Can Proof-of-Stake Lead to Centralization Problems?

Proof-of-stake leads to centralization when it favors token quantity. It’s not necessarily a problem as big holders don’t want to lose money on bad decisions. But it creates fewer opportunities for new validators and higher security risks.

Can Proof-Of-Stake Lead To Centralization?

Decentralization is the essence of blockchain. Yet, one of the most centralized consensus models seems to be the most successful. Proof-of-stake rewards users for locking tokens and selects validators based on those holdings.

You don’t need to buy mining devices in proof-of-stake. You can win many block rewards by just having the most coins for the longest. Hence why proof-of-stake tends to centralize:

• Validators can rig probability by staking over 51% of the total-value locked.
• Big holders earn interest and compound faster than smaller ones
• There are conflicts of interest. Big holdings mean liquidity. Fewer validators mean more rewards and network efficiency.

Maybe the consensus model is the problem. Bitcoin is fairly decentralized and still performs well with millions of users. So is it worth switching from proof-of-work to proof-of-stake?

Proof-Of-Stake VS Proof Of Work: Which One Is More Centralized?

While both consensus mechanisms tend to centralize, proof-of-stake is more centralized than proof-of-work. Mainly, because networks put too much weight on token quantity rather than validator numbers, staking length, or randomness. Proof-of-stake might be profitable for everyone, but governance comes down to a few holders.

Whales aren’t necessarily the problem. Interest rewards increase the gap between stakeholders, even if that gap was just $100. Proof-of-stake needs smart contracts that prevent early holders from taking up the network.

Proof-of-work is based on computing power. “Mining” is a competition between powerful devices to solve cryptographic puzzles to win block rewards. The network security increases proportionally to the energy spent.

Thus, the most competitive miners have dozens of devices and live in countries with low electricity costs. These tend to centralize by creating mining pools. All miners from a pool contribute computing power and get proportional rewards for their contribution. If they win.

Still, proof-of-work is fairly decentralized and more secure than proof-of-stake.

How Proof-Of-Stake Can Avoid Centralization

Does Proof-of-Stake lead to centralization? Not always. You can design for it by rewarding smaller users, reducing validator entry barriers, and expanding token utility.

Build For Decentralization

Developers can design to prevent centralization before it becomes a problem:

• Worried about having too few validators? Program the network to require a minimum of, say, 50 from different IP addresses.
• Too many tokens at stake from one user? Limit how much yield you can generate after a certain range. Or limit the monetary influence on governance.
• Validators teaming and centralizing the network? Increase randomness. Instead of only full nodes, randomize all token holders. Randomly give different weights to different factors on every block (token quantity, transaction history, staking length, validation speed…).
• Are people cashing out with governance tokens? Make them financially worthless to remove incentives. Or require those tokens as a requirement to unlock staked cryptocurrency.
• Too much risk of cyber-attacks? Remove the admin keys. If they’re required for whatever reason, use multi-signature to require several confirmations per action.

Decentralization isn’t hard to design. The challenge is to balance the network security and scalability (see blockchain trilemma).

If you want your network to become successful, it has to be more competitive (scalable) than the leading ones. The best scalability comes from centralization, unfortunately.

Reduce Validator Requirements

Decentralization doesn’t mean much in practice when there’s a high entry barrier. If users need high-tech devices, technical knowledge, or thousands of dollars, fewer will become validators. In order to attract validators in the first place, the consensus model should be rewarding enough for the whole.

To become an Ethereum validator, for example, you need to stake over 32 Ethereum, have enough hardware storage, and keep your device always online. Validators expect minimum hardware requirements to increase after the Merge in September 2022. AKA centralize.

Other examples are:

• BNB Chain. You need to stake over 10,000 BNB, have 250+ GB of storage, and run a full node. Even then, only the 21 largest holders will validate blocks.
• Solana Chain has around 2,000 validators. There’s no minimum amount, but it requires 128+ GB RAM, ~1 SOL per day for voting transaction fees, and $500-$1000 a month for server maintenance.

To get more validators, reduce the hardware requirements and minimum staking amount. One way to split data storage is by sharding the blockchain. It also increases performance, so validator numbers go up without raising fees.

As for staking minimums, lower ones increase the risk of bad actors misusing the network. So it’s recommended to update smart contract penalties as well.

Develop the Ecosystem

As more people use the network, scalability issues arise. When a blockchain doesn’t scale well, users cause network congestion and expensive fees. This drives away new users and incentivizes people to hold rather than use those tokens.

But utility tokens aren’t a store of value. When treated that way, proof-of-stake tends to centralize. So instead of having a stake feature, have several ways for users to get value from tokens.

Ethereum has a large ERC-20 ecosystem. Users can spend Ether on DeFi services, NFT marketplaces, metaverse projects, web infrastructure, governance, or advertising. A diverse ecosystem attracts more users (potential validators) while dissuading big spenders from locking tokens.

Sometimes, you can “borrow” these systems rather than build them. If the blockchain is EVM compatible (Ethereum’s code) or interoperable, it can export functions from Ethereum, Binance, or Solana. For example, Avalanche is highly compatible, scalable, and decentralized. So developers from different blockchains create an AVAX version of their apps. This benefits the network and decentralizes the overall crypto space.

e.g., Polygon also borrows many ERC-20 tokens as an Ethereum sidechain.

Is Proof-Of-Stake The Best Consensus Model?

There are enough reasons to think of proof-of-stake as the best consensus mechanism. While it’s not as safe as proof-of-work, it’s more flexible, accessible, and efficient. Not everyone can stock mining hardware devices, and users wouldn’t be happy to pay more fees for higher electricity costs.

Proof-of-stake is a good foundation, but it’s not enough to solve the so-called blockchain trilemma. Today’s most successful blockchains don’t use proof-of-stake, but a variation of it:

• Polkadot uses Nominated PoS to allow probability and voting to choose validators on every block
• EOS, Cardano, and Tron use Delegated PoS, which is like nominated PoS, except it favors stakeholders the most.
• Algorand’s pure proof of stake allows ALL token holders to become validators by running a weighted “lottery” every few microseconds.
• Solana is one of the fastest blockchains, and Avalanche is one of the most decentralized. They both designed a unique consensus model based on proof-of-stake.

Further proof of this is Ethereum 2.0. and The Merge. While it seems to use the classic proof-of-stake, this mechanism alone isn’t efficient enough. Thus why Ethereum needs dozens of scalability solutions, such as roll-ups, state channels, sidechains, or load-sharing forks like the upcoming PulseChain.