Bitcoin’s Estimated Hashrate Fell 25% After the Halving

Bitcoin’s seven-day rolling estimated hashrate reached approximately 90 exahashes per second on May 23, 2020, down from 120 EH/s on May 11. That was a 25% calculated decline across the twelve-day window and the clearest operational evidence that miners were adjusting to the network’s third subsidy halving.

The retreat did not mean Bitcoin had stopped processing transactions or that a quarter of identifiable mining machines had been disconnected. Hashrate is inferred from mining difficulty and the random pace at which blocks are found; it is not measured by a network-wide meter. A seven-day average reduces short-term luck but remains an estimate.

Market conditions were comparatively subdued. Kraken’s May 23 exchange report quoted bitcoin at $9,193, up 0.23%, with $79.6 million of bitcoin volume and $114 million traded across all Kraken markets. Those were Kraken-specific figures covering its crypto and fiat markets—not a consolidated global close or audited industry-wide turnover.

The subsidy shock became an operating test

Bitcoin block 630,000 was timestamped May 11, 2020 at 19:23:43 UTC. Under the protocol’s issuance schedule, that block reduced the subsidy from 12.5 BTC to 6.25 BTC. Transaction fees remained additional compensation, but the fixed portion of each new block reward was cut in half immediately.

Glassnode reported on May 22 that estimated hashing power had fallen by more than 20% over two weeks. It interpreted the decline as a response to reduced miner revenue and noted that fees were supplying more than 15% of miner revenue. Those were analytical estimates, not disclosures from a complete census of mining operators.

The economic mechanism was straightforward: a miner receiving half as much newly issued bitcoin per block would need some combination of lower costs, higher fees, greater equipment efficiency or a higher bitcoin price to preserve the same revenue profile. The public record does not identify which operators shut down equipment, how many machines moved between networks or each miner’s electricity price.

Congestion exposed the other side of the adjustment

The reduction in effective mining power temporarily slowed block production. When blocks arrive less frequently while transaction demand remains high, pending transactions accumulate and users compete through higher fees for limited block space.

A Decrypt report published May 23 said BitInfoCharts had recorded an average transaction fee of $6.64 on May 20, compared with $0.28 on January 1. The increase was $6.36, or approximately 2,271% when calculated from the rounded figures. Decrypt reported 2,213%, demonstrating why rounded endpoints should not be used to reproduce a provider’s more precise percentage.

Bitcoin had already lowered mining difficulty by approximately 6% on May 20. That automatic retarget provided some relief, but the May 23 hashrate estimate showed that the post-halving adjustment remained unfinished.

What May 23 established

The defensible conclusion is narrower than a claim that Bitcoin’s security fell exactly 25%. The seven-day hashrate estimate indicated materially less effective computing power than on May 11, while elevated fees showed that available block space had become more expensive during the transition.

Kraken’s modest bitcoin price change also cautions against treating network statistics as mechanical price signals. The exchange record documents a quiet session on one venue; it cannot establish that traders had fully priced the mining adjustment or that hashrate caused subsequent market moves.

Later context

On May 27, CoinDesk reported that the seven-day hashrate average was still near 90 EH/s while mean block intervals had shortened to approximately 11 minutes. It also documented declining congestion and fees. That later observation clarifies the adjustment path but does not change what the May 23 record established.