| Index | Recent Threads | Who's Online | User List | Search |
|
|
 |
Portfolio123 » List all forums » Forum: Ranking Systems and Factors/Functions » Thread: Ranks appear to be incorrect |
|
Total posts in this thread: 6 |
[Request a Feature]
[Post new Thread] [Add To My Favorites] [Watch this Thread] |
| Author |
|
|
hihint
Advanced Member
|
The rankings associated to my screener "Fundamental Rank Active" appear to be incorrect. According to my own calculation, PAAS, SVM, RIC should have lower ratings than BPI and CPLA, but currently the ranking system rate them higher. Please take a look. Thanks. |
||
|
|
marco
|
Looks like you are using Industry ratios in your ranking system. PLease see my latest post regarding problems with industry ratios. ---------------------------------------- Portfolio123 Staff |
||
|
|
marco
|
Still investigating. ---------------------------------------- Portfolio123 Staff |
||
|
|
hihint
Advanced Member
|
Thx marco. Were you referring to "Pr2BookQInd are N/A for all stocks" as your "latest post regarding problems with industry ratios"? The N/A problem was already fixed and all the data appear to be correct. I downloaded the data (supposedly to be correct) and run the calculation on my own but my calculation gives different ratings... In another post I noticed that the ranking system was "rearchitectured". Maybe it introduced some sort of glitches... |
||
|
|
marco
|
In trying to recover from reuters dropping several ratios we introduced a problem in ranking factors using industry ratios. Please try it now. Please note that we are still not happy with the current situation of industry ratios. At the moment some are still coming from reuters and others are being calculated by us as medians (our quick fix when we found out). But for the ones that are still coming from reuters, the industries we use are being mapped to industries in their new taxonomy (as I explained in another post) therefore there's a historical dis-continuity. We need to completely replace the industry ratios with our own methodology which we have developed but not yet released. This will ensure historical continuity and consistency. It's a painful but necessary step, but it is an improvement. Our methodology was developed by our in-house analysts and it's a combination of market cap weighted averages (with some trimming for outliers) when the industry is large enough, and medians when it's not. We tried very hard to reproduce reuters industry ratios unsuccessfully. The logic that calculated reuters industry average is done by some old legacy system and nobody at reuters knows exactly what was being done (the developers of that logic are long gone). Rest assured that our algorithm is perfectly reasonable and we will explain it in detail shortly. But it will produce somewhat different results depending on your usage. Marco ---------------------------------------- Portfolio123 Staff |
||
|
|
hihint
Advanced Member
|
Thanks a lot marco. Now I see it producing similar results with my own calculation. Industry ratios are a crucial component in my strategy. Also I believe continuity and consistency are crucial to any investment strategies. Your effort trying to provide consistent industry ratios is highly appreciated. It's understandable that there might be differences here and there. But it shouldn't be a problem as long as it's coherent. I'm looking forward to the new in-house ratio engine. Hopefully I can see it soon. Best |
||
|
| [Show Thread Printable Version] [Post new Thread] |
