I'm not a medical researcher, but I think you're misinterpreting these results. The efficacy does not "turn negative". It's negative for some time interval lag

I'm not a medical researcher, but I think you're misinterpreting these results. The efficacy does not "turn negative". It's negative for some time interval lagged after vaccination. This is expected, because some of the unvaccinated control will have been infected before that time interval. If you look at test and control 200 days after infection, the number of vaccinated individuals who were infected at any time is much smaller than unvaccinated. Do I have this right?