holdings based vs returns based

it says in the text that returns based can be subject to window dressing, while holdings based can not. What exactly do they man by that? can anyone elaborate?

I think u might have it backwards.

Window dressing is basically when a manager tries to cover up poor performance by, say, adding a position near the end of the reporting period that has high momentum or isn’t part of the strategy (taking a position in an in-favor sector that isn’t in line with the IPS for example) to improve returns.

If you analyze returns based on holdings, all you’re doing is looking at the most current composition of the portfolio- this is subject to window dressing because the fund manager can pretty-up the portfolio before reporting and might not be subject to scrutiny through analysis of past returns- so he could add Amazon to the portfolio at quarter end to make it look good and all you would see is the current positions, assuming that they were probably part of the investment strategy all along.

If you use a returns-based analysis, then you’re identifying the specific drivers of return for that period and using actual past factor exposure data- effectively scrutinizing returns more and making it much harder for fund managers to cover up any less-than-stellar performance with a quick position change. This has its own disadvantages however (that I can’t remember tbh).

Someone lmk if this in inaccurate

i see. That’s great! thank you!