Why do different sources give different figures for Bitcoin's security budget?
Because there is more than one honest way to build the ratio. The same public data gives 0.92%, 0.86% or 0.82% depending on two choices nobody is hiding. These pages publish 0.82% and show the other two.
The two choices that move the number
Which revenue. A trailing year of miner revenue is a nominal cash sum: each day valued at that day’s price. A spot run rate takes recent blocks and annualises them at today’s price. The first is what miners were paid over a year that is gone; the second is what Bitcoin pays now. Neither is wrong.
Which market cap. Spot market cap is today’s. An average market cap over the same window matches the revenue window. Pairing a trailing-year revenue with a spot cap mixes two periods, which is defensible but worth declaring.
All three, computed on the same data
| Basis | Ratio | Gap / yr | Market cap to cover it |
|---|---|---|---|
| Trailing-year revenue over spot market cap | 0.92% | $1.27B | $24.40B |
| Trailing-year revenue over average market cap | 0.86% | $2.13B | $40.93B |
| Spot run-rate revenue over spot market cap (published) | 0.82% | $2.77B | $53.28B |
Three ways to measure the same ratio, all published. None of the three closes the gap.
Why we publish the lowest one
Basis C answers the present-tense question — what Bitcoin pays today — which is the question these pages are asked. It is also the least flattering to the argument made here: it produces the largest gap, $2.77B a year, where the most generous basis produces $1.27B. Choosing the basis that makes your own case hardest is the only way a reader can check you are not choosing the basis that makes it easiest.
If you see a different number elsewhere
Ask two questions of it: over what window is the revenue measured, and against which market capitalisation. A tracker quoting a lower figure than 0.82% is usually using a longer or older revenue window, a higher one is usually pairing recent revenue with an older cap, and a figure quoted without either is not yet a measurement. The full derivation of every figure here, and the source of every input, is on the methodology page.
What none of the three changes
The 2028 figure. Halving the subsidy on schedule takes the ratio to about 0.41% on every basis, because the halving removes the same fraction of revenue whichever way the revenue was measured, and price cancels in all three. The bases disagree about today. They agree about the direction.