Parallel execution claims face validator scrutiny
Aptos faces centralization risks due to a hand-picked validator structure and supply concentration, where over 80% of APT remains with the team and investors. Technical throughput promises of 160,000 TPS clash with launch performance of only 4 transactions per second.
I find that the network’s technical throughput promises clash directly with its centralized validator structure and supply concentration. Aptos processes 16 million transactions in a single day, according to reports from early July. This follows a June 2026 performance where the network hit 83.7 million transactions in one week. However, earlier reports noted that the parallel execution engine only managed 4 transactions per second at launch, a figure that falls far short of expectations. I find the discrepancy between the promised 160,000 TPS and the actual 4 TPS performance at launch impossible to ignore. The total supply hits a 2.1 billion APT cap, yet 821,111,362.91 APT currently sits in staking. This means a bit over 80% of the supply remains under the control of the team and investors because no airdrop ever occurred. The April 2026 governance changes established a hard supply cap of 2.1 billion APT and mandated that 100% of transaction fees burn, which creates a mechanical link between network demand and token supply reduction.
Economic stability and liquidity
Monthly emissions of 1.6 million APT stay higher than the current burn rate of 235,200 APT, which means the net supply still grows. Since the mainnet launched in October 2022, cumulative burns reached 1.4 million APT. Even though average transaction costs hold at $0.0005, the network struggles to maintain significant capital retention for its user base. You probably know that high TPS means nothing if liquidity disappears. Aptos saw its total value locked drop from a 2026 peak of $1 billion to approximately $275 million. I think the network fails to retain capital despite its technical speed. The stablecoin market cap hit $1.65 billion after a $1.9 billion peak in the first half of 2026. The network also hosts more than $540 million in tokenized real-world assets, including BlackRock’s BUIDL fund and Franklin Templeton’s BENJI. Staking rewards currently sit at approximately 2.6% APR.
| Metric | Value |
|---|---|
| Peak Testnet TPS | 30,000 |
| Aggregate Shard TPS | 1,000,000 |
| Current TVL | $275 million |
| Monthly Emissions | 1.6 million APT |
| Monthly Burn Rate | 235,200 APT |
The network recorded 12.3 million daily transactions on May 2, 2026, but engagement remains inconsistent. APT currently trades around $1.11 with a market cap near $890 million. This follows an initial supply of 1 billion tokens where 510 million went to community members and 190 million went to core developers. The remaining 300 million tokens belong to the Aptos Foundation and private investors. The project holds a valuation between $1 billion and $4 billion.
Validator risks and security
Running validators requires a minimum of four CPU cores, 8GB of RAM, and 300GB of storage. The network maintains roughly 140 active validators across 23 countries and 54 cities. Aptos uses the Raptr consensus mechanism and the VelociRaptr pipeline to reduce block proposal time to the tens of milliseconds, providing a faster way to achieve finality. This architecture relies on a quorum store that replaces traditional leader-centric mempools. Only batches that reach a quorum are eligible for ordering.
A developer named Paul Fidika stated the foundation hand-picked all 101 validators. He also noted that validators signed an arbitration agreement in the Cayman Islands to join the set, which they had to do to be included. This hand-picked structure creates a centralization risk that contradicts the goal of a decentralized protocol. The network also maintains 32 threads working in parallel through the Block-STM design to achieve 100,000 TPS. Each shard produces between 35,000 and 40,000 TPS. The proposed encrypted mempool uses a batched threshold encryption scheme to reduce latency by less than 20 ms per batch. This update aims to hide transaction intent from validators until after block ordering. Will the community approve this upgrade given the existing centralization concerns regarding validator selection?
Join the discussion