Recently, I reviewed some data related to the Ethereum Shanghai upgrade and wanted to share some interesting observations. At that time, everyone's biggest concern before the upgrade was whether the wave of unstaking would crash the market; now it seems that this issue isn't so straightforward.

Let's start with the numbers. Before the upgrade, there were approximately 562k validators on the beacon chain staking over 18 million ETH, with an average of 34.01 ETH per validator. The distribution of staked ETH was quite uneven, with liquid staking protocols accounting for 33.3% of the market share. The largest player, Lido, controlled 93.78% of the liquid staking market, staking over 5.67 million ETH. Other major exchanges also participated, handling over 4 million ETH.

Interestingly, only 28.7% of validators were in profit at that time, while 71.3% were in loss. This means that not many validators had the motivation to immediately withdraw and sell, because most validators were actually losing money.

Regarding the impact of the Ethereum Shanghai upgrade timing, several scenarios were predicted. In the worst case, there could be a daily sell pressure of 284k ETH, accounting for 5.27% of the 24-hour trading volume at that time. However, considering factors like liquidity, profit distribution, and new staking, the actual sell pressure might only be 59k ETH per day, which is about 1.1% of trading volume. That’s a significant difference.

What changes did the upgrade itself bring? The most critical is EIP-4895, which allows validators to finally withdraw their staked ETH and accumulated rewards. Previously, the beacon chain was one-way—once staked, you couldn’t withdraw. After the upgrade, there are two withdrawal methods: one is partial withdrawal, which only takes out rewards exceeding 32 ETH, allowing validators to continue participating in validation; the other is full withdrawal, which completely exits staking.

There are some technical limitations worth noting. Fully exiting can process up to 1,800 validators per day, equivalent to 57,600 ETH. For profit withdrawals, since there’s no queue limit, theoretically, up to 115.2k validators’ worth of rewards can be withdrawn daily. This design is actually a balance to mitigate market impact.

Looking at the staking distribution, 53.7% of staked ETH has already gained liquidity through liquid staking derivatives like stETH and bETH. These users could have traded on the market long ago, so the selling pressure after unstaking shouldn’t be too high. The real concern is the 46.3% of non-liquid staked ETH.

The conclusion at the time was that there might be significant volatility in the short term (the first five days after the upgrade), but long-term concerns are unnecessary. After all, the motivation to stake more has been increasing, especially with the unstaking feature now available, encouraging more participation. Looking back, this judgment was pretty accurate. Market manipulation and panic emotions were the real risks on the day of the upgrade.