The chances of this data being accurate is close to zero. I have seen data looking at factors for major markets outside the US, and it is wildly inaccurate even as recent as the 1970s. Mainly trading costs and liquidity but delistings too.

In addition, I know someone who looked through the paid data offerings for a similar purpose. He ended up employing ten to twenty people year round to build out a proprietary dataset because it was just not accurate.