New P123 screens

Marc Gerstein and I have created eight new screens that can be found in the “Special Themes” category (https://www.portfolio123.com/app/opener/SCR?cat=8982 ). Besides two “Millennial Money” screens added in January, these are the first new screens we’ve added since 2015. We hope that both new and old users will find some ideas in them that you can use in your own systems.

Marc’s screens are:
Sensible Sentiment: https://www.portfolio123.com/app/screen/summary/232715?st=1&mt=1
Guardrails for Yield Hogs: https://www.portfolio123.com/app/screen/summary/232716?st=1&mt=1
Learning to Love Mediocrity: https://www.portfolio123.com/app/screen/summary/232721?st=1&mt=1
My screens are:
Small-Cap Winners: https://www.portfolio123.com/app/screen/summary/232655?st=1&mt=1
Focus on R&D: https://www.portfolio123.com/app/screen/summary/232653?st=1&mt=1
Focus on Sales: https://www.portfolio123.com/app/screen/summary/232654?st=1&mt=1
Ultra-Safe Stocks: https://www.portfolio123.com/app/screen/summary/232656?st=1&mt=1
Negative Book Value: https://www.portfolio123.com/app/screen/summary/232714?st=1&mt=1

Cool - thank you both and looking forward to checking them out!
Jerome

Hello Yuval.

That’s great. Many thanks to you and Marc.

Just two questions;

  • I backtested few of your screens and I realize that some of the screens start the backtest in 2004 and some of them in 2009. Is there any reason for that or it’s just a coincidence?

  • In case we want to implement one of those screens as a sim, for instance the “Small cap winners”, which ranking system do you recomend? Or what factors do you recommend to add in the rank in order to complement the screener?

Thanks for sharing, Yuval and Marc! This is much appreciated.

Thanks for sharing, gents!

How is it possible to add a ‘wizard rule’ to a custom universe using your screen rules?

See difference screen vs universe attached…



No good reason, really. Sorry for the inconsistency.

Well, you could experiment with the various core ranking systems: the default one, Core Combination, is a good one. Or you could adapt some of the criteria in the screens into your own ranking system. For example, with the “Small Cap Winners” screen, you could create a ranking system that consisted of Pr2SalesTTM, ProjPECurFY, Core: Sentiment, and Core: Quality, and get rid of the last four rules in the screen. The same applies to the other screens.

In the screen, click on the wand next to the rule. Then click on “Show Free Form Code.” Then you can, if you want, click on Update Free Form Rule and it’ll substitute the Wizard rule with a Free Form Rule that can be used in a Universe.

While you certainly can backtest these screens (I did) bear in mind that there’s more to life than just backtested automated strategies. P123 is also great for just-plain idea generating: get a list of names, check them (or some) out, and hit some interesting ideas. Watching all that’s been done and discussed on p123, it’s come to me that we’re lacking the fun of investing. I’ve lately been getting back to that and hope to inspire more of it going forward.

Yuval and Marc,

Thank you for the screens. I have been playing around with them a little and I wanted to pass along my slightly modified version to promote further idea generation in the P123 community.

I rebalance every 4 weeks, have 0.25% slippage above the daily mid-point and have the port 75% dollar hedged with the IWC Russsell Microcap ETF. While microcaps in general have been performing badly, they have been doing alright when you hedge them against a broad smallcap ETF. There is still alpha and it can be extracted.

Anyways, here is the modified version of the screen with hedging. I am curious to see what else people come up with. Thanks for a great screen by the way.


Forgot to provide link to the public screen.
https://www.portfolio123.com/app/screen/summary/232729?st=0&mt=1

This seems pretty cool to me, but–and please forgive my ignorance–how does hedging 75% by shorting an ETF work in practice? I mean, let’s say it’s time to rebalance. What do you actually do? What orders do you place? Do you have to use leverage? Can you do this in a retirement account? I’ve never done any shorting so I have no idea how this actually works. And why did you choose IWC? Is it because its performance is worse than any other benchmark ETF? Thanks, - Yuval

Sorry, I do not understand the hedging method (rules) here, since it do not appears in the screen.

Could you explain it?

Thanks.

I use IWC as it’s the closest thing to a microcap ETF. It takes it the good and the bad. So if you long good microcaps and short a broad group of microcaps (this ETF), you get the alpha. Every 4 weeks you calculate your liquidation value. This shows up in your TRADE account under account and positions. Just short 75 percent of that value. It requires leverage. The cash from the short sale will be placed in your account. If you can’t do it in a retirement account, just hold the ETF short in a separate account. Calculate the liquidation value by hand. Just make sure your long side is equal to liquidation value.

The hedge is IWC ETF and it is always on.

Very cool Kurtis. Learned something new today.

Can this be done in a sim/port? so it automatically spits out a trade? It is a lot of trades :slight_smile:

Thank you Kurtis, Yuval and the P123 team

To build on this and contribute as well…

I have not yet tried to turn it into a simulation but quickly looked at first steps in this direction starting from Kurtis’ version:
→ I suggest using the “core combination” ranking system with 30 stocks. Considering the micro-cap nature of the screen the more stocks the better (with some limits) to diversify away some of the risk of a blow-up.
→ There is a fairly high turnover (and since micro-caps → add implementation risk in getting the fills as expected) → rebalance every 2 weeks works well enough (as opposed to every week)

At this point and before spending more time on this, I am possibly concerned by curve fitting at inception (have seen too many R2G micro-caps blow up). Of course, we can wait for another 3 years and see what happened but…

Questions:

@Yuval → how much of the screen was built theoretically first (valuation logic in your head) vs testing all sorts of factors to see what worked in the past? And when implementing the screen with factors did you use the whole timeset 1999-2019 or first a subset of that then do a roll-forward to a new previously unseen period?

@Kurtis: same question

Thank you guys

Jerome

I wouldn’t be surprised if a strategy sim based on this screen performs poorly. The slippage schedule for sims is much more aggressive and, in my experience, kills many micro-cap screens.

Kurtis,
Problem is that your screen’s slippage of 0.25% is too low for small- and micro-cap stocks. Slippage should be more like 0.5%. Also a 0% carry cost of short position is unrealistic. With 5% carry cost and 0.5% slippage annualized return is about half of what is shown, but still has positive alpha.

Geov,

Slippage is different for everyone. It depends on how aggressive you are and the capital you are trading.

I set the min liquidity at $200K per stock and $2 per share. My experience is that I can trade over 5% of the daily volume (up to 10% in some instances) using the IB VWAP best efforts algo (non-aggressive) and we get 0.25% above the daily mid-point (not the open). There are 50-70 stocks.

You really should have no issue trading $1mm portfolio in this one idea alone. And because it rebalances every 4 weeks, why not take a few days to get in and out of positions. You can either lower your slippage even further on the $1mm portfolio or triple your portfolio size if that’s what you are after.

5% carry cost to short a popular ETF seems rather high to me. I don’t short it (my system shorts individual stocks of high liquidity) so I can’t say what the number is but that does seem a lot high.

As for trading systems and when they were developed, I have systems like this which trade 250 stocks. Many of my systems were developed years ago. They performed very well out of sample. To date, only one family office I consult with is interested in microcap and smallcap systems and that is without the hedging component. To me, this seems like an excellent way to lower volatility while harvesting the size premium. I even offered to run the portfolio for the firms through the P123 Trade interface. It is funny but there is just little interest in this sort of product for some reason. I have no idea why.

Kurtis,
It is not a bad strategy. One can add market timing to the buy rules which require rank to be higher during down-market periods. This result in the model reducing number of positions during down-markets.

Here are my rating rules:
Rating (“Core: low volatility”) > 25 (This is an excellent addition to the rules.)
Eval($comp<50,Rating (“Core: Sentiment”) > 75 ,Rating (“Core: Sentiment”) > 35)
Eval($comp<50,Rating (“Core: Quality”) > 65 ,Rating (“Core: Quality”) > 50)

$comp is the market timing formula based on the Composite Market Timer. An entry signal for the stock market arises when the combined weight of the six component models is equal to, or greater than 50. But any good timer will do.
https://imarketsignals.com/2016/composite-market-timing-increases-returns-and-reduces-drawdown/