Post New

Any hope for the return of GTR1?
sigh...

A real tragedy if not.

I barely know how to work it, compared to others, but its still the most valuable investing tool I've ever found. That such a thing should have been available for free all these years, developed and revised as the centerpiece of a serious interactive community -- it is an all-too-rare example of the original Web dream coming true.

Baltassar

I can tell you the latest I've heard was that Robbie's drive had damaged ROM and he was trying to find
a business with expertise in recovering data when that situation exists. That update was from the end of last week.
I'm hoping to hear something more sometime soon.

Let Robbie know we're willing to help and track down resources. I emailed him directly mentioning that I know a small company that builds/repairs computers.
Didn't hear back, and I'm not sure really what he needs, and whether he is looking for resources in Australia.

Mark

At that last update he was considering sending his drive to a company in the US.

I don't know where that stands at this point.

NEW Update

"Well, it turns out that Recovery Squad damaged the platters of my drive beyond any possibility of repair and were probably lying to me every step of the way.

It will be a long (probably multi-month) programming project to get the GTR1 database updating again."

:(

What a shame. So sorry to hear that. : (

GTR1 is amazing and a joy to use. I greatly appreciate the many years it was available, and earnestly hope it returns. Can I do anything to help? I haven't found anything close. Maybe I'll sign up for P123 and try to duplicate some GTR1 results. (First step would be ETFs.)

Finding quality fundamental data sources would help. Has anyone looked at what's available lately? finbox uses Standard & Poors Market Intelligence data. CRSP data is used by many academics (e.g. Fama French data library). There are other data sources (e.g. refinitiv, factset). Maybe there's enough of us to afford better data.

There are also several quant websites: quickfs.net quantrocket.com businessquant.com ...

I would be interested in having a conversation, maybe offline, about the opportunities here and possible ways to divide up the work.

Last time I checked the list price for a full featured P123 subscription for new members was $2400/year. And unless your grandfathered in some of the oldest data isn't available. Very powerful, you can make different buy and sell criteria, write script to automate testing, active blog among some very capable investors and much easier to use than GTR1. But in my opinion GTR1 although more difficult to master when you add up all the advantages and disadvantages in a comparison GTR1 is overall on a par with the top level P123.
It is in the best interest of what's left of the MI community to help or support Robbie in any way we can.
If indeed he wants any help.

"Well, it turns out that Recovery Squad damaged the platters of my drive beyond any possibility of repair and were probably lying to me every step of the way.

It will be a long (probably multi-month) programming project to get the GTR1 database updating again."


That's very disheartening.

I've been using the GTR1 ranking feature for two screens that I use, and now I'm adrift. The two screens are variants of Mungofitch's ROE_Cash screen. The screens turn over very slowly, so there wasn't much of a rush. I could put off the screen update for a month or two, but with a longer wait it becomes a problem. I've been producing the rankings with GTR1 using SIPro data. I wonder if anyone else is using the same screens on GTR1, or perhaps producing the same rankings through a different mechanism, such as directly within the SIPro database, or using VL data. If you have, I'd appreciate you sharing your method.

Elan

I've been using the GTR1 ranking feature for two screens that I use, and now I'm adrift. . . variants of Mungofitch's ROE_Cash screen

I'm in the same boat. The only screen I've been using on GTR1 was my own version of ROE_Cash.
Borisnand posted a version of ROE_cash using only sip data on 10/25/2020. I tried to make a version
that would run on my SIP data only backtester. My first pass:
This version uses SP500 or SP400 as a quick filter of crap instead of top 1700 Mcap

(ci.SP = 500 OR ci.SP = 400) \
and CI.IND_2_DIG <> 55 \
and MLT.PFCPS >0 \
and CASH_Q1 > 0 \
and BSQ.Equity_Q1 > 0 \
and NETINC_12M >0 \
and fcfps_12M > 0 \
and 100*ISA.NETINC_12M/BSQ.EQUITY_Q1 >18 \
ORDER BY ((CASH_Q1 - LTDEBT_Q1)/MKTCAP) Desc

Doesn't backtest as well as the GTR1 version:
ROE_CASH_ 2004-10-01 thru 2023-03-31
Top 15 tilldrop 20 period 21 days slipRndTrp 0.70% , Dividends True , CashForNoPicks True
CAGR 14.0% GSD 28.5% CAGR/GSD 0.5 UlcrIndx 10.8% MxDD 54.7% AT 200%
11.37 X accumulated with 555 Buys, 4815 Holds and 100.0% Avg Invested

This can be screened using the SIP built in screener.
Boris also suggested creating a custom field in SIP which I didn't do which would probably improve the results.
Custom SIPro field for CashMinusDebt:

IIF(IsFieldNull([Cash Q1])=0,0,[Cash Q1]) + IIF(IsFieldNull([Short-term investments Q1])=0,0,[Short-term investments Q1]) - IIF(IsFieldNull([Long-term debt Q1])=0,0,[Long-term debt Q1]) - IIF(IsFieldNull([Short-term debt Q1])=0,0,[Short-term debt Q1])

RAM

Elann and RAMc,

If you can give me a Radiscript definition for that screen, I should be able to get the picks by using either Radiscreen or the program I wrote called Screen Picks. I have made a few attempts to make a translator from GTR1 "speak" to Radiscript but each time I end up getting bogged down.

Larry

Elann and RAMc, If you can give me a Radiscript definition for that screen, I should be able to get the picks by using either Radiscreen or the program I wrote called Screen Picks

Here is boris's version he posted on 7/13/20 except using SP500 & 600. Reviewing boris's post I just noticed that from way on back I had played with several versions of the final sort. For a while the preferred sort was (cash-debt)/market_cap which is what I posted. But boris's version dropped the market cap devisor.
I should have looked at the version I have been using because I wasn't dividing by market cap.
The good news, I just backtested the screen w/o a market cap divisor. Much Improved:

ROE_CASH 2013-10-01 thru 2023-04-14 Top 15 tilldrop 20 period 21 days slipRndTrp 0.70% ,
Dividends True , CashForNoPicks True
2013-10-01 start CAGR Fri 20.6% UlcrIndx 8.1%
2013-10-08 start CAGR Fri 18.8% UlcrIndx 7.6%
2013-10-15 start CAGR Fri 18.2% UlcrIndx 7.6%
2013-10-22 start CAGR Fri 20.3% UlcrIndx 8.1%
GSD 22.9% CAGR/GSD 0.9 MxDD 28.0% AT 141%
5.96 X accumulated with 201 Buys, 2475 Holds and 100.0% Avg Invested
Delayed trading till Monday using an average of O,(H/L)/2,C gave an 1.0003 within the noise advantage.

Define {ROE_Cash}
Uses [SI Equity (common) Q1] [SI Net income 12m] [SI Price/FCFPS] [SI Sector] [SI Long-term debt Q1] [SI Short-term debt Q1] [SI Cash Q1] [SI Short-Term Investments Q1]
Deblank [SI Equity (common) Q1] [SI Net income 12m] [SI Price/FCFPS] [SI Sector]
Keep :OR ([SI Standard and Poor stock] = "500",[SI Standard and Poor stock] = "400")
Keep :[SI Sector]<>"07-Financial"
Keep :[SI Sector]<>"55 - Financials"
Keep :[SI Price/FCFPS]>0
Keep :[SI CASH Q1] > 0
keep :[SI Equity (common) Q1] > 0
Create [ROE] :100*[SI Net income 12m]/[SI Equity (common) Q1]
Keep :[ROE]>18
Create [CashMinusDebt] :[SI Cash Q1]+[SI Short-Term Investments Q1]-[SI Short-term debt Q1]-[SI Long-term debt Q1]
Sort Descending [CashMinusDebt]
End

Friday's Picks GOOGL MSFT META TSLA CSCO CFR VRTX MRNA COST ACN ZION MOH REGN CMA ANET

RAMc

RAMc,

Thank you so much for posting this listing. I've been trying to work with SI screener without much success.

A couple questions. Why did you limit your screening to just the SP 500 & 600? What about mid caps? While true that this screen mostly picks large cap stocks, it has pick and held in the past a couple solid mid cap stocks for several month and nice gains. I would assume that 95+% of your picks are going to come from the SP500, very little added value from the 600 small cap index.

The original screen definition was top 1700 stocks by market cap. When I was making my feeble attempts at building this screen I was using the SP 1500 minus financials which was leaving me just over 1300 stocks. Maybe a larger pool of more likely mid cap stocks that will pass the filters and be lower on everyone radar? I like the original idea of top 1700 market cap because there are many great stocks that don't make the SP 1500 list because the index already has "too many" stocks from x sector. It also allows a more fluid movement of stocks as their price increases or decreases instead when the SP decides to update their index once or twice a year.

Also any way to expand the screener to the top 30-40 so the original idea of HTD can be implemented? I believe the original was buy top 20 HTD till 40?

Great work, appreciate it and gives me another stepping stone in the right direction towards my sorry efforts to learn SI screener.

Peace,
Opihi

Also forgot to note when I was playing with the screener, I was using a ROE> 19 or 20 as it would filter the list closer to the original screen definition of ROE top 500. ROE>19 would pass like 480 stocks, >20 would pass around 520.

Peace,
Opihi

There is hope for the GTR1 backtester updating again, but it's going to take some time, probably months.

One thing that would speed things up is if anyone knows any professors or graduate students in university finance departments. TMF user MarginOfBuffett helped with me this in the past, but I don't have a current email address for him.

Why did you limit your screening to just the SP 500 & 600?

Note that the 600 is probably a typo. In the actual Radiscript it says 400, so it's large and mid cap.

Elan

I can make contact with a guy here at the U of M business school. Don't know him, but he was recommended to me by a friend. Name Kai Petainen.
https://www.forbes.com/sites/laurengensler/2014/12...
Let me know if you want me to make an introduction.

lol sorry meant to DM that, oh well.

RAM,

Using the following Radiscript for ROE_Cash:

Define {ROE_Cash}
Uses [SI Equity (common) Q1] [SI Net income 12m] [SI Price/FCFPS] [SI Sector] [SI Long-term debt Q1] [SI Short-term debt Q1] [SI Cash Q1] [SI Short-Term Investments Q1]
Deblank [SI Equity (common) Q1] [SI Net income 12m] [SI Price/FCFPS] [SI Sector]
Keep :OR([SI Standard and Poor stock]="500",[SI Standard and Poor stock]="400")
Keep :[SI Sector]<>"07-Financial"
Keep :[SI Sector]<>"55 - Financials"
Keep :[SI Price/FCFPS]>0
Keep :[SI CASH Q1] > 0
keep :[SI Equity (common) Q1] > 0
Create [ROE] :100*[SI Net income 12m]/[SI Equity (common) Q1]
Keep :[ROE]>18
Create [CashMinusDebt] :[SI Cash Q1]+[SI Short-Term Investments Q1]-[SI Short-term debt Q1]-[SI Long-term debt Q1]
Sort Descending [CashMinusDebt]
End

I got the following picks (last Friday's data) using my program called Screen Picks:

Screen	1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25	
ROE_Cash	GOOGL	MSFT	META	TSLA	LMT	NXPI	CSCO	VRTX	HAL	MCO	KLAC	MRNA	COST	MCHP	ACN	CF	MOH	REGN	DLTR	ANET	NVDA	TJX	ROST	EXPD	ADBE

Similarities but lots of differences.

I will include ROE_Cash when I do the SI Pro Rankings tomorrow morning so we can compare.

Larry

Using the following Radiscript for ROE_Cash:

Awesome. The versions I use for ROE_Cash are a bit different.
Instead of screening for S&P500 (Lg cap) and S&P400 (Mid cap), I screen for Market cap top 1700.
Instead of ROE>18, I screen for ROE top 500.

The second variant which I call ROE_Cash2 is very similar with different cutoffs.
Market cap top 1600.
ROE top 60.

Is it possible to add those to your SI Pro Rankings?

Thanks,
Elan

or perhaps producing the same rankings through a different mechanism

I am not using anything else, but filling the time by looking at some alternative data sources. Quandl (now known as Nasdaq Data Link) was mentioned a few times in the old forum.

Re: A SPY alternative screen, 06/01/2020
"Yahoo gives about the same values for ROE, Cash, and LT debt as Tiingo. There are other data providers. Which I didn't bother to cross-check. Quandl and SimFin are two others."
http://www.datahelper.com/mi/search.phtml?nofool=y...

Sharadar US fundamental data (including adjusted prices) is about $800/year for personal use.
https://blog.data.nasdaq.com/the-comprehensive-gui...

Elan - I screen for Market cap top 1700.
Instead of ROE>18, I screen for ROE top 500.

I haven't used radiscreen for more than 10 years but this might be what you want. At least it runs in Runradis.
Define {ROE_Cash}
Uses [SI Equity (common) Q1] [SI Net income 12m] [SI Price/FCFPS] [SI Sector] [SI Long-term debt Q1] [SI Short-term debt Q1] [SI Cash Q1] [SI Short-Term Investments Q1]
Deblank [SI Equity (common) Q1] [SI Net income 12m] [SI Price/FCFPS] [SI Sector]
Sort Descending [SI Market Cap Q1]
Top :1700 plusTies
Keep :[SI Sector]<>"07-Financial"
Keep :[SI Sector]<>"55 - Financials"
Keep :[SI Price/FCFPS]>0
Keep :[SI CASH Q1] > 0
keep :[SI Equity (common) Q1] > 0
Create [ROE] :100*[SI Net income 12m]/[SI Equity (common) Q1]
Sort Descending [ROE]
Top :500 PlusTies
Create [CashMinusDebt] :[SI Cash Q1]+[SI Short-Term Investments Q1]-[SI Short-term debt Q1]-[SI Long-term debt Q1]
Sort Descending [CashMinusDebt]
End

Last weeks picks (I haven't loaded this weeks yet)
GOOGL MSFT TSM PDD TSLA BNTX EQNR CSCO VRTX MRNA ABNB COST ACN BPOP MOH

I can't run this version on my backtester so I would run a GTR1 backtest using only *.s factors to verify before having any confidence.
RAM

RAM,

I think you can delete that first Keep line: Keep :[SI Sector]<>"07-Financial". As I look at the data under {SI Sector] when sorted ascending (7123 rows last Friday) the list starts with "50 -Energy" and ends with "63 -Academic & Educational Services". There is nothing that looks like "07-Financial"

The rest of the Radiscript seems to work fine in my program called Screen Picks. I haveave temporarily called the screen ROE_Cash_elann. For last Friday I get the picks: GOOGL MSFT TSM PDD TSLA LMT BNTX NXPI EQNR CSCO ASML VRTX MCO KLAC MRNA ABNB COST MCHP ACN CF MOH ENLC FERG UMC M

For this Friday I get: GOOGL MSFT TSM PDD TSLA BNTX NXPI EQNR CSCO ASML VRTX MCO KLAC MRNA ABNB COST MCHP ACN CF MOH ENLC FERG UMC M JWN

Also, your last Friday list shows BPOP which according to my data is a "55 -Financials".

Larry

Larry,
I left the <> 07 financials in there to allow the screen to work in backtesting before 2018-10-26
When Stock Investor changed the designations for financial. Agreed it is not needed for current screening.

Date		ind2dig=7count		ind2dig=55count
10/19/2018	1250			0
10/26/2018	0			1481

RAM

your last Friday list shows BPOP which according to my data is a "55 -Financials".

Thanks for catching that!
The "55 - Financials" was typed in with 2 extra spaces before and after the - , should be "55-Financials"

Is there a way to donate money to the cause? Perhaps the community could at least cover the financial costs. I'm pretty sure I've contributed in the past, but it's been a few years and I can't find a record or remember how I did it. Maybe we could set up a Go Fund Me?

Define {ROE_Cash}
Uses [SI Equity (common) Q1] [SI Net income 12m] [SI Price/FCFPS] [SI Sector] [SI Long-term debt Q1] [SI Short-term debt Q1] [SI Cash Q1] [SI Short-Term Investments Q1]

Attempted to run a copy of the ROE_Cash screen on the Fidelity Screener again, but not quite the same without the specific criteria.

Fidelity Screeen ROE_Total D/E_GMI-Fin-Health

Security Type
Common Stock 7975

Company Headquarters Location
North America selected 5932

Market Capitalization
$1.88B and Above 1429

Sector/Industry/Sub-Industry
Energy, Materials, Industrials & 6 more selected 1187

Return on Equity (TTM)
18.46% and Above 441

Total Debt/Equity (TTM)
70.55% and Below 183

Free Cash Flow
0.00 and Above 170

Total Cash
All selected 170

Quick Ratio (Most Recent Qtr)
All selected 170

Current Ratio (Most Recent Qtr)
All selected 170

Debt to Capital Ratio (Most Recent Qtr)
All selected 170

S&P Global Market Intelligence Financial Health
61-100 99


GD_

Post New

Mechanical Investing FAQ

---

Contact Shrewd'm
Contact the developer of these message boards.