For me I use it for back-testing. At a separate website, I select stocks to follow. They have a very useful tool that ranks about 8000 stocks. The best ones come to the top. I then look at the charts for the top 50 or so and then use my WL code developed back tester to select stocks to trade. This is working quite well. But how about you. I think some of you are using WL to select stocks. Is this working? I'm curious how these strategies work. What indicators are used?
Take Care.

QUOTE:
some of you are using WL to select stocks. Is this working?

This is an interesting question.

I run a three-tiered screening process.

1) For the first tier, I use the Fidelity stock screener. It creates some Excel worksheets of the winners, which a stand alone C# program reads and knocks out any stocks that are already in WL datasets. What's left gets saved into a WL screener dataset. The problem with these stock screeners is that they employ linear metrics (trading indicators) to evaluate the stocks. That's helpful, but that isn't going to reveal how well these stocks do in an event-oriented trading simulation, which is a nonlinear modeling (simulation) process.

2) For the second tier, I use a WL-based screener to evaluate how well each stock has recently followed a tread upwards. Both slope and R-squared are evaluated for a simple linear and 2nd-degree polynomial regression model. I would like to do more, but without running the WL optimizer against each candidate, I really can't fit a "real" event-trading model. This step knocks out about 70% of the tier 1 candidates. Yes, some good stocks get eliminated because they aren't currently doing well, but those are the breaks. (Part of the goal of the first tier screening is to pick stocks that are currently doing well.)

3) For the third tier, I run both production trading strategies against the remaining candidates. One strategy is a buy-high approach (Bull run); whereas, the other is a buy-low approach (Voss predictor)--they are unrelated. Yes, that means running the WL optimizer to fit PV parameters for all these final candidates. Then I examine the Equity Curve for each strategy side-by-side for each stock candidate. From there, I pick which stock belongs to which strategy or if it should be discarded altogether. Only about 10% of tier 2 make the final cut.

On the Equity Curve, I'm looking for a positive slope especially over the most recent months. A slope of zero doesn't count, and any negative slope behavior is bad. If you have a better way, I'm all ears.




The only step that matters in this process is the third step. One needs to match their new stock up with a strategy that shows a positive going equity curve. The first two steps aren't that important and can be done numerous ways.

I guess I have a 2 step process. The first is the 3rd party list and then I go to my WL strategy. I like to look at a 3 month window and I use 60 minute bars in the WL strategy. Here is my trading strategy.


The viewer _tab_2_5_2021 is the list from the 3rd party service. Take a look at a couple charts on the list.

QUOTE:
Take a look at a couple charts on the list and you will how good they are.

However, the question remains, how well do these stocks work with your strategy? You need to look at the Equity Curve of your strategy to determine that. Does the Equity Curve have a predominately positive slope over many recent trades?

It's very "normal" for one stock to do well with one of my strategies, and horrible with another. So you need to match up the stock with a particular strategy based on the results of its Equity Curve; otherwise, you're just fooling yourself.

I just looked at the first stock in your list, APPS, with my buy-high Bull-run strategy. As shown below, it indeed does very well in 2020 producing winning trades nearly every time. But if you look at the bottom of the screenshot, you'll see the profit-per-bar for "trading" is $454; whereas, the profit-per-bar for "buy-and-hold" is $2562--much higher. So this stock shouldn't be traded! Rather, one should invest in it, not trade it. Treat it as a growth stock to "hold", not trade. You'll make more money that way. You don't need a trading program (like WL) for this stock today.

Hi Superticker,
Yes, in hindsight you may be correct. But if one has only been trading this stock for a few months and the stock has not been around for several years, there is not enough information as to buy and hold or trade with stops etc. The strategy I'm using uses the DecyclerOscillator(Close, 40, 0.3) for decisions and a backup of a 10% trailing stop. This strategy has been working quite well. This keeps me busy too. Thank you for the comment.

P.S. That website also has tools for searches to look back for a list of stocks that have an average of a defined gain for the last specified number of years. This would be very useful for a buy and hold strategy.

QUOTE:
This [stock screener] website ... has tools [to create] ... a list of stocks that have an average ... defined gain for the last specified number of years [which may] be very useful for a buy and hold strategy.

A couple comments:

WL7 has a feature to create "themed" datasets automatically. One could certainly create one to scrape a screener website and place those stocks in a WL7 dataset. You might try doing that with WL7.

I don't think creating an event-driven buy-and-sell trading simulation makes sense for a buy-and-hold strategy. However, I do have a ranking WL "strategy" that ranks Fidelity sector mutual funds. By "mutual funds", I'm talking about instruments that you would not trade by design. And this ranking strategy, which does not trade at all, is designed to merit rank the different Fidelity sector funds. It's awkward to setup in WL6 (One needs to setup Lists<ScoreCard> and sort them.), but it may be easier to setup in WL7, which is better designed to compare the performance of one instrument against another.

QUOTE:
uses the DecyclerOscillator(Close,40,0.3) for decisions

I'm not familiar with this indicator, but I'll look at it. Start a new topic in the indicator forum if you want to discuss this.

Just out of curiosity, what was the stock selection criteria employed in the bottom listing of Post# 3? Was it simply looking at the top recent SMA performance of each stock in ranking order?

There are many stock screeners out there. Most have the most basic selection criteria available. Some are for free. The few that are very programmable cost money, which is reasonable since they are offering an expanded service. But most are based on linear parametrics, which won't correspond to an event-driven simulation model (as WL employs). But favorable linear performance can translate into opportunity for the event-driven simulation model.

I have also thought about screening stocks based on analyst's ratings. But once again, those ratings are hard to cross correlate with an event-driven simulation model.

You really don't know how your event-driven simulation strategy is going to perform until you test it and observe its resulting Equity Curve. There's no shortcut to avoiding that step. And how to best evaluate that Equity Curve is a topic in its own right.

Superticker,
That list is my main Watch list. The watch list was on the 3rd party's tool and the stocks are ranked by their proprietary ranking using their VST that combines several criteria. Were you referring to QDEL listing (it was at the bottom)? I once had QDEL, but several months ago, it dropped 10's of per cent at the opening. I don't like that one any more, but it has high value. By the way, yesterday I canceled all my stops as I was getting tired of getting stopped out of a high performing stock and then have it go higher an hour later. I will have to come up with another criteria as to when to sell. All my stocks right now are looking very good. How about a 20 day SMA?

QUOTE:
... the stocks are ranked by their proprietary ranking using their VST that combines several criteria.

So their ranking is proprietary. That's what I don't like about trying to grade stocks by analyst's ratings as well. There's no way to reproduce their results. At least with the Fidelity screener I know precisely what the criteria is. The only advantage of using a proprietary ranking is that they may wrap some salient fundamentals (earnings per share, P/E ratio, leading/trailing PEG ratios) into their merit metric, which I don't do directly in my mathematical model.

You're probably best off picking a merit ranking that works well with your strategy (relative to creating the most positively sloped equity curve) that is also open to how it obtains that ranking.

Scoring points for proprietary ranking, however, I did pick up 5 stocks out of your Post# 3 listing that my Post# 2 selection process dropped in its tier 1 or tier 2 steps. So proprietary rankings have their place as well.

Superticker,
What I like about their listing method is that it's not defined by a person or persons. Each stock in their list is rank by 4 criteria, they use RV (relative value), RS (relative safety), RT (relative timing), and CI (comfort index). These 4 are combined and the result is VST. One can use their spreadsheet listing and sort for VST or any of the other criteria across the top. They list 8000 stocks or so. I like to view the top 50 stocks ranked by VST, but my own watch list may other stocks that I have heard about and would like to keep track of them.

QUOTE:
Each stock in their list is rank by 4 criteria, they use RV (relative value), RS (relative safety), RT (relative timing), and CI (comfort index). These 4 are combined and the result is VST.

Yes, but how is that VST ranking going to play with the equity curve of your WL strategy?

You'll always have to test each VST ranking stock against the equity curve of each of your production strategies. That requirement will never go away. However, some stocks are easier to make money from than others--regardless of the WL strategy employed. And that is where the VST ranking can help. Even a bad WL strategy might be able to make money from a high ranking VST stock for a short time.

I really have only one Wealth-Lab strategy that I use, but it can and does evolve from time to time. Since a couple days ago when canceled all my Stops, I use the strategy to check out new stocks to buy. For about a 6 month period I look for at least 70% gain in the Performance area. Also a minimum of sharp drops. These stocks may see a rapid rise and that is a sign for some traders to say, "we need to take it down". But they come back in a day or so.