A client wants to minimize ongoing rebalancing costs and tracking error. He has chosen to index a part of the portfolio to the S&P 500 Index. Which construction method would be more appropriate and why:
Stratified sampling
optimization
full replication
Optimization.
I’d say full replication. There was almost inentical question in Schweser or Curriculum and full replication was adviced. When index contains large-cap, liquid stocks and there <1000 of them then this method is right.
It may be the case that investor’s ‘part of portfolio’ is too small to invest in SP500 but we have no such info. Insted minimization of rebalancing and tracking error are, again, in favour of full replication.
I agree with kobi. CFAI explanation is found in Question 20 of Reading 23.
full replication is costly. i choose optimisation. whats the correct answer?
Full replication has the advantage of being “self rebalancing” (given the fact that is value-float weighted). Full replication minimizes tracking error.
Regards,
Thanks guys, but isn’t full replication more costy?
It is upfront, but the client wants to minimize “ongoing rebalancing costs,” not upfront transaction costs.
Just went reviewed this chapter. Answer is full replication. No rebalancing costs and lowest, if any, tracking risk. IsThereAny made a good point about “upfront” costs, but that isn’t mentioned as a constraint in the problem.
My mistake: it was way too late at night for me to be answering questions.
Yes, for a market-value weighted index that likely has little change in its constituent securities, full replicatoin is the answer.
U blow your load with full replication and your done for the night.