Been gone awhile so happy to see the MFO site activity remains compelling and dynamic and new design much easier to navigate. Earlier advice in 2011 on Daniel Solin and Burton Malkiel books saved me serious money by not hiring a portfolio manager for my IRA. Much appreciated. So now I'm wondering if M* performance results factor in expenses. For instance, if a fund shows a 3-year avg annual return of 10 percent and its expense ratio is 2 percent, is that 10 percent net after expenses? Thanks.
Long answer: "Annual total returns are calculated on a calendar-year and year-to-date basis. Total return includes both capital appreciation and dividends. The year-to-date return is updated daily. For mutual funds, return includes both income (in the form of dividends or interest payments) and capital gains or losses (the increase or decrease in the value of a security). Morningstar calculates total return by taking the change in a fund's NAV, assuming the reinvestment of all income and capital gains distributions (on the actual reinvestment date used by the fund) during the period, and then dividing by the initial NAV. Unless marked as load-adjusted total returns, Morningstar does not adjust total return for sales charges or for redemption fees. Total returns do account for management, administrative, and 12b-1 fees and other costs automatically deducted from fund assets."
Just hit "edit" on your post, and you can change anything that you'd like to, if it didn't come out exactly as you were thinking the first (or second, or third...) time. I do that frequently myself, as I can hardly ever get it right the first time.
Since you haven't been around for a while, you might want to take a look under "Resources"... there's a fairly decent "User Guide" there which can help anyone to get the most out of the site.
A fund's expense ratio already includes 12b-1 fees. If M* were to take note of the 12b-1 fees explicitly, it would be double-counting those fees.
Note that M*'s performance results don't really explicitly consider any fees - they just report the net performance of the fund. If a fund starts at $10.00 at the beginning of the year, ends at $10.50 after distributions, has $0.50 in distributions (on Dec 31st), then its performance is 10% for the year. The fund itself deals with the expenses - not only those reported in the ER, but other expenses like trading costs (commissions). That's all baked into the NAV.
On the other hand, M* considers all loads when computing its ratings - that includes front end loads and contingent deferred sales loads (back end loads) as well as those loads already incorporated into the ER (level loads). M* "adjust[s] returns for maximum front-end loads, applicable deferred loads, and applicable redemption fees." The latter includes even Vanguard fees for buying or selling a few of its funds where the fee goes into the fund assets (and is thus not considered a load). See: M* Data FAQ.
Above, I spoke disparagingly of double counting. There is one area where M* does double count. When it computes overall ratings, it combines a fund's 3 year, 5 year, and 10 year ratings. In doing so, it triple-counts the performance (and risk) data from the first three years (since those data are used in computing the 5 and 10 year ratings as well as the 3 year ratings), and double-counts the data from years four and five. This is not undesirable, as it has the effect of weighting recent performance more heavily. Aside from the overall rating calculations, I know of no other calculation performed by M* that double-counts data.
I stand corrected, looks like, on M* not incorporating loads into its ratings. I'm not sure where I got impression they did not, but thanks for catching...I hate being wrong and hate perpetuating wrong info even more.
I'll have to double-check, but it may be because I believe load is not reflected in M* Sharpe and Sortino calculations. I've always assumed its star assignments were based on Sharpe, or Sortino, or some variation thereof...but digging a little deeper, it's more subtle.
Below are a few publications that provide insight into formulation of categorized risk adjusted return used in Morningstar's trademark star ratings. I believe it involves subtracting a fund's monthly variability from its excess return then comparing the result with category averages to assign star numbers.
Here is a 1997 paper from Stanford's website: Moringstar's Performance Measures.
And, the discussion of "expected utility theory" in Morningstar's Ratings Methodology publication.
Finally, a detailed 2002 M* publication with more specific formulation of how the utility function is applied: The New Morningstar Ratings Methodology. Load incorporation is specfically noted.
It reminds be a bit of the risk formulation used in the legacy ThreeAlarm methodology.
In any case, thanks to you, I will be more sensitive to these subtleties going forward.