For a continuous semimartingale , its quadratic variation is the continuous increasing zero-starting process obtained as the limit, uniformly on compact time intervals in probability, of sums of squared increments along deterministic partitions whose mesh tends to zero. If is its continuous local-martingale/finite-variation decomposition, then ; finite-variation terms and their cross sums vanish.
Write . If , then for every ,
First send to infinity and then to infinity. Thus convergence in probability under implies convergence in probability under , also for a supremum on a compact time interval. The same squared-increment sums converge under to the version of . Since is also a semimartingale under , these sums converge to its quadratic variation. Uniqueness of limits in probability gives equality at every fixed time; continuity and countably many rational times give
Absolute continuity suffices; equivalence is not required. This is quadratic variation under an absolutely continuous measure change.
Use just the parameters and . Positivity and adaptation of their local martingales give
Thus both processes are adapted. A positive continuous local martingale has a semimartingale logarithm, by Itô formula after localization away from zero. Since and has finite variation, is a continuous semimartingale. Write , where is a zero-starting continuous local martingale and is continuous finite variation, also starting from zero.
Apply Itô formula to each exponential:
Because is itself a local martingale, the finite-variation term is zero by part (a). Divide its finite signed measure by the strictly positive . For , respectively,
Adding and subtracting show and . Therefore the exponential criterion for a continuous local martingale and its bracket gives
The semimartingale property was established before applying Itô's formula to ; it was not assumed from the outset.
If and a continuous semimartingale is a semimartingale under both measures, its quadratic variations agree -indistinguishably. The same squared-increment sums converge uniformly on compacts in probability under both measures: absolute continuity transfers the original convergence, and uniqueness of the limit identifies the two continuous versions. Equivalence of measures is unnecessary.