The significant increase in the Unspent Transaction Output (UTXO) set, from approximately 90 million to over 140 million UTXOs within seven months, has resulted in a noticeable expansion of the chainstate database size managed by Bitcoin Core, rising from around 5GB to roughly 9GB.

Observations indicate that during periods of low transaction volume, such as early weekend mornings, there is a pattern of bursts of small, high-fee transactions related to ordinals/BRC-20 tokens. These appear to be strategically timed and priced to ensure blocks remain full and the mempool doesn't clear, possibly under the pretext of token minting. Data from the mempool.space website provides evidence for this, with examples including large clusters of high-fee transactions within specific blocks.

This pattern suggests the possibility of deliberate manipulation of transaction fees, potentially by miners or a group of miners. If one were to profit from high transaction fees, they would either need substantial funds or benefit directly from the high fees. This leads to speculation about whether such actions could be an attack on the Bitcoin network's fee structure, which is designed to deter DoS attacks. If an entity could profit from artificially inflating transaction fees, this would undermine the network's economic defenses against abuse.

A theoretical analysis explores the profitability of such an attack, concluding that it would depend on several factors: the percentage of miners participating, the volume of fee-stuffing transactions needed to maintain high fees, and the costs associated with these transactions. A key point is that the higher the participation by miners, the lower the cost of the attack, as they would recoup more of the fees spent on the attack. In an extreme scenario where all miners are involved, the attack would have no cost since all fees would be returned to the attackers.

The real-world impact of such an attack involves fee estimation algorithms, which many wallets use. These algorithms often base their estimates on historical data, leading to overpayment of transaction fees following artificially induced high-fee periods. An example provided is the aftermath of a flood of transactions with fees up to 700 sat/vB, which caused Bitcoin Core's fee estimation to advise significantly higher fees than necessary for subsequent transactions.

In summary, while there is no concrete evidence of malicious activity, the patterns observed in transaction timing and fees warrant consideration of the theoretical profitability of manipulating Bitcoin's fee market. It is important to note that this analysis is speculative and based on observable patterns rather than definitive proof of intentional fee inflation.