Great to see that your scores seem to correlate with performance :)
One kind suggestion: could you provide more robust statistics than the simple average, such as the median, i.e. the 50th percentile? The simple average can easily be skewed by outliers while the median is a more skill-revealing imo and straightforward measure.
But three of the four categories only had 6 points of data, so how much more does the median mean. Also, had you invested equally in each I think you’d be more interested in the average return.
I will rank the data groups next time by return rather than by rating so medians can easily be plucked out though.
What an interesting write-up! Thanks Wolf. I'm surprised how well the 3-star ratings have performed. Also great to see you put your results out there like this to be scrutinized.
"I took the last closing price prior to when their financials were released" - this should really be something like the closing price 1-2 trading days after your rating is released, to be a realistic measure of the value of your rating to users who need time to read and consider it. You could measure it twice; once for when subscribers see it and once for when non-subscribers get it, to show the value to subscribers.
I think you're giving yourself all the credit for the movement in the price in the first day or two, and none to the financials themselves. I'm not assuming anything about the speed of the review, I'm focusing in the speed of realistically and responsibly actioning them. Some of us do actually do our own DD.
Great to see that your scores seem to correlate with performance :)
One kind suggestion: could you provide more robust statistics than the simple average, such as the median, i.e. the 50th percentile? The simple average can easily be skewed by outliers while the median is a more skill-revealing imo and straightforward measure.
Cheers
I suppose that’s fair and I get your point.
But three of the four categories only had 6 points of data, so how much more does the median mean. Also, had you invested equally in each I think you’d be more interested in the average return.
I will rank the data groups next time by return rather than by rating so medians can easily be plucked out though.
fabulous - confirms my bias to loving your research and generally being a skeptical mofo.
4.75/5
What an interesting write-up! Thanks Wolf. I'm surprised how well the 3-star ratings have performed. Also great to see you put your results out there like this to be scrutinized.
"I took the last closing price prior to when their financials were released" - this should really be something like the closing price 1-2 trading days after your rating is released, to be a realistic measure of the value of your rating to users who need time to read and consider it. You could measure it twice; once for when subscribers see it and once for when non-subscribers get it, to show the value to subscribers.
That would measure or at least try to my reviews impact on the price.
I suppose that’s interesting but immeasurable IMO. 90% of reviews are read in the first 24 hours, many are done pre-market, some several days later.
If you were to try to measure it, how long after do you measure the results? Certainly, not over a year. Not sure I see the benefit.
I think you're giving yourself all the credit for the movement in the price in the first day or two, and none to the financials themselves. I'm not assuming anything about the speed of the review, I'm focusing in the speed of realistically and responsibly actioning them. Some of us do actually do our own DD.
How do you arrive at that? I’m measuring reviews from 11-14 months ago.