In the 2018 essay question from CFAI, the vignette was stated as follows and the question is asking which method is appropriate to construct the portfolio:
Välimaa next considers the transition of the Fund’s portfolio holdings, which have a total market value of EUR 150 million. She is constructing the portfolio using individual equities and considers the following methods: full replication, stratified sampling, and optimization. The benchmark for the portfolio is the FTSE Eurotop 100 Index, which is based on market capitalization and consists of 100 of the largest publicly traded European companies. The investment committee prefers not to use sophisticated algorithms that are difficult to understand.
The answer results says Full replication since the portfolio of EUR 150m is already large enough to replicate the Index.
I think here they’re focused on the 150 million size of the given portfolio, not the size of the benchmark index. Since they want the simplest approach (replication) the question is whether the portfolio is big enough to fully replicate the index.
You are correct that knowing the exact market caps of the index components will be necessary for replication and rebalancing. But you have 1.5 million euros to allocate on average to each of the 100 index holding names even in the most evenly distributed market cap weighting scenario. So I guess they are saying that’s enough to successfully replicate, you won’t hit a wall trying to replicate 100 stocks with 150 million in funds.
