I get that there are different opinions in the community regarding buy/sell rules, so I'd like to know what your buy/sell rules are and if you are willing to share them. For simplicity, let's assume a modified Easy2Trade Universe (~3k stocks). where I integrate multiple other restriction rules into my Universe, so I'm now excluding many buy/sell rules and ONLY using the Ranking system.
Here are my Buy/Sell Rules against a 25-position long system:
Buy: Rank>99
Sell: RankPos > 60 and NoDays > 5 and StaleStmt = 0
Why: The Highest Ranks are the best stocks to buy based upon my ranking system. Sell with a little wiggle room when stocks fall out of the top 60 positions and have been held for at least a week, and sell if the company fails to report on time.
For super high turnover systems I will sometimes have a rule where: rank<99 and (gainpct-benchpct<-2 or gainpct-benchpct>2)
It doesn't have to be 2%. It could be 5 or 10 or whatever. But the idea is that I wait until the stock does something - good or bad, before selling it just on rank.
As I said, more so for higher turnover systems where it tries to sell every week.
I exclude stocks with special dividends since the latest filing, because they change the large calculations in my system. I don't want stocks that are late on filing, because when that happens, they will often just go dark at 120 days and you are fcked. With the RankPos sell rule i always have the best stock according to my ranking system in the system, just 10 ranks of wiggle room (30 stock system).
These rules where developed after large unexpected losses and so far i don't think i need more. (But lets see what the market throws at me in the future)
Metrics which are good screens (e.g. AvgVol, Exchanges, Marketcap limits, threshold profitability, etc.) --> use in universe rules
Metrics which are good sorts (classic factors like multiples, momentum, etc.) --> use in ranking system rules
Rules which depend on live portfolio data (Days since last sell/buy, maxcorrel to other holdings, Country or Sector Count in Portfolio) --> use in Buy/Sell Rules (together with a basic Sell if Rank<x
Point 1 avoids noise in the ranking due to stocks you wouldn't buy anyway.
Point 2 helps to make the most of the best criteria to rank stocks with max. flexibility regarding market situation and universe (e.g. strict a momentum rule in either universe or buy rules instead, messes a lot with universe size and/or # of skipped stocks in the ranking depending on market regime. That can be exactly what you want, but imo makes the whole system less robust and adds much noise).
Point 3 makes sure that you keep your strategy robust and simple while giving you the power to control risk and concentration.
Not sure if my current 40 stock port will live up to expectations. It was just recently launched in place of a model used about 4 years with lots of buy and sell rules. Here is my basic approach and why. The port expects to hold 40 fundamentally strong stocks in normal times.
All buy rules (all fundamental in nature rather than technical) except a turnover test rule (to avoid difficult to trade stocks) are in the custom universe. By doing that, the onus is on the universe to filter candidates and provide a sell signal when fundamentals deteriorate and on the ranking system to recommend replacement stocks. Ranking stocks that won't be bought because of buy rules could require judgement regarding where the holding rank cutoff should be. I preferred to focus effort on the universe and ranking interplay so that the top buckets consistently outperform over time and are not full of stocks that won't pass buy rules applied after ranking.
There are no sell rules, so stocks are held until they drop out of the universe. The only buy rule requires a stock's turnover to be sufficient for a likely easy trade. I currently require a minimum daily turnover within the last 15 trading days to be 50 times what the purchase would normally be (so I am likely to buy no more than 2% of the day's turnover).
The port currently has no sell rules, so turnover only happens when a held stock fails a fundamental rule in the universe. Since the universe is heavily filtered, it is still possible for fewer than 40 stocks to be held at any time, but in general more stocks are held than if I relied on a rank or rankpos rule. My previous port suffered from holding few stocks at times when it would have been better to allow lower ranked stocks in.
This approach should allow me to focus more on the ranking system's makeup and consistency over time. Good luck with your efforts!