Net nets

[Rayvt]

I used to do everything in Excel. But with these complicated factors the spreadsheets got pretty complex. And slow. It is very cumbersome to do multiple sorts non-manually in Excel. Being cheap, I have not upgraded Office 2003 to a newer version. Perhaps LibreOffice could do it?

Now I generally do everything in bash scripts. Using Cygwin (Linux on Windows) for the (few) screens I run on Windows, and native bash on the (majority) of the screens I run on a Linux box.

On Windows my script that does the 4-factor screen that Jim just mentioned takes only 4 seconds to go from the raw data file to spitting out the list of buys/sells/keeps. Most of that time is disk I/O.

[Inspired2learn]

Once again, thank so much. I tried to diligently do this several times, but my list has only some of the names that Jim mentioned…and I am wondering if this is simply because I am using the librbary access, and the paid subscription version would better allow these screens to produce the accurate results? In other words, is there anyone else who is using the library access and still getting the same picks that Jim got? I am more than happy to get the Valuleline subscription if that is the right way to do it, but just wanted to make sure (also, there seems to be numerous Value line services, and so not use which one to buy either)

Also, when I did it, it took a lot of time (not complaining at all as I really wanted to learn) but felt there must be a more efficient way, which you all must be using. So, assuming everyone else here is using the paid version, does it automatically allow the 4 screeners to run at the same time?

And do you then manually assign a score from 1 to 800 or each of those stocks on the 4 criteria…and manually compute the sum of 4 criteria?

Thanks so much for patiently trying to teach…eager to learn, and admittedly I am in the Pre-K in this university, but definitely trying to learn and course correct what has been a disastrous journey of investing for me.

Thanks a lot,
Charlie

[comorgan]

Subtle distinction: since VL 52 week highs are anywhere up to a month old, the ratio of price to [lagged] 52 week high can be quite a bit above 1.
Technically you want the highest ratio, not “closest to 52 week high”.

I think I run screens differently than most of you. It sounds like most of you download the database and then run your screens? I run my screens in the Classic Screener, then download to exel. When I’ve spot checked my results against what Jim and others have posted in the past, I’ve gotten the same results, provided I use the correct fields.

In this case, it sounds like something is different. I’ve never gotten a price ratio higher than 1. I also get a few stocks that have no 52 week high listed. I don’t know what to do with this information, but mention it as it may have some relevance to Inspired2learn’s comment about the library edition. I don’t know, my library only offers the print edition.

Find the top 800 stocks with highest 5-year sales growth rate
Jim, I’ve found that each week there are a fair number of stocks that do not list a value for 5-year sales growth rate. I think you’ve mentioned in the past that it is a good sign for a stock to have the value listed, so I always delete stocks with no 5-year sales growth rate from my screens that utilize this field.

What do you do with stocks that list no 5-year sales growth rate, ROE, or 52 week high? For example, say a small stock has none of these, so scores zero for each, but is small enough to be included in the screen. I know you replied to Dragontales that is why we back test, but including a passing stock solely because it’s small cap doesn’t sound like a good idea. Obviously, the back test is good, but that step intuitively seems like an opportunity for improvement without overtuning. It just seems like giving the best stock an 800 score and summing for the highest total would be better?

[mungofitch]

In this case, it sounds like something is different. I’ve never gotten a price ratio higher than 1.
I also get a few stocks that have no 52 week high listed. I don’t know what to do with this
information, but mention it as it may have some relevance to Inspired2learn’s comment about the library edition.
I don’t know, my library only offers the print edition.

It’s reasonably likely the 52 week high gets updated when you log on at the library, and this is causing the difference.
In the weekly download files, the “52 week high” field is updated only at the end of each calendar month, I’m pretty sure.
Actually, the first weekly release that is at least a couple of days after the month end.

In the case of this particular screen, there is only one thing which depends on the current price.
The other fields change very slowly. (the market cap ranking isn’t a big deal)
And it’s pretty well known that “distance from high” momentum, like total return 1 year, benefits from a bit of lag.
So one hack solution is simply to use the ranks a week or two after you calculate them.

Each week there are a fair number of stocks that do not list a value for 5-year sales growth rate.
I think you’ve mentioned in the past that it is a good sign for a stock to have the value listed, so
I always delete stocks with no 5-year sales growth rate from my screens that utilize this field.
What do you do with stocks that list no 5-year sales growth rate, ROE, or 52 week high?

The general rule for all our screens is to deblank required fields (unless the screen is very specifically constructed not to).
If the field isn’t populated, the stock is not considered further.
But…
This is a bit complicated in the case of a screen-of-screens doing a sum of ranks, as it has four subscreens.
So the stock would be eliminated from one but perhaps not the other of the subscreens.
I think (?) this would just end up giving it a worse score, but I’m not sure.
I could check, but I’m a bit lazy/busy today.

The two main reasons for lacking a 5-year-sales-growth field are that either the firm has been trading for less than five years, or it’s a financial firm with no meaningful revenue.
A bank, thrift, or fund, basically.
As Value Line does not report sales figures for banks, there is no sales growth figure calculated.
(This makes some sense…the revenue of a bank that most sources report is actually their gross margin. But you have to remember it)

The traditional way to do all this is to get a Value Line “Analyser” subscription, export the weekly data yourself,
create the screen in Radiscript once, and run the Radiscreen Excel macro to generate the picks.
That way you get all the same behaviour that has traditionally been used in all the backtesting.
Freebies at the library are not always good deal.

Jim