Here's a statement of the obvious: The opinions expressed here are those of the participants, not those of the Mutual Fund Observer. We cannot vouch for the accuracy or appropriateness of any of it, though we do encourage civility and good humor.
[snip] ”The automated essays "analyzing" those funds where the "Q" is used is utterly comical.”
Yes - They sound computer written. There’s an uncomfortable “sameness” to the composition. They seem to rely a lot on total years of experience of a fund’s management team. And, their overall assessment of the firm seems to weigh heavily in their individual fund appraisal.
Morningstar launched their Quantitative Rating for funds in 2017 and later (2021?) added complementary written reports. Their objective was to greatly expand the number of funds under coverage. Machine learning is utilized to mimic how a M* analyst would rate a fund. I haven't investigated the efficacy of the Quantitative Rating. However, the corresponding written reports are often nonsensical. Consequently, I pay scant attention to the "Q" rating. The quality of the original M* Analyst Ratings varies and largely depends on which analyst scrutinized the fund.
- Charts with rolling returns (customizable) - Returns of entire asset classes / categories over various time frames (1 month to 5 years) - Brokerage availability of mutual funds / share classes showing TF / NTF
and many more. All gone.
Of course, M* Discuss was destroyed. Largely during the tenure of Chief IT Office Gregg Goff, who drew a salary of $1.4 million. Go figure.
Starting THIS year, Q ratings designation was removed (or, merged into regular ratings) because M* felt that those were almost as good as live analyst report. However, some flags for Q remain and I can sense a Q report just from the gobbledygook.
M* has left the individual investors behind. It wants to be a mini-Bloomberg by targeting professional investors. Some of its professional software now cost $10,000/seat/yr, and one sale there may cover lots of old Premium subscriptions. Much of that software was developed and refined with free and low-cost individual subscribers.
It's no use complaining. I have moved on to Stock Rover (SR).
”However, the corresponding written reports are often nonsensical.”
Yes, I think it’s the written reports that I’m seeing. Some interesting content but mostly mundane garbage IMHO. If they pop up for free worth a quick read.
”However, the corresponding written reports are often nonsensical.”
Yes, I think it’s the written reports that I’m seeing. Some interesting content but mostly mundane garbage IMHO. If they pop up for free worth a quick read.
I use Stock Rover too, and it's helpful. But not much for FUNDS.
That is my concern as well! I will attempt to manipulate my brokerage available data, but so far, it is far less detailed than what has previously been offered by M*.
Only Stock Rover (SR) premium & above have portfolio transactional capabilites. Others I looked at (PV, etc) are really for portfolio analytics, so info with % weights or $amounts must be uploaded via Excel or brokerage connection (I won't do that).
Unfortunately, most are moving to just analytics bypassing the complicated transactional stuff. M* was far ahead of its time in providing free portfolio transactional capabilities with auto-updates for distributions, but that may go away beyond 2023.
[snip] ”The automated essays "analyzing" those funds where the "Q" is used is utterly comical.”
Yes - They sound computer written. There’s an uncomfortable “sameness” to the composition. They seem to rely a lot on total years of experience of a fund’s management team. And, their overall assessment of the firm seems to weigh heavily in their individual fund appraisal.
Morningstar launched their Quantitative Rating for funds in 2017 and later (2021?) added complementary written reports. Their objective was to greatly expand the number of funds under coverage. Machine learning is utilized to mimic how a M* analyst would rate a fund. I haven't investigated the efficacy of the Quantitative Rating. However, the corresponding written reports are often nonsensical. Consequently, I pay scant attention to the "Q" rating. The quality of the original M* Analyst Ratings varies and largely depends on which analyst scrutinized the fund.
Color me skeptical (cynical?), but I suspect the real objective was not so much to expand the number of funds "analyzed" as it was to decrease the number of funds intelligently analyzed (i.e. by real, human analysts).
Before M* introduced Q ratings, it had been criticized for reducing the number of funds analysts covered. Adding Q ratings allowed M* to say that it was still providing medal ratings (albeit computer-generated), yea verily even expanding this coverage. My suspicion (unverified) is that the number of funds covered by real live analysts has continued to decline. By adding computer-generated "explanations" and subsequently dropping the "Q" designation, M* is able to reduce coverage somewhat undetected.
As @hank said, there is a sameness to the computer generated text. The writeups have a fill-in-the-blanks feel, like Mad Libs.
The medal ratings are generated by a machine learning program that is fed lots of training (historical medal rating) data to come up with a classification (medal rating) heuristic. This heuristic works cryptically without explanation. Note the mention of "correlation" in the M* interview cited. Correlation offers neither causes nor explanations. See Super Bowl indicator.
In contrast, the written text feels as if it comes from filling in a template using a much smaller (hand crafted?) set of explanatory factors. Again as hank noted, years of experience being one such factor.
I've not tried to look for discrepancies between the text ratings (process, people, parent) and the medal (gold, silver, ...) ratings, as people did when (real) analyst ratings first came out. Not worth the effort.
M* Investor has good analytics, but lot of it just showcases M* databases, & not what many investors may need or want. Its shortcomings are in its portfolio functions (transactional or overview portfolio data).
M* originally promised not to retire the old M* Portfolio until the M* Investor catches up. But that promise is now only through 2023.
I still have a M* subscription because of the stock coverage. I own two of the Van Eck “moat” ETFs and I occasionally dabble in an undervalued stock or two. The MF coverage continues to grow worse and worse as many others have pointed out.
"Color me skeptical (cynical?), but I suspect the real objective was not so much to expand the number of funds 'analyzed' as it was to decrease the number of funds intelligently analyzed (i.e. by real, human analysts)."
"Before M* introduced Q ratings, it had been criticized for reducing the number of funds analysts covered. Adding Q ratings allowed M* to say that it was still providing medal ratings (albeit computer-generated), yea verily even expanding this coverage. My suspicion (unverified) is that the number of funds covered by real live analysts has continued to decline. By adding computer-generated 'explanations' and subsequently dropping the 'Q' designation, M* is able to reduce coverage somewhat undetected."
I just hope people aren’t paying for this “analysis” - or, for that matter, buying whatever the robot likes and doing themselves harm. I will say, there’s a trove of useful data on M* if you care about return history and what is inside a fund. I appreciate that much of that is free, Of course, it’s available elsewhere as well.
I have depended on M* for many years, as part of my bond oef investing experience. I took full advantage of the analytical data associated with bond oefs. I have built a pretty effective "watchlist" system at M*, focusing on "bond category" listings. I have screened a large number of funds, and included about 10 funds in each category. For each fund in the category, I include TR performance over different periods of time, and include a number of risk data points. That has allowed me to visually see performance trends, and risk/reward parameters. It has served me well, but it appears that is about to end as M* devalues services to individual investors. I have not found a good alternative, so it appears my ability to "return" to bond oef investing will be much more challenging, with much less detailed information. I have not been investing in bond oefs since March of 2022, and I have to evaluate how I might do that in the future, without the historical M* support features, that have allowed me to make more informed decision making of the past.
Comments
Morningstar launched their Quantitative Rating for funds in 2017
and later (2021?) added complementary written reports.
Their objective was to greatly expand the number of funds under coverage.
Machine learning is utilized to mimic how a M* analyst would rate a fund.
I haven't investigated the efficacy of the Quantitative Rating.
However, the corresponding written reports are often nonsensical.
Consequently, I pay scant attention to the "Q" rating.
The quality of the original M* Analyst Ratings varies
and largely depends on which analyst scrutinized the fund.
Link
- Charts with rolling returns (customizable)
- Returns of entire asset classes / categories over various time frames (1 month to 5 years)
- Brokerage availability of mutual funds / share classes showing TF / NTF
and many more. All gone.
Of course, M* Discuss was destroyed. Largely during the tenure of Chief IT Office Gregg Goff, who drew a salary of $1.4 million. Go figure.
M* has left the individual investors behind. It wants to be a mini-Bloomberg by targeting professional investors. Some of its professional software now cost $10,000/seat/yr, and one sale there may cover lots of old Premium subscriptions. Much of that software was developed and refined with free and low-cost individual subscribers.
It's no use complaining. I have moved on to Stock Rover (SR).
Yes, I think it’s the written reports that I’m seeing. Some interesting content but mostly mundane garbage IMHO. If they pop up for free worth a quick read.
Unfortunately, most are moving to just analytics bypassing the complicated transactional stuff. M* was far ahead of its time in providing free portfolio transactional capabilities with auto-updates for distributions, but that may go away beyond 2023.
Before M* introduced Q ratings, it had been criticized for reducing the number of funds analysts covered. Adding Q ratings allowed M* to say that it was still providing medal ratings (albeit computer-generated), yea verily even expanding this coverage. My suspicion (unverified) is that the number of funds covered by real live analysts has continued to decline. By adding computer-generated "explanations" and subsequently dropping the "Q" designation, M* is able to reduce coverage somewhat undetected.
As @hank said, there is a sameness to the computer generated text. The writeups have a fill-in-the-blanks feel, like Mad Libs.
The medal ratings are generated by a machine learning program that is fed lots of training (historical medal rating) data to come up with a classification (medal rating) heuristic. This heuristic works cryptically without explanation. Note the mention of "correlation" in the M* interview cited. Correlation offers neither causes nor explanations. See Super Bowl indicator.
In contrast, the written text feels as if it comes from filling in a template using a much smaller (hand crafted?) set of explanatory factors. Again as hank noted, years of experience being one such factor.
I've not tried to look for discrepancies between the text ratings (process, people, parent) and the medal (gold, silver, ...) ratings, as people did when (real) analyst ratings first came out. Not worth the effort.
I can see portfolio analysis at M* Investor
Overview Holdings Xray Performance Stock intersection
M* originally promised not to retire the old M* Portfolio until the M* Investor catches up. But that promise is now only through 2023.
If I cancel, it will be the first year in probably 30 years I have not had a membership or sent money their way
Nice to get something right for once.
"Before M* introduced Q ratings, it had been criticized for reducing the number of funds analysts covered. Adding Q ratings allowed M* to say that it was still providing medal ratings (albeit computer-generated),
yea verily even expanding this coverage. My suspicion (unverified) is that the number of funds covered
by real live analysts has continued to decline.
By adding computer-generated 'explanations' and subsequently dropping the 'Q' designation,
M* is able to reduce coverage somewhat undetected."
@msf,
Interesting observation and hypothesis.
Thanks!