Forecasting rETH Yields

The ETH staking rate is the most important interest rate in crypto. It is the baseline for ETH returns that all other investments and yield opportunities are judged against. Given this importance, it’s helpful to understand what drives this yield and forecast what ETH staking rates might be in the future.

For the purposes of this blog post we will focus on a specific liquid staking token - rETH from Rocket Pool. We will explore the different factors that drive ETH staking returns and rETH returns in particular, and we will analyze recent trends in these yield drives to forecast what rETH yields will be in the next 6 and 12 month periods.

We find that the yields are likely to decline as the total ETH staked supply increases - our base case is that average rETH yield over the next 6 months will be 3.51% and will be 3.14% over the next 12 months.

How to calculate rETH yields

At a basic level, Rocket Pool takes in ETH deposits, stakes them, and delivers the yield to rETH holders. So to forecast rETH yields, you need to understand what generates the ETH staking yield itself and you also need to understand how much of that yield Rocket Pool delivers to rETH holders.

There are two major components of the ETH staking rate - the base protocol reward and the MEV reward. The base protocol reward comes from issuance paid by the Ethereum blockchain and the MEV reward comes via Flashbots from on-chain activity.

There are two additional factors involved in converting the true ETH staking yield into the rETH yield. These factors are the Rocket Pool fee and the rETH effectiveness percentage. The Rocket Pool fee is the percentage of the ETH staking yield that Rocket Pool pays to validators to incentivize them to do the work. This fee is currently 14.58% of the total ETH staking yield.

The effectiveness percentage is the percentage of ETH held by Rocket Pool that is actively staked and paying yield to rETH holders. The effectiveness percentage is not 100% - at any time, some amount of ETH is held unstaked and not earning yield while it is facilitating withdrawals or is stuck in the validator entry queue. (More details on this later in the post).

To sum it all up, here is the equation for the rETH yield:

rETH yield = (Base protocol reward yield + MEV reward yield) * rETH effectiveness percentage * (1 - Rocket Pool validator fee percentage)

Let’s explore each of these components in turn and forecast how they will impact rETH yields in the future.

Base protocol reward

The base protocol reward is the ETH that the blockchain mints to reward stakers. Today, there are 25,081,536 ETH staked and annual base protocol rewards of 832,971 ETH. This implies a base protocol reward yield of 3.32%.

The annual base protocol reward yield declines as the total amount of ETH staked increases but the decline is nonlinear because annual base protocol rewards increase at the same time. Annual base protocol rewards are a function of the number of active validators. Assuming 32 ETH per validator, annual base protocol rewards are equal to 940.8659 * sqrt(numValidators).

Here are the calculations as of today.

Base protocol rewards: 940.8659 * sqrt(25,081,536 / 32) = 832,971
Base protocol reward yield: 832,971 / 25,081,536 = 3.32%

If the amount of ETH staked were to double to 50,163,072, the base protocol reward yield would decline to 2.35%:

Base protocol rewards: 940.8659 * sqrt(50,163,072 / 32) = 1,177,998
Base protocol reward yield: 1,177,998 / 50,163,072 = 2.35%

MEV reward

The MEV reward is an extra incentive that ETH stakers earn if they choose to support MEV-boost from Flashbots. All ETH staked through Rocketpool supports MEV-boost and earns these extra incentives.

The MEV reward is tightly linked to trading activity on Ethereum. The more markets move and the higher the on-chain trading volume, the more opportunities for MEV and corresponding rewards for ETH stakers. Unlike the base protocol reward, the annual pool of MEV rewards does not increase as the total amount of ETH staked increases. This means that MEV reward yield decreases linearly as total ETH stake increases.

MEV reward yield’s correlation to on-chain activity makes it highly volatile. Weekly average per-block MEV rewards from Flashbots have been as low as 0.07 ETH per block this year to a high point of 0.28 ETH per block.

Per block reward rates spike occasionally, but most of the time rewards sit between 0.1 and 0.2 ETH per block with a rough average of 0.15 ETH per block.

These per block reward rates translate to annualized MEV rewards of between 183,960 ETH and 735,840 ETH with 2023’s average equal to 394,200 ETH. Here is the MEV reward yield contributions today at these different activity levels:

Low activity (183,960 ETH per year): 0.73%
Average activity (394,200 ETH per year): 1.57%
High activity (735,840 ETH per year): 2.93%

If the total ETH staked were to double, this is what the MEV reward yield contributions would be:

Low activity (183,960 ETH per year): 0.37%
Average activity (394,200 ETH per year): 0.79%
High activity (735,840 ETH per year): 1.47%

rETH effectiveness percentage

Like all liquid staking providers, Rocket Pool does not actively stake 100% of the ETH they hold. The percentage of their ETH that they do stake is called the effectiveness percentage. For example, if the effectiveness percentage is 90% and you stake 10 ETH through Rocket Pool, you’re only going to earn the staking yield with 9 of them.

Rocket Pool keeps some percentage of their ETH unstaked for two reasons:

1. To facilitate withdrawals and smooth other technical factors like reward accumulation. ETH can’t be instantly unstaked, so to provide a better redemption experience for users, Rocket Pool will consistently hold a small percentage of its total ETH in its unstaked form.
   
2. Because it’s stuck in the entry queue. Users can’t stake instantly, they have to join the entry queue and wait their turn. While they’re in this queue, their ETH is sitting idle and not earning staking yield.

In “normal” times, you would expect that the rETH effectiveness percentage would be ~95%. But in reality, the demand to stake ETH and the size of the entry queue is so large that the amount of ETH stuck in the queue materially impacts the rETH effectiveness percentage.

For example, here is the rETH effectiveness percentage over the course of 2023:

The effectiveness reached a low of ~70% soon after withdrawals were enabled and currently sits at ~90%. Over time the effectiveness percentage should normalize to ~95% but the time it takes to get there will depend on the strength of the demand to stake more ETH going forward. As long as this demand stays strong relative to the TVL of rETH, we should expect the effectiveness percentage to remain somewhat below its long term equilibrium.

Forecasting future rETH yield

To forecast future rETH yields, you need to make assumptions about the rate of increase in total ETH staked, the future levels of on-chain activity and MEV, and the rETH effectiveness percentage. These assumptions are speculative, but we can make reasonable near-term estimations of these different drivers based on recent trends.

Here is analysis and rough estimations of what these determinants will be over the next 6 and 12 month periods.

• Total ETH staked: total ETH staked has been steadily increasing since withdrawals were enabled earlier this year. Total ETH staked currently grows at a MoM rate of ~5.5%. If we extrapolate this trend forward 12 months, we get an average total ETH staked of 28,793,817 in the 6 month period and 34,247,982 in the 12 month period.
  
• On-chain activity: on-chain activity levels are difficult to forecast. It’s mostly reasonable to think that average levels will be within the previous range of the past year, but it’s a wide range. Given that, I think it’s most helpful to look at three potential scenarios - low activity (0.07 ETH per block), average activity (0.15 ETH per block), high activity (0.28 ETH per block).
  
• rETH effectiveness percentage: change in the rETH effectiveness percentage will mostly be driven by the size of the entry queue and the total amount of ETH staked with Rocketpool. Given that the rETH supply has been growing and the size of the entry queue has been falling, it’s reasonable to expect that the rETH effectiveness percentage will continue to increase. We can assume 92% effectiveness over the next 6 months and 95% effectiveness over the next 12.

Results

These assumptions produce the following results for our estimations of the average rETH yield over the next 6 and 12 month periods broken out by on-chain activity levels:

For example, here's the calculation detail for the rETH yield forecast with average activity over the next 6 and 12 months:

6 Month Forecast

Average Total ETH Staked = 28,793,817
Base Protocol Reward = 940.8659 * sqrt(28,793,817 / 32) = 892,487
Base Protocol Reward Yield = 892,487 / 28,793,817 = 3.10%

MEV Reward = 394,200
MEV Reward Yield = 394,200 / 28,793,817 = 1.37%

rETH Yield = (3.10% + 1.37%) * 0.92 * 0.8542 = 3.51%

12 Month Forecast

Average Total ETH Staked = 34,247,982
Base Protocol Reward = 940.8659 * sqrt(34,247,982 / 32) = 973,352
Base Protocol Reward Yield = 973,352 / 34,247,982 = 2.84%

MEV Reward = 394,200
MEV Reward Yield = 394,200 / 34,247,982 = 1.15%

rETH Yield = (2.84% + 1.15%) * 0.92 * 0.8542 = 3.14%

A lot can happen in the next 12 months, but these numbers provide a solid baseline for what to expect from rETH yields going forward based on what we know today.

I think the most significant unknown is the level of MEV rewards going forward. If the bull market comes back, we could see total MEV rewards dwarf what we have seen in the past year. On the other hand, if further advances in DEXes like Uniswap X succeed in mitigating MEV opportunities we could actually see the MEV reward number decline even if on-chain activity increased!

But other than MEV rewards, I think we can be confident about these forecasts. The largest yield driver in the near-medium term will be the growth in the total ETH staked and I think it’s a safe bet that number will continue to go up. You can never know for sure, but hopefully this analysis provides a good baseline from which to base ETH-denominated investment decisions over the next 12 months.