Pricing Network Trust: Attack Cost, Hash-Rate, and the Illusion of Stability in Bitcoin Markets

This study re-examines the relationship between the cryptographic security of a blockchain network and the volatility of the cryptocurrency market that operates on top of it. Much of the prior literature has proxied network security with a simple normalization of hash-rate, which is by construction a linear transformation of hash-rate itself and therefore carries no independent economic information. This paper instead constructs an attack cost index that combines hash-rate with mining hardware efficiency and electricity price, and tests its relationship with GARCH(1,1) conditional volatility using a full sequence of unit root tests, cointegration analysis, trend-controlled regressions, and first-differenced (growth-rate) regressions on daily data from January 2019 to March 2026 (n = 2,630). A levels regression reproduces the negative relationship reported in prior work, but the coefficient loses all statistical significance once a linear time trend is added, and the relationship disappears almost entirely in first differences. The price–hash-rate correlation that prior work treated as its most robust finding is reproduced here in levels but likewise collapses in growth rates and at every lag from 7 to 90 days. The paper argues that these results are consistent with spurious regression between non-stationary trending series, and shows that introducing a more economically grounded security index does not, by itself, resolve this problem unless the non-stationarity of the underlying series is explicitly addressed.