The Bank for International Settlements published research examining whether high-frequency trading (HFT) affects firms’ cost of equity, using exchange speed upgrades as quasi-natural experiments. The paper finds that faster trading can, on average, increase firms’ cost of capital, with materially different effects across stocks depending on their beta and liquidity. Causal estimates are identified using NASDAQ’s introduction of co-location in Q2 2005 and a latency-reducing technology upgrade in Q2 2010, applying a difference-in-differences framework with propensity score matched NYSE stocks as controls. The results indicate higher post-upgrade costs of equity on NASDAQ, with the paper attributing the net effect primarily to a “systematic risk” channel: HFT increases the cost of capital for low-beta stocks by raising their sensitivity to market-wide risk, especially among more illiquid names. In contrast, for the most liquid stocks, HFT is associated with a lower cost of capital through a reduced liquidity premium. To test whether the findings depend on US-style market fragmentation, the analysis is replicated in the largely unfragmented Hong Kong market using the Stock Exchange of Hong Kong’s Orion Central Gateway introduction in Q2 2014, with matched Shanghai Stock Exchange firms as controls. The paper reports similar patterns across settings, and notes regulatory relevance in that HFT’s costs and benefits appear concentrated in different parts of the equity universe rather than uniformly distributed.