From a 2016 meta analysis of whether or not the general wealth and social infrastructure, or the individual specific wealth and level of social access (here C and E), or measured IQ (here Genes, or A), is more correlated with long term Socioeconomic Status, basically, networth normalized against the rest of your society.
Notice how IQ Variance increases dramatically as SES goes up.
IE, the wealth of a person becomes a less and less reliable indicator of their assesed intelligence, the more wealth they have... whereas their general starting point in life can be predicted more and more reliablely, the wealthier they currently are.
... will update with link to the study when I remember where I put it.
EDIT: found it
https://pmc.ncbi.nlm.nih.gov/articles/PMC4749462/
EDIT 2:
Probably also worth noting that this study encompassed the US, Europe and Australia, possibly more places... but the graph here is ...
... well it is specific the to the US, this dramatic and statistically significant discrepancy only exists in the US, and the authors end up devoting the majority of the paper to explaining how many ways this discrepancy is statistically valid, and how odd this is.
They don't fully attempt to explain why this is, but I think its fairly obvious: The US in particular is extremely unmeritocratic, extremely nepotistic, extremely financially exploitative, and of course has an extremely laughable education system in comparison to the other countries with data sets included in the study.