Skip to main content
Message Font: Serif | Sans-Serif
 
No. of Recommendations: 8
A nice performing screen that uses a Zacks proprietary rating.

In 2006, the screen returned 74% with a -4.5% drawdown

1) price times volume >500K
2) market cap > 100M
3) Zacks rank = buy or strong buy
4) (4 week return + 12 week return + 26 week return) /
(P/E times P/S) >20

The idea behind statement #4 is to get a high numerator
and a low demominator.

Looking back over 3 years, this screen had a CAGR
9 times better than the S&P.
Some months no stocks made the grade, but over a 3 year
period an average of 5 stocks were selected each month.

AVT
PNCL
are the current selections

(as usual with all backtest results and market screens,
past results may not be duplicated in the future)

Print the post Back To Top
No. of Recommendations: 3
Tpoto,

I ran several versions of this screen, and found the best one over 5 years to be adding a 5th line returning the top 3 only (duplicate line 4 but stipulate top 3 only).

In fact this screen beats my other screens tested over the same dates by a nice margin.

Great work, again!

Thanks,
Pat
Print the post Back To Top
No. of Recommendations: 3
Pat-

Thanks for doing the 5 year backtest for me.

I originally had the top 3 for line 4, but then
I found than I wanted it to be fairly strong so
I started with a minimum value of 20.
But value of 20 and then top 3 is a great addition,
as you suggested.

For the 3 year backtest, 4 week hold the CAGR comes in at over 100%
averaging the first 4 of the weekly starting periods.

I might want to try to follow that screen using
MSN rating of 10 (in place of Zacks) to see how a free
screener would do.

With this, I wonder if some of the screens that I use
should have some minimum before selecting the top x
like in this case.


Thanks.
Print the post Back To Top
No. of Recommendations: 2
tpoto, Pat,

Do you think that introducing some weighting coefficients for the returns to the formula in #4 might help? In other words, say, giving it a little more tilt towards recent performance.

For example;
(k4, k12, k26) as multipliers for the respective returns, where sum(k's)=3.

In fact, it would be interesting to test both, emphasizing earlier and recent performance, something like 3*(0.20, 0.35, 0.45) and 3*(0.45, 0.35, 0.20), resp..


Print the post Back To Top
No. of Recommendations: 1
I hope Tpoto will reply, because I have to admit I can't figure it out.

Print the post Back To Top
No. of Recommendations: 0
Do you think that introducing some weighting coefficients for the returns to the formula in #4 might help? In other words, say, giving it a little more tilt towards recent performance.

I think by definition it already is.

Because the 4-week return is part of both the 12-week and the 26-week return and the 12-week return is part of the 26-week return.
Print the post Back To Top
No. of Recommendations: 6
the best one over 5 years to be adding a 5th line returning the top 3 only (duplicate line 4 but stipulate top 3 only)

I have been playing with the new advanced Zacks backtester, which goes back to 1999. This screen as described in post 24 has AMAZING results.

I'm almost too embarrassed to post them, but here they are:

4 week hold
12/31/99 to 3/30/07
Total Return(average of 4 runs): 12,301%
CAGR: 94
MD: 32
CAGR/MD: 3.20

2 week hold
12/31/99 to 3/30/07
Total Return(average of 2 runs): 19,271%
CAGR: 110
MD: 35
CAGR/MD: 3.20

I tested this screen for the 2006-YTD period, as well as the worst
period of all, 3/17/2000 to 3/7/2003, and all results are astonishing.

The CAGR during the bear market was 73%, and the CAGR for the 06-07 period was 119%.

This is an amazing screen no matter how you test it.

Many thanks to Tpoto!
Print the post Back To Top
No. of Recommendations: 0
A nice performing screen that uses a Zacks proprietary rating.

In 2006, the screen returned 74% with a -4.5% drawdown

1) price times volume >500K
2) market cap > 100M
3) Zacks rank = buy or strong buy
4) (4 week return + 12 week return + 26 week return) /
(P/E times P/S) >20


Is the price times volume for the daily volume or average 50-day volume or some other volume figure?
Print the post Back To Top
No. of Recommendations: 1
mingusmonk-

Yes, that is quite a screen. Even better is the fact
that there was just one very ugly period in which
the picked stocks took a big drop. Putting a very
high (liberal) stop (mental or market) would have curtailed
the drop with a even higher CAGR.

I'm tempted to try to tweak it, but might be datamining
and besides why fool with success!

P.S. I see you have a 2 year anniversary today!
Congrats!
Print the post Back To Top
No. of Recommendations: 0
Is the price times volume for the daily volume or average 50-day volume or some other volume figure?

This is the average volume over the past 20 days.
Print the post Back To Top
No. of Recommendations: 0

1) price times volume >500K
2) market cap > 100M
3) Zacks rank = buy or strong buy
4) (4 week return + 12 week return + 26 week return) /
(P/E times P/S) >20


Have you considered backtesting the v4 screen except with the momentum component revised to be the same as in the screen above?

v4 from:
Market cap > 100 million
Price > 5
Price * volume > 750000
Zacks rank = 1
Average Broker Rating = 1 (Strong Buy)
24 wk pr chg/ P/S (top 3)

to:
Market cap > 100 million
Price > 5
Price * volume > 750000
Zacks rank = 1
Average Broker Rating = 1 (Strong Buy)
(4 week return + 12 week return + 26 week return) /
(P/E times P/S) top 3



Print the post Back To Top
No. of Recommendations: 0

1) price times volume >500K
2) market cap > 100M
3) Zacks rank = buy or strong buy
4) (4 week return + 12 week return + 26 week return) /
(P/E times P/S) >20


I eliminate steps 1) and 2) and trade the first stocks that are window tradable at foliofn. Steps 1) and 2) are not a direct correlation to window tradable stocks. I rarely find a stock rated Z<3 that is not window tradable.
Print the post Back To Top
No. of Recommendations: 0
Have you considered backtesting the v4 screen except with the momentum component revised to be the same as in the screen above?

Fred,

I tested your revision and compared it vs the original V4.

The CAGR steadily improved as I shortened the hold time. The 4 week runs were similar, the two week runs of your revision show a slight improvement in CAGR (from 59 to 62, with similar max drawdowns), while the 1 week run showed the greatest increase in CAGR, from 76 to 84.

Interestingly, the max drawdowns were similar across all runs.
Print the post Back To Top
No. of Recommendations: 0
Has anyone attempted this in value line/

Here is what I have:
http://backtest.org/0206SBT12XOpriMvoluCG500000XmcpG100XOtrw4Ar13Ar26CDOcpeMprsrCT3

I do not appear to get any results. Am I missing something obvious?

Print the post Back To Top
No. of Recommendations: 3
A nice performing screen that uses a Zacks proprietary rating.

In 2006, the screen returned 74% with a -4.5% drawdown

1) price times volume >500K
2) market cap > 100M
3) Zacks rank = buy or strong buy
4) (4 week return + 12 week return + 26 week return) /
(P/E times P/S) >20
=== === ===
What is the annual subscription fee at Zack's to run screens that use Zacks proprietary ratings ?
--BigBunk
Print the post Back To Top
No. of Recommendations: 2
This week's picks for this screen (at least in my version) were:

RUSHA
RYI
RS

I bought these via FolioFN in the 11 AM window.

RYI hada 7.3% gain today and RS had a 4.2% gain. That is great. Unfortunately the Folio window missed the entire gain and resulted in a loss for both purchases today.

Here is the question:
Has anyone run this screen at Folio long enough to say if this is a common occurrence?

Great screen - but is it O.K. for FolioFN?

Thanks for any info.

knigit
Print the post Back To Top
No. of Recommendations: 1
I bought these via FolioFN in the 11 AM window.

I know the feeling, but before blaming the window time note that the stocks were on the screen the later part of last week. An advantage with Zacks is that the stock picks are daily, not weekly.
Print the post Back To Top
No. of Recommendations: 2
I know the feeling, but before blaming the window time note that the stocks were on the screen the later part of last week. An advantage with Zacks is that the stock picks are daily, not weekly.

Fred04-

I appreciate the comment, and see where you are coming from. Are you rebalancing daily?

I am not trying to place blame on Folio, so to speak, as much as I am trying to determine the compatibility of this screen's picks with the Folio window trading method. In this case, the gain/timing suggests that today would have be a good day to use the "free" market orders I have - however, there is little (no) sense in instituting a global change based on a single day's data. I am seeking additional data points - even if they are somewhat anecdotal.

Unless one rebalances daily, I do not see how daily ranking can help this. Also, daily rebalancing would make the backtest more of a general guideline than anything else. Perhaps this is a superior approach, I cannot say.

The idea of trading on a different day (but still weekly or bi-weekly) with "fresh" data is a valid one. I have considered it more than once, knowing that many of the picks might "survive" through the weekend, but this is again moving away from the backtester and the great proven success of Market Screen #3.

Has anyone looked at non-Monday trading performance?

Thanks

knigit

Print the post Back To Top
No. of Recommendations: 2
I know the feeling, but before blaming the window time note that the stocks were on the screen the later part of last week. An advantage with Zacks is that the stock picks are daily, not weekly.
=== ===
What is the cost of an annual Zacks' subscription with the proprietary Zacks' ratings ??
--BigBunk
Print the post Back To Top
No. of Recommendations: 1
BigBunk-

You have asked that questions several times.
It all depends on the type of subscription you want.

Go to zacks.com and look up their contact information-
and e-mail them or call them.

Judging from your previous posting history, it appears
that you are not a serious investor and just might
be trying wasting our time.

Print the post Back To Top
No. of Recommendations: 0
Print the post Back To Top
No. of Recommendations: 0
Print the post Back To Top
No. of Recommendations: 2
Print the post Back To Top
No. of Recommendations: 0
I've been away from the boards for awhile. Tpoto, I've used several of your screens in the past, and I'm amazed at the number, quality, and ease of implementation of the screens you come up with.

I'm trying to replicate this screen in MSN using a SS rating of 9+ instead of a Zack's buy/strong buy.

1) price times volume >500K

Is this really supposed to be 500,000? It does not filter out any stocks when I run it (in conjunction with the other criteria), but maybe it is only intended to keep out the lowest liquidity stocks.

4) (4 week return + 12 week return + 26 week return)
(P/E times P/S) >20


Even using returns that are expressed as numbers (versus percentages), no stock exceeds a value of 9, so none are meeting the screen. I realize this is plausible, but just want to be sure I'm doing it right.

Perhaps if anybody is running this on MSN, or even Zack's, they could post current picks so I could compare.

Thanks,

Chris
Print the post Back To Top
No. of Recommendations: 0
Even using returns that are expressed as numbers (versus percentages), no stock exceeds a value of 9, so none are meeting the screen. I realize this is plausible, but just want to be sure I'm doing it right.

I used the screen in MSN and found that I got a full list. The top three (based on maximizing the 4th line sort) is STP, PBR and COGN. That uses 9+ stock ratings. However, a slight variation is to use Buy and Sell ratings in MSN (Analyst Proj/Mean Rec). With that approach, the top three are: FSLR, STP and PBR. The explanation of the differences being that COGN is rated as a Hold and FSLR is not rated by the SS. My inclination is to use MSN's Strong Buy/Medium Buy ratings.

Phil
Print the post Back To Top