The zacks historical database contains more stocks than VLine and the Zacks rating system has outperformed the VLine #1 timeliness stocks the last 10 years.I still don't feel Vline ratings belong in a mechanical strategy as a portion of the ratings are subjective whereas Zacks rates stocks on earnings surprise.Also, VLine's limited database (just 1400 stocks 10 years ago) is limiting.rick
From the Zacks website (www.zacks.com)....Zacks Rank:The Zacks Rank rates stocks in terms of their expected price performance over the next three to six months. The Rank is a quantitative model based on trends in estimate revisions and EPS surprises. Stocks are ranked from 1 to 5, with 1 being the best rank ("strong buy") and 5 being the worst ("strong sell"). Consider purchasing stocks with a Zacks Rank = 1 ("Strong Buy"). During the 19 years, 1979 - 1998, the Zacks Ranked #1 stocks provided an average annual return of 35.1% vs. only 17.4% for the S&P 500. The Zacks Rank is also a key underlying component of many of the portfolios recommended in the Zacks Advisor.
One last note...In reviewing what i just submitted regarding alternatives to Value Line, it seems that the Zacks #1 ratings have outperformed the majority of the 5 or 10 stock strategies we look at. Wouldn't this make it a great starting place?and "no", I don't work for Zacks. However, I have used it in the past.
So ? What's next. Was not aware of the Zacks database, is it readily avail ?The S&P certainely is. The idea is not to develop new screens but rather to exply all the existing ones on S&P.
Find out if Zacks sells data disks for the past. If not, we'd have to start collecting data now. We probably wouldn't want to start using it to develop screens for 10 years.
why isn't the idea to develop new screens as well?
Looks perfect to me. Would love to see our favorite screens applied on Zacks#1.How can I contribute ?
they actually sell a backtesting program ($5,000) but it is extremely user unfriendly.would there be a way to take the hisorical data from this program and put in a user-friendly form?
I have my own tool (selling that Decision Support Software around the world was my business for 20 years) terribly UNfriendly but very powerful. If we can extract ASCII data from Zacks disk, then I can start working. Still have to foot the bill though.
see http://zrs.zacks.com/docs/backtest.htm for info.I would love to contribute $ if we can work on the user format. this database also allows holding by marketcap (proportions) and also has historical Zacks ratings.
I would love to contribute $ if we can work on the user format.**Why not talk to Jamie Gritton? His backtest software is already written!!-Greg
Indded, and by the way I can contribute too in purchasing back data. What about future data
I believe the backtesting tool is part of an overall package that includes updates
Greg,would it be possible to utilize the database in a fashion similar to Q Investor?
>>Indded, and by the way I can contribute too in purchasing back data. What about future data?Future data! I'll chip in for that!
I think I mentioned this quite some time ago, but before I started following MI as practiced here I had spent a few years doing screening using Zack's and Telescan software and data. Since I was only a few blocks from them, Zack's would hand deliver disks, diskettes actually, to me every Friday. By that mechanism, I had accumulated at one point data for about two years. The problem, as I recall, is that their historical data base did NOT include companies which were defunct either due to M&A or natural causes. I am not sure whether their historical data reflects their rankings adjustments.How much that matters for Zack's data is uncertain, but for VL it matters a lot more than you might think. About a month ago I looked at symbol lists that MakeItJake had been kind enough to send me for VL T1 data. In comparing the Jan 98 list with the Dec 98 list, I found that only 4 disappeared due to M&A in 98, another 7 went in 99 and 4 more in 00. But alarmingly, at least to me, only 15 of the 100 T1 stocks on the Jan 98 list were still on it in Dec 98, so great is the cumulative effect of VL's constant revision. Zack's backtester also did not allow ranking of criteria but instead required numerical values, much like the filtering criteria in VL currently. All things considered, one should not underestimate the value of the unadjusted historical data that are used here. arezi
I've been using the Zacks Research Wizard product since mid 1997 to screen for stocks to invest in and list on my web site. Similar to using Value Line, I first download the weekly data (on over 6000 stocks) from the Zacks web site . The download is almost 6mb in size so it takes around 30min to download with a 56k modem and around 3min with a cable modem. I then use the Zacks Research Wizard software to export the data items I want to follow to a text file and then import the text file into Excel for analysis. I keep the past historical data to test ideas. I've zipped and posted a recent weekly tab-delimited text file (6,244,861 bytes after un-zipping) created with Research Wizard at:http://briefcase.yahoo.com/arthaynerandhttp://q5research.com/20000929rw.zipInterested readers of this board can download it and get some idea of the info available. The Zacks ranking is in column X if you import the file into Excel. You can find additional info at: http://firstname.lastname@example.org
The Rank is a quantitative model based on trends in estimate revisions and EPS surprises. Stocks are ranked from 1 to 5, with 1 being the best rank ("strong buy") and 5 being the worst ("strong sell").Is this all that is known about the Zacks ranking system? If the rules are not fully known it becomes somewhat of a black box from our standpoint.I have previously mentioned Rabbitt Analytics, whose ranking rules are disclosed.I have also raised the point that a Value Line T1 or T2 ranking could in and of itself introduce a bias which influences a stock's future performance. Unfortunately there is no way to prove or disprove this.
Best Of |
Favorites & Replies |
Start a New Board |
My Fool |