sense, it may have the core of truth, so it is worth further exploration.

Think about it from another perspective: It is obviously better to pledge $ 5 and then take it back later than to burn $ 5. In order to feel the cost, you have to pledge $ 100 within a year to match the $ 5 burn. Even though the personal costs are the same ($ 5 to $ 5), the penalty is not ($ 100 to $ 5)! Now, PoW coins that use dedicated hardware usually face the risk of hardware investment. Once attacked, these investments may lose value, but it will never match the risk value of the pure staking protocol because PoS pledgers will The reward does not need to bear the cost of burning a lot of electricity.

We can say that the penalty of PoS and the penalty of PoW are asymmetric. When the block reward is equal, PoS will have a larger amount of money at risk.

This has some meaningful results. If PoS achieves higher economic security at the same cost, this means that PoS's security budget can be smaller, but it can generate the same economic guarantee. This means that we can reduce the block reward, but the economic security of this PoS chain can be the same as that of the PoW chain.

The retort Paul: you can't control the size of the block reward

p3

The PoS system can reduce the payment of block rewards caused by this asymmetry. This view does not really work under the mindset of Bitcoin lovers. Bitcoiner's impression is that at some point, all block rewards from the new supply will disappear. At this point, the only factor that ensures the security of these two chains is transaction fees, and transaction fees are mainly set by the block space market, not consensus algorithms. It is further assumed that these fees are sufficient to ensure the security of the blockchain, so you cannot benefit from asymmetry. If we assume that the PoW chain will be secure without it, it just means that the PoS chain is "overprotected" by the mortgagor. of.

Eric's objection

p4

Paul makes sense, but the arguments don't seem to be very strict. A somewhat vague assumption is that transaction fees will always be sufficient to ensure the security of the PoW chain. In blockchain systems, transaction costs can sometimes be very different. Whether the fees alone can reliably generate the security we need is a major source of controversy in bitcoiner. We can certainly see that it is difficult for transaction costs to generate the necessary security budget. In this sense, if the PoS protocol can provide higher security for the same return, then it becomes very attractive.

In the worst case, if there is no supply inflation, this may cause the PoW protocol to become unstable, while the PoS protocol can rely on transaction costs to maintain its own operations. Annoyingly, this means that PoS systems may even become a higher stock-to-flow (S2F) asset! Of course, these are just speculations, but it emphasizes that this asymmetry does not bring disgraceful consequences, and this asymmetry should not be covered up.

Dan Robinson's point of view: money is not a resource

p5

Dan's argument is that the wealth of our world cannot be measured by how much "money" we have. It is easy to understand that the world cannot become richer because of continuous printing of money. Such an action will only redistribute those who have the ability to buy goods, but the entire society will not suddenly have more goods and services.

However, in a world of Proof-of-Stake (PoS) currencies, assets that are withdrawn from circulation through staking are simply money. So, is the quantity of goods or services in the world reduced just because someone is pledged their assets?

No! This simply means that someone has given up the ability to buy goods and services and moved to pledge, which means that there are more things for others to do. In the real world, staking does not destroy any assets, nor does it make the world poorer.

Dan further argues that there is no such thing as "making capital out of the wider economy." Just as government printing money will make others poorer, locking one's money will make others richer (during this period). As everyone becomes a little richer, the net capital available to the economy remains the same.

Retort Paul: Money is actually a resource

p6

Paul's argument is that money is not like a glass of water, and if you put a glass of water on a flat surface, it is balanced no matter how you put it. In fact, there are many places where capital is relatively concentrated, and money flows to useful places. This is really important. Concentrating capital injections on a company that has the potential at the right time will have a much greater impact on society than a group of people on the street carrying a 5-dollar bill.

Dan's point that the unit of money itself has no value is valid, but it generates huge value in distribution. Ideally, the market distributes money to those who create value and relies heavily on "money in your pocket" to be effectively used. For a long time, Staking has effectively eliminated the opportunity to enter these pockets, which means that some companies that could have obtained funds will not, and this is the social cost of Staking.

Eric's View (# 1)

I understand Paul's idea by thinking about how entropy works. In thermodynamics, we say:

Low entropy = ordered = high energy

High entropy = disorder = low energy

p7

I think similar logic can be applied to economics:

Low entropy = ordered = high economic potential

High entropy = disorder = low economic potential

"Money in your pocket" is essentially an "island of order" in a disorderly system. The intelligent distribution of currency increases the overall economic potential of the system.

Even when an entity locks money, everyone may become richer, but the wealth is evenly distributed on a gray spot. Net economic potential remains affected. So Dan's argument is basically "We won't take any particles out of the jar!", And Paul's argument is "Yes, but we are unleashing economic potential by creating more chaos!" The redistribution of wealth is carried by Energy is different from the energy carried by the original, concentrated capital.

Eric's View (# 2)

So the views above are not opposing, I just agree with Paul. Now, let's attack Paul's view with a truly opposing view.

Although I think Paul is right in principle, the whole situation changes when we think further about who will staking.

1.Green energy for mining

Let us first consider the fact that staking is as competitive as mining. In mining, as mining operations increase the difficulty of mining, a miner with cheaper hardware and energy will have an advantage over other miners.

So, in general, mining will tend to use the cheapest energy on the planet. Interestingly, no energy is cheaper than "trapped" energy because they cannot be used for any other purpose.

These trapped energy sources include natural gas, solar energy, wind energy, and excess unused electricity. These types of resources are used by mining. Whenever the topic of PoW's impact on the environment gets public attention, bitcoiners often boast of this. However, these only solve the energy problem, not the externalities of manufacturing mining hardware.

2.Staking green capital

The same principle applies to staking. For staking, what is the cheapest “funding source”, that is, funds that you can never compete with. Let's take Ethereum as an example, those who are bullish on ETH. If a person is bullish on ETH, and he may have decided to bullish ETH in the next 5-10 years, then he has almost no economic cost of staking. Even better, if he only holds or invests, it has no impact on the overall economy. In addition, these investors are willing to accept lower and lower yields that will outperform other investors! This "trapped capital" will exceed any "non-trapped capital" that could otherwise be used for production elsewhere, as in the mining example above.

p8

When the size of the Staking pool increases due to increased competition, the return on Staking will decrease ( source )

When someone says "PoS is not so wasteful"

Back to the picture at the beginning of the article, the reason why it is humorous is that it captures the argument that PoS is just a vague representation of PoW.

p1

However, there is a turning point. Facts have shown that using a diesel generator to charge an electric vehicle does not mean directly using diesel to power the car. A Tesla charged with a diesel generator will still drive farther than a diesel-powered vehicle consuming the same amount of diesel! Electric engines just use energy more efficiently, and they do reduce waste.

By the same token, the economic significance of proof of equity seems to be greater than what is covered by mainstream bitcoiner perspectives. The "green capital" part of the above argument, I personally think it is very attractive, and it is also the main reason why I now consider exploring the "PoS potential advantages" from the perspective of social impact.

What does this mean for PoW?

From a narrow and economic perspective, PoS seems to have some advantages that I have not fully considered before, but I still believe that Proof-of-Work (PoW) is the most robust way for us to obtain a secure chain, and it is also the most robust way to design cryptocurrencies. Reliable method.

In this article, we completely ignore some of the technical security issues in PoS systems, and many of them are related to the security dependency of proof of stake. I am still confident that Proof-of-Work (PoW) is the most secure system we have comprehensively considered at present, but my thinking may change in the future, and I will monitor the progress of the PoS system keenly.

Many thanks to Hasu, James Prestwich, Dan Robinson, Brian Crain, and Torbjørn Bull Jensen for contributing their thoughts to this article, and to Paul Sztorc, Alex Mizrahi, Emin Gün Sirer, Zaki Manian, David Schwartz, Georgios Konstantopoulos, and others for their thoughtful participation The entire discussion.