Author: rgearyiii

Subject: GTR1 Database Update 1/12/14

Date: 1/12/2014

Recommendations: 64

Index Membership My daily updates of the GTR1 database now include the constituent lists for three indexes: the S&P 500, the Nasdaq 100 and the Dow Jones Industrial Average. Each index has its own field file, described below, where 1 indicates membership and 0 indicates non-membership. The only possible subtlety about these field files is in the timing of membership changes. Each field file is constructed in such a way that, with the default lag of one market day, membership changes take effect in backtests on the first market day that the indexes reflected the changes. For example, on December 11, 2013, it was announced that Alliance Data Systems (NYSE:ADS) would join the S&P 500 on December 20, 2013, after the close of trading. That same week (on December 13), it was also announced that NXP Semiconductors (Nasdaq:NXPI) would join the Nasdaq 100 on December 23, 2013, prior to the market open. In both cases, the respective indexes did not reflect their new members until the regular trading session on Monday, December 23, 2013, so as far as the GTR1 backtester is concerned, the two changes happened at the same time. That is, Monday, December 23, 2013, is the first market date in GTR1 backtests where ADS is flagged as a member of the S&P 500 and NXP is flagged as a member of the Nasdaq 100, provided the default field file lag of one market day is in effect. However, with zero lag, both membership changes take effect at the close of Friday, December 20, 2013. This timing of membership changes in GTR1 backtests is fully consistent with its timing of data from other sources, such as Value Line, where data updates distributed after Friday`s close take effect for trading (at close) the following Monday when the default field file lag of one market day is in effect. The field file sp500.a is built from monthly snapshots of the S&P 500 from December 31, 1973 through October 31, 2006, and daily membership lists from November 1, 2006 through the present. This field file has existed in the GTR1 database since 2010, when the monthly data was donated to me by a major contributor to this message board, but I had too many other tasks (such as working towards daily updates of the GTR1 backtester) keeping me from updating S&P 500 memberships beyond the end of 2009. Now that this field file is kept up-to-date, it is no longer necessary to fall back on the SI Pro field file sp.s for S&P 500 membership in backtests beyond the end of 2009. The new (or rather, previously unannounced) field file nas100.a provides daily Nasdaq 100 membership from January 1, 1995 according to historical changes indicated at http://www.nasdaq.com/indexshares/historical_data.stm and http://en.wikipedia.org/wiki/NASDAQ-100#Components . The current list is obtained directly from nasdaq.com. The single day`s worth of work involved in building the historical membership lists was inspired by comments by Zeelotes suggesting that Nasdaq 100 membership is extremely important in his own successful trading. I have not had much time to explore this data myself due to a long to-do list for the backtester, so I am hoping others will start the research and share results. If anyone can demonstrate the value of this data, I`ll consider looking into what it takes (a lot more than a single day!) to obtain Nasdaq 100 constituent lists back to the index`s inception in 1984. The new field file dji.a provides daily Dow Jones Industrial membership since January 1, 1974 through the present. Historical membership lists were obtained from http://online.wsj.com/public/resources/documents/DowMemberHi... . I have not bothered with automating daily updates to DJI membership because they are extremely rare, and another web query would be one more thing that could go wrong in my daily update routine. If you see that I`ve missed a change in membership (no password is required for current screening based on these field files, or any field files with extension ".a"), let me know and I`ll manually apply it. (Also let me know if you`re aware of any errors in the sources I have cited; it`s just about guaranteed there are some.) Since there have been questions about it lately, here is a GTR1 backtest for Dogs of the Dow from 1974 with an annual hold: http://gtr1.backtest.org/2012/?h252::dji.a:et1:Yield:gt0:Yie... Results (averaged over 252 cycles) are nothing to bark about. Fama/French Industry Classifications I recently created a new set of the following field files: File Name Type Description Default Value Start Date End Date Last Modified fincd5.a 115 [Fama/French 5 Industry Code] 0 19740104 99991231 20140107 fincd10.a 115 [Fama/French 10 Industry Code] 0 19740104 99991231 20140107 fincd12.a 115 [Fama/French 12 Industry Code] 0 19740104 99991231 20140107 fincd17.a 115 [Fama/French 17 Industry Code] 0 19740104 99991231 20140107 fincd30.a 115 [Fama/French 30 Industry Code] 0 19740104 99991231 20140107 fincd38.a 115 [Fama/French 38 Industry Code] 0 19740104 99991231 20140107 fincd48.a 115 [Fama/French 48 Industry Code] 0 19740104 99991231 20140107 fincd49.a 115 [Fama/French 49 Industry Code] 0 19740104 99991231 20140107 Each field file corresponds to one of Fama/French`s eight industry classifications, whose definitions (in terms of SIC code) can be downloaded from http://mba.tuck.dartmouth.edu/pages/faculty/ken.french/ftp/I... . The number in the field file name indicates how many industry groups stocks are divided into, and industry codes range from 1 through that number. Thus, for example, fincd10.a divides stocks into 10 industry groups (usually called "sectors" at that resolution), with proper codes ranging from 1 through 10 (and zero indicating that no classification exists for a stock). The previously existing field file fincd.a is identical to the new fincd49.a but will continue to exist at gtr1.backtest.org/2012 so that old URLs keep working. These field files are built by mapping stocks to their Fama/French industry groups via their SIC codes stored in the field file sic.a. But it`s important to realize that there is nothing at all standard about the assignment of Standard Industrial Classification (SIC) codes. If you ask five different data vendors what a company`s SIC code should be, you`ll typically get two or three different answers. The SEC requires publicly traded companies to provide an SIC code with their filings, but these codes are never reviewed and companies rarely ever change them after their initial filings, even if their businesses undergo significant transformation. Thus most data vendors assign SIC codes themselves and come up with their own industry classification systems. The SIC codes found in sic.a are assigned by my primary daily data source (which in turn gets most of them from Interactive Data Corp) through the most recent year`s end. I extend them forward through the next year myself with statistical analysis that uses industry classifications by other sources such as Yahoo!, Marketsmith, SI Pro, etc to guess what codes my primary source will assign in its next annual update. Thus, there is always quite a bit of arbitrariness in any fields derived from SIC codes, and any industry classifications in general. A screen that uses industry classification should show good backtest results regardless of whose industry classification is used, be it Value Line`s, SI Pro`s, Marketsmith`s etc. Zacks Rank In http://boards.fool.com/Message.asp?mid=30418571 I introduced the field file rank.z containing the historical weekly Zacks ranks extracted from Research Wizard, Zacks` screener and "backtester" (I`m not just being a backtester snob, by the way--their backtester, riddled with all kinds of crystal ball effects, really is garbage). As described in that post, establishing reliable links between the historical Zacks ranks and stocks in the GTR1 database was a huge challenge, with survivorship bias being the certain consequence of error. The field file rank.z at gtr1.backtest.org/2012 now contains the daily archives of Zacks ranks that I have been keeping since then. Updates to rank.z are part of my daily GTR1 database update routine, but the updates for this file will not always happen as soon as the rest of the database is updated. For example, on Fridays I typically update the basic GTR1 data and Value Line data that night and SI Pro the next day (Saturday). Zacks rank, however, is not reliably updated in Research Wizard until Sunday night, so I delay updates until then (though the ranks I obtain Sunday night are treated as "Friday" ranks with zero lag). You can find the date through which Zacks ranks have been updated by running the backtester with the command line argument list:rank.z, which shows a Starting Date, an Ending Date and a Last Modified date, with Ending Date indicating how recent ranks are. But this is only of interest to those with screening access, which requires a password. If you do have permission for Zacks screening, then the date as of when Zacks ranks (and all other displayed data) are current is also indicated in red font just above the results. If you want screening permission for rank.z, we can discuss in private email about how that can be obtained. Also email me if you are a Zacks subscriber, because keeping this field up-to-date is going to be a challenge, and there are a couple issues that need to be researched and resolved. For example, someone with access to both Research Wizard and the premium web service has reported that there are discrepancies between the two sources of Zacks rank that are not merely differences in timing. I would much prefer to obtain updates through a web interface due to the ability to automate them more easily, but I need to know that the data is correct. Marketsmith 2011 appears to have been the year of tree-saving at IBD, as the "IBD 100" was reduced to the "IBD 50", and less famously, the print version of Daily Graphs was trimmed from covering 2,800 to 2,000 stocks. However, in April of that year they rolled out a new service called Marketsmith to supersede Daily Graphs Online. Marketsmith includes about 150 technical and fundamental data fields from the William O`Neil database, including all those available through Daily Graphs Online. Marketsmith covers the entire GTR1 universe (all NYSE/AMEX/Nasdaq stocks, excluding preferred stock, ETNs, ETDs, structured products, warrants and rights), plus a smattering of ETNs and ETDs (Daily Graphs Online was expanded to cover the same stocks as well). The data is updated continuously, but only available through its (hideous) web-based Microsoft Silverlight platform. The highly customizable charts available through Marketsmith are apparently excellent, for those who are interested in them. Marketsmith is of interest to myself mainly as a laboratory for investigating, tweaking and possibly extending (beyond its possible discontinuation at any time) IBD`s WER. The official WER list is only published on Thursday nights, but Marketsmith users can produce WER lists any day of the week by following the right screening steps, with results closely matching the official WER on Thursday nights. IBD appears to have an additional screening step that has never been stated publicly, and I`m hoping that the GTR1 backtester with both Marketsmith and WER data will lead to its discovery. More importantly, I`d like to use Marketsmith data to reverse-engineer IBD`s RS and EPS rating systems, allowing us to backtest WER-like strategies back to 1987 using SI Pro and CPiT data. Marketsmith data is contained in the field files listed below (all with file extension ".k", `k` being a relatively rare letter I might not get to use again). Marketsmith fields that are easily calculated with GTR1 field functions (mainly total returns) do not have field files. All field files are included in my nightly GTR1 database update routine. "Friday" updates, however, do not take place until Sunday night in order to capture as much of the weekend data updates as possible. Sunday night data is treated as Friday close data (or Thursday close data on Easter weekend) with zero lag, so as always, accepting the default field file lag of one market day is recommended for realistic backtesting. A screening password for Marketsmith field files is available to the group of mechanical investors who support its archiving. File Name Type Description Default Value Start Date End Date Last Modified a.k 105 [MS Alpha] -999999 20110407 20140110 20140110 acceps3q.k 105 [MS EPS Accel 3 Qtrs] -999999 20110407 20140110 20140110 accmarg.k 105 [MS AT Margin Accel] -999999 20110407 20140110 20140110 accsls2q.k 105 [MS Sales Accel 2 Qtrs] -999999 20110407 20140110 20140110 accsls3q.k 105 [MS Sales Accel 3 Qtrs] -999999 20110407 20140110 20140110 ad.k 105 [MS A/D Rating] 7 20110407 20140110 20140110 adpw.k 105 [MS A/D Rating - Pr Wk] 7 20110407 20140110 20140110 adr.k 100 [MS ADR] 0 20110407 20140110 20140110 adv50.k 105 [MS 50-Day Avg $ Vol (1000s)] -999999 20130328 20140110 20140110 atmarg.k 105 [MS AT Margin] -999999 20110407 20140110 20140110 atmarg2qa.k 105 [MS Avg AT Margin 2Q] -999999 20110407 20140110 20140110 atmarg3qa.k 105 [MS Avg AT Margin 3Q] -999999 20110407 20140110 20140110 atmarg4qa.k 105 [MS Avg AT Margin 4Q] -999999 20110407 20140110 20140110 atmarg5qa.k 105 [MS Avg AT Margin 5Q] -999999 20110407 20140110 20140110 atmarg6qa.k 105 [MS Avg AT Margin 6Q] -999999 20110407 20140110 20140110 atr.k 105 [MS Avg True Range] -999999 20130328 20140110 20140110 av50.k 105 [MS 50-Day Avg Vol (1000s)] -999999 20110407 20140110 20140110 b.k 105 [MS Beta] -999999 20110407 20140110 20140110 cfepsq1.k 105 [MS CF vs EPS % Last Qtr] -999999 20110407 20140110 20140110 cfepsy1.k 105 [MS CF vs EPS % Last Yr] -999999 20110407 20140110 20140110 cmp.k 100 [MS Comp Rating] 0 20110407 20140110 20140110 cr.k 105 [MS Current Ratio] -999999 20110407 20140110 20140110 cso.k 105 [MS Shares (1000s)] -999999 20110407 20140110 20140110 dgpg.k 102 [MS DG Book Page] 0 20110407 20140110 20140110 divxamt.k 105 [MS Expected X Dividend Amount] -999999 20110407 20140110 20140110 dtcsm1.k 105 [MS Days Vol Short Current] -999999 20110407 20140110 20140110 dtcsm2.k 105 [MS Days Vol Short 1 Period Ago] -999999 20110407 20140110 20140110 dtcsm3.k 105 [MS Days Vol Short 2 Periods Ago] -999999 20110407 20140110 20140110 entval.k 105 [MS Enterprise Val (mil)] -999999 20110407 20140110 20140110 eps12m.k 105 [MS EPS Trailing 4 Qtrs] -999999 20110407 20140110 20140110 eps12mgt4ya.k 105 [MS EPS Trl 4Q > EPS 4 Yrs Ago] -999999 20110407 20140110 20140110 eps12mgty1.k 105 [MS EPS Trl 4Q >= EPS Lst Fiscal Yr] -999999 20110407 20140110 20140110 epsg1y.k 105 [MS EPS % Growth 1 Yr] -999999 20110407 20140110 20140110 epsg1ygt3y.k 105 [MS EPS % Chg Lst Q > 3-Yr Growth] -999999 20110407 20140110 20140110 epsg2qa.k 105 [MS Avg EPS % Chg 2Q] -999999 20110407 20140110 20140110 epsg3qa.k 105 [MS Avg EPS % Chg 3Q] -999999 20110407 20140110 20140110 epsg3y.k 105 [MS EPS % Growth 3 Yr] -999999 20110407 20140110 20140110 epsg3yal5y.k 105 [MS 3-Yr EPS Growth >= 5-Yr EPS Growth] -999999 20110407 20140110 20140110 epsg4qa.k 105 [MS Avg EPS % Chg 4Q] -999999 20110407 20140110 20140110 epsg5qa.k 105 [MS Avg EPS % Chg 5Q] -999999 20110407 20140110 20140110 epsg5y.k 105 [MS EPS % Growth 5 Yr] -999999 20110407 20140110 20140110 epsg6qa.k 105 [MS Avg EPS % Chg 6Q] -999999 20110407 20140110 20140110 epsgq4q0.k 105 [MS EPS Est Cur Qtr %] -999999 20110407 20140110 20140110 epsgq5q1.k 105 [MS EPS % Chg Last Qtr] -999999 20110407 20140110 20140110 epsgq5q1e.k 105 [MS EPS % Chg Last Qtr (-/+)] -999999 20110407 20140110 20140110 epsgq6q2.k 105 [MS EPS % Chg 1 Q Ago] -999999 20110407 20140110 20140110 epsgq6q2e.k 105 [MS EPS % Chg 1 Q Ago (-/+)] -999999 20110407 20140110 20140110 epsgq7q3.k 105 [MS EPS % Chg 2 Q Ago] -999999 20110407 20140110 20140110 epsgq7q3e.k 105 [MS EPS % Chg 2 Q Ago (-/+)] -999999 20110407 20140110 20140110 epsgq8q4.k 105 [MS EPS % Chg 3 Q Ago] -999999 20110407 20140110 20140110 epsgq8q4e.k 105 [MS EPS % Chg 3 Q Ago (-/+)] -999999 20110407 20140110 20140110 epsgy0ynxt.k 105 [MS EPS Est Next Yr %] -999999 20110407 20140110 20140110 epsgy1y0.k 105 [MS EPS Est Cur Yr %] -999999 20110407 20140110 20140110 epsgy2y1.k 105 [MS EPS % Chg Lst Yr] -999999 20110407 20140110 20140110 epsgy3y2.k 105 [MS EPS % Chg 1 Yr Ago] -999999 20110407 20140110 20140110 epsq0ord.k 102 [MS EPS Due Date] -999999 20130328 20140110 20140110 epsr.k 100 [MS EPS Rating] 0 20110407 20140110 20140110 epsstab.k 100 [MS Earnings Stability] 0 20130328 20140110 20140110 epsvest.k 105 [MS EPS Vs Est Last Q %] -999999 20110407 20140110 20140110 epsy1.k 105 [MS Fiscal EPS Lst Yr] -999999 20110407 20140110 20140110 epsy1gt4ya.k 105 [MS EPS Lst Yr > EPS 4 Yrs Ago] -999999 20110407 20140110 20140110 epsy2.k 105 [MS Fiscal EPS 1 Yr Ago] -999999 20110407 20140110 20140110 epsy3.k 105 [MS Fiscal EPS 2 Yrs Ago] -999999 20110407 20140110 20140110 epsy4.k 105 [MS Fiscal EPS 3 Yrs Ago] -999999 20110407 20140110 20140110 epsy5.k 105 [MS Fiscal EPS 4 Yrs Ago] -999999 20110407 20140110 20140110 epsy6.k 105 [MS Fiscal EPS 5 Yrs Ago] -999999 20110407 20140110 20140110 epsy7.k 105 [MS Fiscal EPS 6 Yrs Ago] -999999 20110407 20140110 20140110 etf.k 100 [MS ETF] 0 20130328 20140110 20140110 etfcef.k 100 [MS ETF/Closed-End Fund] 0 20110407 20140110 20140110 evfcf.k 105 [MS EV to FCF] -999999 20110407 20140110 20140110 exdivord.k 102 [MS Ex-Dividend Date] -999999 20110407 20140110 20140110 grs.k 105 [MS Ind Group RS] 7 20110407 20140110 20140110 gsust.k 105 [MS Sustainable Growth %] -999999 20110407 20140110 20140110 h52.k 105 [MS 52-Wk High] -999999 20110407 20140110 20140110 ibdarticle.k 100 [MS IBD Article 2 Yrs] 0 20110407 20140110 20140110 ihch.k 105 [MS Funds % Increase] -999999 20110407 20140110 20140110 ihltmed.k 105 [MS Inst Ownership < Median] -999999 20110407 20140110 20140110 indcount.k 105 [MS Number of Stocks] -999999 20130328 20140110 20140110 indmcp.k 105 [MS Ind Mkt Val (bil)] -999999 20110407 20140110 20140110 indr.k 105 [MS Ind Group Rank] -999999 20110407 20140110 20140110 indr3m.k 105 [MS Ind Grp Rnk 3 Mo Ago] -999999 20110407 20140110 20140110 indr6m.k 105 [MS Ind Grp Rnk 6 Mo Ago] -999999 20110407 20140110 20140110 indtr3m.k 105 [MS Ind Grp % Chg 3 Mo] -999999 20110407 20140110 20140110 l52.k 105 [MS 52-Wk Low] -999999 20110407 20140110 20140110 lmord.k 102 [MS Last Mark-up Date] -999999 20130328 20140110 20140110 ltdeq.k 105 [MS Debt %] -999999 20110407 20140110 20140110 ltdwcap.k 105 [MS LT Debt to Working Cap] -999999 20110407 20140110 20140110 ltoaltind.k 105 [MS Liab to Assets < Ind Median] -999999 20110407 20140110 20140110 mcp.k 105 [MS Market Cap (mil)] -999999 20110407 20140110 20140110 nih.k 105 [MS Number of Funds] -999999 20110407 20140110 20140110 nnhind.k 105 [MS # New Highs in Group] -999999 20110407 20140110 20140110 nnlind.k 105 [MS # New Lows in Group] -999999 20110407 20140110 20140110 nwceo12m.k 105 [MS New CEO 12 Months] -999999 20110407 20140110 20140110 omalind.k 105 [MS Op Marg >= Ind Median] -999999 20110407 20140110 20140110 opt.k 100 [MS Options] 0 20110407 20140110 20140110 pbv.k 105 [MS Price to Book] -999999 20110407 20140110 20140110 pcf.k 105 [MS Price to CF] -999999 20110407 20140110 20140110 pe.k 105 [MS P/E] -999999 20110407 20140110 20140110 peest.k 105 [MS Est P/E] -999999 20110407 20140110 20140110 peg.k 105 [MS PEG] -999999 20110407 20140110 20140110 pegdvadj.k 105 [MS Dividend-Adjusted PEG] -999999 20110407 20140110 20140110 pelt5ya.k 105 [MS P/E < 5-Yr Avg] -999999 20110407 20140110 20140110 peps12mgq21q1.k 105 [MS EPS % Growth 5 Yr Pct Rnk] -999999 20110407 20140110 20140110 pevsp.k 105 [MS P/E vs S&P 500 P/E (%)] -999999 20110407 20140110 20140110 pih.k 105 [MS Funds %] -999999 20110407 20140110 20140110 pmh.k 105 [MS Mgmt %] -999999 20110407 20140110 20140110 pnhind.k 105 [MS % New Highs in Group] -999999 20110407 20140110 20140110 pnlind.k 105 [MS % New Lows in Group] -999999 20110407 20140110 20140110 poh.k 105 [MS % Off High] -999999 20110407 20140110 20140110 ppe.k 105 [MS P/E Percent Rank] -999999 20110407 20140110 20140110 ppeind.k 105 [MS P/E Ratio Rank in Grp] -999999 20110407 20140110 20140110 profalind.k 105 [MS Prof Marg >= Ind Median] -999999 20110407 20140110 20140110 ps.k 105 [MS Price to Sales] -999999 20110407 20140110 20140110 ptmarg.k 105 [MS Pre-tax Margins] -999999 20110407 20140110 20140110 roe.k 105 [MS ROE] -999999 20110407 20140110 20140110 roeav5y.k 105 [MS ROE 5-Yr Avg] -999999 20110407 20140110 20140110 rslnh.k 105 [MS RS Line New High] -999999 20110407 20140110 20140110 rslnl.k 105 [MS RS Line New Low] -999999 20130328 20140110 20140110 rsr.k 100 [MS RS Rating] 0 20110407 20140110 20140110 rsr3m.k 100 [MS RS 3-Month Rating] 0 20110407 20140110 20140110 rsr6m.k 100 [MS RS 6-Month Rating] 0 20110407 20140110 20140110 sales.k 105 [MS Annual Sales (mil)] -999999 20110407 20140110 20140110 short.k 105 [MS Short Volume] -999999 20110407 20140110 20140110 sich.k 105 [MS Shrt Int % Chg] -999999 20110407 20140110 20140110 sipfloat.k 105 [MS Shrt Int % of Float] -999999 20110407 20140110 20140110 slsg2qa.k 105 [MS Avg Sales % Chg 2Q] -999999 20110407 20140110 20140110 slsg3qa.k 105 [MS Avg Sales % Chg 3Q] -999999 20110407 20140110 20140110 slsg3y.k 105 [MS Sales Growth 3 Yr] -999999 20110407 20140110 20140110 slsg4qa.k 105 [MS Avg Sales % Chg 4Q] -999999 20110407 20140110 20140110 slsg4y.k 105 [MS Sales Growth 5 Yr] -999999 20110407 20140110 20140110 slsg5qa.k 105 [MS Avg Sales % Chg 5Q] -999999 20110407 20140110 20140110 slsg6qa.k 105 [MS Avg Sales % Chg 6Q] -999999 20110407 20140110 20140110 slsgq5q1.k 105 [MS Sales % Chg Lst Qtr] -999999 20110407 20140110 20140110 slsgy2y1.k 105 [MS Sales % Chg Lst Yr] -999999 20110407 20140110 20140110 smr.k 100 [MS SMR Rating] 7 20110407 20140110 20140110 spr.k 100 [MS Sponsor Rating] 0 20110407 20140110 20140110 tim.k 100 [MS Timeliness Rating] 7 20110407 20140110 20140110 timpw.k 100 [MS Timeliness Rating - Pr Wk] 7 20110407 20140110 20140110 tr126vsp.k 105 [MS Trl 26 Wk % Perf vs S&P 500] -999999 20110407 20140110 20140110 udv.k 105 [MS Up/Down Vol] -999999 20110407 20140110 20140110 vch50.k 105 [MS Vol % Chg vs 50-Day] -999999 20110407 20140110 20140110 vol.k 105 [MS Volume (1000s)] -999999 20110407 20140110 20140110 yield.k 105 [MS Yield %] -999999 20110407 20140110 20140110 Daily Graphs Daily Graphs data, which has not been in the GTR1 database since the 2009 version ( gtr1.backtest.org/2009 ), is now back in the 2012 database at gtr1.backtest.org/2012 . It has been extended forward from 20110407 to the present as a subset of the Marketsmith data, with coverage intended to match that of the traditional print version of Daily Graphs. Both Daily Graphs Online and Marketsmith have a field with Daily Graphs Book Page (contained in the Marketsmith field file dgpg.k), which is supposed to indicate whether a stock is covered in the print version, but that field was corrupt in both Marketsmith and Daily Graphs Online through the end of 2011, when IBD finally corrected it. I therefore had to reconstruct Daily Graphs lists for 2011 by other means. Daily Graphs book membership is correctly indicated from 1999 through the present in the field file dg.d, with 1 indicating membership and 0 indicating non-membership. Discrepancies between dgpg.k and dg.d in 2011 should be expected, and in all cases, dg.d should be assumed correct. As already mentioned, the traditional Daily Graphs universe (in print) was cut from 2,800 stocks to 2,000 stocks in April, 2011. Clearly, IBD does not have our backtesting in mind ;-) (especially when it comes to their halving of the IBD 100). While this change probably doesn`t render backtests anywhere near close to worthless (as the changes to the IBD 100 might, for those who have the historical data), this raises some doubts about the validity of backtests of Daily Graphs screens. In any case, here is a list of the Daily Graphs field files (note that some file names have changed since the 2009 version of the database): File Name Type Description Default Value Start Date End Date Last Modified a.d 115 [DG Alpha] -999999 19990106 20140110 20140110 ad.d 115 [DG A/D Rating] 7 19981231 20140110 20140110 b.d 115 [DG Beta] -999999 19990106 20140110 20140110 cmp.d 115 [DG Comp Rating] 0 20020815 20140110 20140110 dg.d 115 [DG Inclusion] 0 19981231 20140110 20140110 epsgq5q1.d 115 [DG EPS % Chg Last Qtr] -999999 20000210 20140110 20140110 epsr.d 115 [DG EPS Rating] 0 19981231 20140110 20140110 grs.d 115 [DG Ind Group RS] 7 19981231 20140110 20140110 pe.d 115 [DG P/E] -999999 19990104 20140110 20140110 rsr.d 115 [DG RS Rating] 0 19981231 20140110 20140110 smr.d 115 [DG SMR Rating] 7 19981231 20140110 20140110 tim.d 115 [DG Timeliness Rating] 7 19990104 20140110 20140110 vch50.d 115 [DG Vol % Chg vs 50-Day] -999999 19990830 20140110 20140110 Robbie Geary