The annualized reward has dropped from 7% to 2.6%, and the dollar income has evaporated by 96%.

Written by: @r2Jamong

Translated by: AididiaoJP, Foresight News

The distribution of Aptos validators is tightening. This is not merely a decrease in the number of nodes; it is a process where network participants are reselecting their operational locations due to performance upgrades, reward designs, and falling coin prices.

As of October 2024, Aptos has 146 validator nodes distributed across 22 countries and 48 cities. Beyond Europe and the Americas, there are nodes in Singapore, Tokyo, Seoul, and Hong Kong, with additional locations in São Paulo, Johannesburg, and Sydney. Overall, it still represents a globally extensive network.

In less than two years, by September 2026, the number of nodes will decrease to only 84, a reduction of 42%; the number of covered countries will drop to 13, and the number of cities to 28, with similar decline rates. The remaining nodes are mostly concentrated in the Americas and Europe. Outside of Europe and the U.S., there is only one remaining node in Tokyo. The previously multi-city layout in Asia has mostly exited.

Two things are happening simultaneously: the performance of the chain is improving, making it harder for validators to operate stably from remote locations; the decline in rewards combined with the drop in coin price has led to financial imbalances for some operators.

The chain is faster, location becomes more important

The Baby Raptr upgrade in June 2025 and AIP-131 (Velociraptr) have compressed Aptos' block time to below 50 milliseconds. This is a clear improvement for users and transactions; for validators, network conditions and the location of data centers are no longer minor variables.

Aptos' rewards are calculated based on "staking amount × reward rate × validator proposal success rate." The further a node is from the validator's main cluster, the easier the proposal success rate drops under the same staking amount, leading to reduced income. Nodes scattered across different continents are therefore the first to be constrained by the rules. Shutting down or migrating to Europe or the Americas becomes a more rational choice.

Performance improvements also raise hardware thresholds. Coupled with AI demand pushing up memory prices, fixed costs for validators increase further. Remote, small-scale, and low-profit operators will struggle to sustain themselves first.

There is an industry contradiction here: a faster consensus mechanism may inversely compress geographical diversity. Latency is a physical constraint, and rewards are tied to proposal success rates, causing nodes to naturally concentrate in low-latency areas. Decentralized distribution may gradually shrink due to performance goals.

Financial constraints determine node retention

The direct reason for node exits is still income. Validators earn through tokens but pay for data center, bandwidth, operation, and labor costs in dollars. The two cost and income curves operate in different market cycles.

Aptos' annualized staking reward dropped from 7% in October 2024 to 2.6% in September 2026, a decrease of 63%. The path is clear: AIP-119 proposed in April 2025 reduced the rate from 7% to 5.19% starting in June that year; the foundation's token economy adjustment in February 2026 further brought it down to 2.6%. Validator income comes from the commission share of delegated staking rewards, and with the decrease in reward rates, both delegators and operators see simultaneous reductions in income.

The real strain on financial balance comes from prices. During the same period, APT dropped from $9.50 to $0.58, a decline of 94%. After the reduction in nodes, the average staking amount per node increased from 5.75 million APT to 8.97 million APT, up by 56%. Staking becomes more concentrated, but the reduction in rewards and falling prices lead to a 96% decrease in annualized rewards when measured in dollars.

Increasing staking by 56% cannot compensate for the income evaporation of 96%. This is the arithmetic foundation of the exit wave.

The shift in token economics towards low issuance and low staking rewards is reasonable from a long-term dilution perspective. However, the cost is that validator business thins out, leading to a narrower network distribution. Tightening on the issuance side and pressures on the operation side make it difficult to balance both.

Other PoS chains face similar issues

Generating income through tokens and paying costs in fiat currency is not unique to Aptos. When the market weakens, marginal regions and small to medium-sized operators often exit first. What remains are more exchanges, institutions, and professional nodes already located in data centers in Europe and the U.S.

The Ethereum community also has similar discussions. EIP-8363 proposes to burn part of the newly issued rewards when staking scales expand, to curb inflation and excessive staking growth. The direction is similar to Aptos reducing rewards, but it also requires an assessment of the economic resilience of validators.

When measuring decentralization, one cannot only look at the number of nodes and staking rates. The more crucial question is: when token prices fall and reward policies adjust, will there still be a sufficiently diverse group of operators remaining? The number of nodes can expand during a bull market but will be reordered by cost during a bear market.

At this stage, participants who can endure short-term reward fluctuations are more important. Exchanges and institutions have other business support, long-term service demands, and a greater capacity to ride through downturn cycles. Therefore, exchange staking and institutional validators have a positive significance for network stability. However, their increased share may also weaken geographical and subjective diversity. Stability and decentralization do not always move in the same direction.

Another key aspect is the entry threshold. Most public chains concentrate engineering resources on enhancing performance and throughput, but they should also invest in efficiency optimization: achieving comparable performance with less hardware, lower electricity costs, and reduced bandwidth. Once the threshold lowers, new operators can enter, and existing operators have space to survive under low rewards. Otherwise, "global distribution" may remain only in documentation, while the actual nodes will shrink to only a few cloud regions.

Long-term decentralization depends on three things

The changes Aptos has undergone in the past two years do not simply prove "decentralization failure," but rather reveal a set of specific constraints:

• Performance upgrades will change who is suitable to become a validator;
• Rewards and coin prices determine who can still bear the startup costs;
• The remaining participants will shape the network map of the next phase.

Long-term decentralization ultimately depends on three points: having validators that can continue to operate during market downturns, effectively lowering operational costs, and keeping the entry environment open to new participants. Lacking any of these, the number of nodes may look good in the short term, but geographical distribution will first narrow.