Why the Volatility Sign Flips
The participation results show that whale alerts reach the Bitcoin and Ethereum-network environments differently. This chapter asks what happens after that response reaches prices: does realized volatility rise, fall, persist or fade, and does the alert affect the asset named in the event or another market as well? The answer helps separate price uncertainty from trading activity and keeps the reported BTC–ETH contrast tied to a horizon and a matched baseline rather than to a broad claim that one network is simply “more volatile.”
Volatility is a separate outcome
Realized volatility measures how much minute-by-minute returns vary over a selected period. It is calculated from the returns themselves, rather than from whether the price finishes higher or lower. If Bitcoin rises steadily, its realized volatility can be relatively low. If it moves sharply up and down and ends near its starting price, realized volatility can be high. A positive post-alert estimate therefore means more return variation than in the comparison window; it does not tell us the direction of the price move.
That distinction matters because a whale alert can affect several dimensions at once. The Bitcoin results in the previous chapter show a broader group of active non-whale wallets and same-direction buying or selling. Those observations describe participation and trading direction. The volatility model asks a different question: how dispersed were minute returns after the alert relative to matched periods before it? More traders can participate in an orderly market, while a smaller number of trades can move prices sharply in a thin market. The measures should not be substituted for one another.
The study compares post-alert windows with matched pre-alert windows and controls for hour of day and day of week. It examines horizons from 15 minutes through 30 minutes, one hour, six hours, nine hours, twelve hours and 24 hours. This time structure is essential. A 15-minute shock that is gone by the next day is a different economic event from a persistent change in volatility, even if both are described casually as “whale-driven turbulence.”
Bitcoin volatility rises, then reverses
The native-Bitcoin results show a clear time profile. A BTC alert is associated with a 0.11 increase in BTC realized volatility at 15 minutes, 0.12 at 30 minutes and 0.15 at one hour. The association reaches about 0.22 at six hours and remains about 0.22 at nine hours. At twelve hours it falls to 0.16, and by 24 hours it is -0.12. The estimated increase therefore builds for several hours in the study's windows before reversing relative to the matched comparison.
The word “builds” should be read as a description of the sequence of estimates, not as a claim that one alert continuously pushes the market in one direction. The model estimates volatility over separate horizons, and the six-hour window contains a different collection of minute returns from the 15-minute window. Still, the pattern is consistent with a burst that is not confined to the instant of broadcast. Attention, order flow and repricing may continue to work through the market for hours before the excess variation disappears.
The 24-hour reversal is equally important. It keeps the result from becoming a story about permanent instability. Relative to the matched baseline, the study finds no continuing positive BTC volatility effect at that horizon; for BTC alerts and ETH alerts, the reported coefficients are negative at 24 hours, while the WBTC estimate is indistinguishable from zero. The evidence supports a temporary response profile in the selected event sample, not a claim that whale alerts permanently raise Bitcoin risk.