Ethereum's killer app could "kill it"? Researchers think DeFi will eat into PoS cybersecurity

Written in front: This article was written by Dragonfly Capital investor Haseeb Qureshi. He believes that, whether it be staking or borrowing, speculators in the network will always choose the most beneficial method of participation. As long as the attacker raises the yield of the on-chain lending market, there will be fewer and fewer people participating in staking, which will eventually affect the security of the PoS network.

defi

The following is the full text of the full text:

At present, on-chain lending has become the most popular decentralized financial (DeFi) application, and the borrowing of MakerDAO, Compound and dYdX platforms has exceeded $ 600 million. On-chain lending has the potential to overturn traditional secured lending. But it seems to do more than that: it can also disrupt PoS consensus.

PoS is another alternative to PoW. In the PoS system, the blockchain is protected by pledged crypto assets, not computing power. Many networks launched last year are PoS networks (such as Tezos, Algorand, Cosmos, etc.), and more will be launched next year.

The PoS system is secure when there are a large number of tokens being actively pledged. In most PoS algorithms, as long as two-thirds of the pledged assets are held by honest participants, the blockchain is secure.

Now suppose you are an attacker trying to damage the PoS system. What would you do?

At a higher level, there are two ways to attack: you can collect one-third of the pledged assets, but this is very difficult and the cost is too high. The second method is that you can convince the current staker to stop staking and then take over this cheaper network.

The second method sounds attractive, but how can we stop the current staker from staking? Here's an easy way: Give them a more attractive rate of return.

PoS will only work if economic incentives can be obtained after staker pledge, and staker will only pledge tokens if the reward is large enough. But if they can get better returns elsewhere, then a rational staker will stop staking and put assets wherever they can get higher returns. If this leads to fewer and fewer people participating in staking, then the network becomes insecure.

To put it plainly , the on- chain lending market competes directly with staking-which means they compete directly with the security of the protocol!

You may have an intuition that we need to understand one of the important interactions. But how to accurately analyze?

Simulation staking game

The best way to simulate a complex economic system like Ethereum is through a technique called agent-based simulation. In agent-based simulation, you need to model a large number of agents with different strategies and risk configurations, and then make them independent of each other. By observing how the emergency system evolves (and repeating experiments thousands of times with different parameters), you can get relevant data to understand how the network behaves in different scenarios.

Gauntlet's Tarun Chitra did exactly that in his latest paper, The Competitive Balance Between Staking and On-Chain Lending. He analyzed the interaction between on-chain lending and PoS pledges, provided that the staker is economically rational of. (Economic rationality means: each agent has a portfolio of assets. These assets are either lent, pledged, held, or traded. The risk profile of each agent is slightly different. They rebalance their investments. Assets in a portfolio to maximize risk-adjusted returns.)

1

(ETH pledged vs. ETH borrowed)

The above figure is a single simulation that simulates how the ETH (orange line) of Compound and the ETH (blue line) in the pledge change over time in a situation similar to Bitcoin deflation block rewards.

The information given above is: Initially, most ETH holders pledged their ETH. But over time, block rewards have fallen, and the returns from holding ETH are no longer attractive compared to borrowing on Compound, so almost everyone has adjusted their ETH to Compound. (You can ignore the initial conversion between borrowing and staking, which is due to random initialization.)

Tarun made several theoretical closed predictions and verified them through simulations. But the most important point is that PoS chains cannot safely use deflationary monetary policies. If the PoS block reward decreases over time, then its long-term equilibrium will be that almost all assets are lent out, rather than pledged.

Let's go further. Knowing this, what can an attacker do?

If an attacker buys the on-chain lending market and gives a better long-term interest rate, this will make the staker switch from staking to borrowing. Then, once the staking funds on the chain are getting scarce, they can enter and dominate the staking market.

Of course, at Compound, the way to lower borrowing rates is to borrow from an asset pool. The risk model then automatically raises interest rates. As attackers continue to borrow and interest rates continue to rise, more and more stakers will turn to lending platforms, and the security of PoS is gradually exhausted. This could lead to a snowball effect: bystanders seeing staking shrink, they would want to short ETH, further increasing the demand for Compound's borrowing. You can think of this network as a sweater with an attacker pulling a line: interest rates. When the attacker pulls the sweater, the sweater responds to pressure, the thread becomes longer and longer, and soon the attacker will unravel the entire sweater.

This is a surprising result! DeFi and consensus seem completely orthogonal, but the competitive lending market actually has a significant impact on the security of PoS machines.

What does this mean for PoS?

First, let's be surprised: oh my god, Turing's complete blockchain is too complicated! Adding smart contracts to the blockchain seems like it should be a pure application layer decision. However, smart contracts make it possible for a complex market like Compound, where Compound interacts with the underlying security of the chain in an inconspicuous way. We often talk about "layer 1" or "layer 2", but unlike the traditional computing model of OSI (Open Systems Interconnection), the blockchain is full of abstract holes.

This also reminds us: we cannot continue to pretend that the blockchain is a closed system, and its only motivation is an internal agreement. The blockchain is too complex and interconnected to analyze in an isolated environment. In this regard, little is known about the actual security of PoS.

As long as the PoS network is in an open ecosystem, any on-chain lending market can erode its security by providing higher returns. In fact, even if the system does not directly support smart contracts (such as Cosmos ATOMs), if the pledged assets can be tokenized and transferred across chains, the tokenized lending market on another chain may have the same effect!

Worrying about these aren't that stupid?

We discussed the possibility of active attacks, and maybe the cost of capital is too high for you. But this can happen even if no one has done evil! It may also be that venture-assisted projects subsidize their own interest rates, try to compete with each other, and inadvertently reduce network security. The end result will be the same: a dangerous and insecure consensus layer.

How should the PoS system prevent this from happening?

The staking network has two options to combat this situation: either force the on-chain lending market to limit interest rates, or compete with the lending market to provide higher returns to the staker.

The first strategy is similar to capital controls. This is obviously impossible on an unlicensed blockchain-even so, borrowers and lenders can simply build the same market off-chain or through adjacent interoperable chains.

The only realistic precaution is to use flexible monetary policy when necessary to provide competitive interest rates. Any fixed inflation system is vulnerable to this attack, because attackers must want to know exactly how much subsidy they need to give the loan market in order to eat away speculators.

This defense is similar to a central bank adjusting interest rates to achieve its economic goals. PoS networks must use its distribution rate as a tool to adapt to real-time market pressures.

In this sense, the current state of Ethereum is actually very good, because it does not promise any fixed monetary policy. However, all PoS networks must take this trade-off into account. There are two possible methods of on-chain governance and off-chain governance, but if the PoS protocol wants to remain secure permanently, it must have an adaptive monetary policy.