This may be trivial pursuit and simple but i just can’t seem to understand this… Under Equity reading page 181 of the curriculum outlines how stratified sampling works and what do in cases when the weight of securities is less than the cell weight.
the example given is: " suppose a cell contains 2 % of weight of the index and there are 2 stocks chosen to represent the index and have weights of 0.3 and 0.5, and by over weighting each security equally by 0.6% you can gain the same exposure to that cell factor"
However in the footnote it gives an example of precise approach where it says to increase the security weights so as to maintain their relative proportion of 0.3/0.5= 0.6. What i don’t understand is how did they end up getting the weights of 0.75% and 1.25% for each security??
The other question i have is could there be a situation where the securities contained in the cell may aggregate to a weight greater than the cell itself? and what to do in such instances?
What is the key feature, I mean an advantage of stratified sapling toward optimization method beside is more appropriate for illiquid stocks, always marked as small caps in questions? Is there any else advantage which optimization can’t do better?
Schweser video focused in on the cost difference was a big factor. Stratified sampling will do the job for much cheaper, but quality wise it won’t be as good as optimization. Mixing the 2 you can still keep costs low but improving it by a good bit in certain specific areas using optimization.
TLDR: it all comes down to cost (bang for buck) is my take away