It completely depends on what you're doing and what you're comparing it to.
Running events has relatively high overhead compared to JS code, so if you have event-heavy code doing things like intensive calculations with lots of loops, the equivalent JS may be faster.
On the other hand, some parts of the Construct engine are extremely well optimised, and have been carefully honed over the years with all our long experience of performance testing JS code. If you write your own scripts to do your own code instead of using the engine, it could well be significantly slower, unless you put in the same effort we have. One example off the top of my head is if you ignore Construct's collision engine and write your own custom one in script, it could well be significantly slower. Not only could it be slower in absolute terms (slower when doing the same work), but it could be slower in algorithmic terms (as in getting exponentially slower as your game gets bigger, whereas Construct's engine is highly scalable and can keep doing about the same amount of work no matter how much bigger the game gets).
As ever performance benchmarks are the only way to know for sure.