Nice work Dutch guyUncorrelated wrote: ↑Tue Nov 05, 2019 3:53 pmI suspect I'm the "Dutch guy online". Here are my results with the momentum factor included:305pelusa wrote: ↑Tue Nov 05, 2019 12:10 pmAh sorry, I did not notice that was the case with ZPRV. I just saw it has an MSCI index, which is easy to get data for.

When I get a chance, I'll regress it and compare values. Some things that matter to me are R^2 values and alphas, which I didn't see for ZPRV in your comment. Also, I'm not necessarily trustful of "a Dutch guy online".

Regardless, provided their values are correct, I think you came to a reasonable conclusion.

IJS daily data (actual fund performance)Code: Select all

`| |annualized alpha|beta |SmB |HmL |RmW |CmA |MoM :--|:--|:--|:--|:--|:--|:--|:-- |IJS, USA small cap value weighted | -0.19%| 1.04| 0.92| 0.21| 0.17| 0.06| -0.06| |t stat | -0.07| 4.04| 48.06| 10.57| 6.58| 1.94| -4.86| expected outperformance: 1.56% Based on 18.8 years of data, 263 data points per year R^2 adj = 0.376`

ZPRV daily data (index)Code: Select all

`| |annualized alpha|beta |SmB |HmL |RmW |CmA |MoM :--|:--|:--|:--|:--|:--|:--|:-- |MSCI USA small cap value weighted | -1.82%| 1.03| 0.68| 0.26| 0.07| 0.16| -0.12| |t stat | -1.41| 5.50| 76.99| 23.37| 5.36| 9.18| -17.90| expected outperformance: 1.25% Based on 4.9 years of data, 308 data points per year R^2 adj = 0.889`

ZPRV monthly data (index)Source: own calculations based on MSCI index data and Ken French data library. Expected outperformance assumes that factors with t-stat below 2 are insignificant (equal to zero) and the remaining factors get 50% of the historical factor performance. Expected outperformance is before fees / turnover / tax inefficiencies.Code: Select all

`| |annualized alpha|beta |SmB |HmL |RmW |CmA |MoM :--|:--|:--|:--|:--|:--|:--|:-- |MSCI USA small cap value weighted | 0.30%| 1.05| 0.75| 0.29| 0.30| 0.07| -0.21| |t stat | 0.35| 2.66| 31.25| 9.35| 8.87| 1.54| -14.33| expected outperformance: 1.17% Based on 24.4 years of data, 12 data points per year R^2 adj = 0.849`

The R^2 on IJS is atrocious. Not sure what's up with that. The factors loadings themselves are almost identical to portfolio visualizer's numbers

What could the low R^2 of IJS mean?