The Simba spreadsheet only tracks annual returns, which makes correlation math rather coarse (just think about the big March 2020 drawdown we experienced, it will probably not show whatsoever in the annual 2020 return). Computing correlation based on monthly returns would be much more meaningful, in truth. And this would be a much better foundation for historical correlation graphs... But monthly returns are outside the scope of this spreadsheet. Tools like Portfolio Visualizer are better at this kind of analysis.

I assume you meant iShares MTUM vs. iShares VLUE (and the MSCI indices allowing to extend corresponding history back in time). Yes, I heard Cliff Asness make this claim multiple times, but this is definitely NOT what the corresponding MSCI indices displayed in the past 40-ish years.calcada wrote: ↑Wed Nov 18, 2020 1:33 pm Also Momentum and Value are supposed to have a negative correlation according to this paper: http://pages.stern.nyu.edu/~lpederse/pa ... ywhere.pdf

"Value and momentum are negatively correlated with each other, both within and across asset classes."

The correlation should be around -0.60 yet it is 0.55 for MTUM:Int'l Value and 0.80 for MTUM:US Value on the spreadsheet. What causes this discrepancy?

I was wondering if this was a side-effect of annual vs monthly returns, so I just ran a quick test with monthly values since 1982 and... same results. Then I varied start/end dates and... same results.

I suspect Cliff assembled a rather theoretical model which might be a tad disconnected from reality. Or he used definitions of value and momentum which are very different from MSCI definitions. Not sure. Could be a good topic of discussion, maybe open a dedicated thread for this?