GIPS: Composite Return Calculation Provisions
If you're looking at customer reviews for a product you're considering buying, lots of positive or negative reviews will carry more weight than a single positive or negative review. Thinking along these lines, what would be the best criterion for investors evaluating their asset manager?
Not really.
There could be external reasons why the portfolio performed this way, including time and withdrawals.
No, actually.
The best portfolio might not capture all the risk and volatility of the investment strategy.
It's also important, though, to consider how the portfolios should be weighted inside the composite, because some larger portfolios may give the manager more flexibility in implementing the strategy. What do you think a good approach to weighting would be?
In using the beginning values, the returns are simply multiplied by the percentage of weighting of the portfolio. But the method that captures both the beginning value _and_ the external cash flows is a little different. It's actually similar to one of the detailed historical rate of return calculations, which factors in beginning values and external cash flows. Which calculation does this?
Not really.
Different portfolio sizes can impact how the manager implements the strategy.
Well, no.
The composite could contain only a few small and large portfolios, in which case the median portfolio wouldn't capture the full strategy.
Not really.
The Dietz method isn't the most detailed because it assumes the midpoint for the external cash flow.
Indeed it does.
The modified Dietz method breaks down returns to the day of the external cash flow, and it's required by GIPS to be used during certain time periods. Recall that the modified Dietz method's denominator calculates the sum of the beginning assets and weighted external cash flows. That makes it a perfect fit, and the denominator equation is
$$\displaystyle V_P = V_0 + \sum^{N}_{i=1} (CF_i \times w_i)$$.
You can then use this equation to calculate the composite return,
$$\displaystyle \displaystyle r_c = \sum \left ( r_{pi} \times \frac{V_{0,pi}}{\sum V_{0,pi}} \right )$$,
where
$$\displaystyle r_c$$ is the composite return,
$$\displaystyle r_{pi}$$ is the individual portfolio's return, and
$$\displaystyle V_{0,pi}$$ is the beginning value of the portfolio.
No.
The geometrically chain-linked return is the current GIPS requirement, so it's not historically based.
Additionally, you can also use the modified Dietz method directly by treating the entire composite as a single portfolio. If you're careful with the direction of cash flows, the equation is
$$ \displaystyle r_{ModDietz} = \frac{V_1 - V_0 - CF}{V_0 + \sum^{n}_{i=1} (CF_i \times w_i)}$$.
With any of the methods used, GIPS wants to ensure accuracy in the composite return measurement. What might GIPS require to help ensure accuracy?
You got it!
GIPS specifies that the composites have individually asset-weighted returns calculated at least monthly for periods after 1 January 2010. That's because frequent weightings increase the accuracy of the composite, which can help investors evaluate the manager.
No.
The GIPS standards are more specific about accuracy; there are clear rules.
Not likely.
Weighting assets just twice won't help accuracy.
To sum up:
[[summary]]
Right!
Just like lots of consumer reviews will tell potential buyers how good a product is overall, an average of all similar portfolios will tell investors how the strategy did overall. This actually describes a __composite__, which groups portfolios with the same investment mandate, objective, or strategy.
Precisely.
You'll want to weight the portfolios in the composite by their asset average. That way, the portfolios capture the manager's ability to implement the strategy. According to GIPS, composites are _required_ to be calculated by asset weighting the individual portfolios by the beginning-period values, or a method that captures beginning-period values and external cash flows. Precisely.
The lowest actual portfolio return
The highest actual portfolio return
The average return of all similar portfolios
They should be asset weighted
They should be weighted equally
Only the median portfolio should be used
Dietz method
Modified Dietz method
Geometrical chain-linked returns
Frequent asset weightings
Infrequent asset weightings
Weightings at only the beginning and ending dates
Continue
Continue
