However many of you have trouble finding a software to do the exact calculations for Double Trouble. The other problem you face with Double Trouble is many possible opportunities on day to day basis. To overcome such problem here is a modified Double Trouble method. I can not think of a new name for it currently, but at some stage will rename it.
This is extremely Telechart friendly version.
- Calculate relative rank using : ((C - AVGC135 ) / AVGC135)
- Take the top 2% stocks ranked by the above scan.
- Run a combined month weakness plus breakout scan on the top 2% stock. :(100 * (C1 - C22) / C22) < 10 AND ( 100 * (C - C1) / C1) >= 4 AND V >= 1000 AND V > V1
- You will get 0 to 5 candidates on daily basis. If you want even lesser number of candidates take 1% only in step 2. In that case you will get 0 to 3 candidates on daily basis in most circumstances.
- Initial stop at 3 days low. Trail after 20% profit with 10% stop.
- Risk 1% of equity.
Members Only
Be a part of the 1000 member strong traders community.....
24 comments:
Hi Pradeep, thanks for the Double Trouble update. What's the best alternative to TeleChart for doing such scans? I heard some brokers may provide similar software, anyone has recommendations or experience to share?
By the way, I'm checking out AmiBroker, while it's got more features than QuoteTracker, I haven't found the justification to pay for it since QuoteTracker is essentially free.
If I decide to subscribe to TC, will the "Gold" service suffice, or I have to buy the "Platinum"? Thanks!
Hi. I have two questions.
1. Can you please explain this formula step by step? I usually try to emulate your formulas in stockfetcher and your step wise instructions will help me in doing that.
2. You have mentioned few times in past that you buy at one-month weakness in stock price and I understand that by weakness you mean that stocks which have grown by less than 10% in past one month (or in past 65 days). Can you please explain that while considering weakness in a stock, would you also include the stocks which have decreased in price from their 1 month (or 65 days) high but are still within 10% of their 1 month (or 65 day) high?
Your insight will be much appreciated. Thanks.
rhoniel
365.25 / 135 * ((C - AVGC135 ) / AVGC135)
c- closing price
avgc135=average of last 135 days closing price
What does this calculate it calculates price change from average price in last 135 days (half year approx) and then converts it in to one year equivalent for comparison purpose. 365.25 is number of days in year when leap year is taken in to consideration.
(100 * (C1 - C22) / C22) < 10 AND ( 100 * (C - C1) / C1) >= 4 AND V >= 1000 AND V > V1
Lets break this in to three different parts for understanding purpose.
(100 * (C1 - C22) / C22) < 10
Means price % change in last one month(21 trading days approx) as of one trading day ago was less than 10%
c1=close yesterday
c22= close 22 days ago (21+1=22) because we are looking at 21 days growth as of one trading day ago.
( 100 * (C - C1) / C1) >= 4
Price change today is 4%
c= close today
c1=close one day ago
V >= 1000 AND V > V1
v=volume in Telechart 1000=100000 shares.
So volume today was >100000
V>v1 means volume is greater than volume one day ago.
Rhoniel said...
2. You have mentioned few times in past that you buy at one-month weakness in stock price and I understand that by weakness you mean that stocks which have grown by less than 10% in past one month (or in past 65 days). Can you please explain that while considering weakness in a stock, would you also include the stocks which have decreased in price from their 1 month (or 65 days) high but are still within 10% of their 1 month (or 65 day) high?
Your insight will be much appreciated. Thanks.
-----------------------------
The one month or 65 day weakness in a strongly trending stock or stock with high relative strength is where technical analyst find their fancy cup and handle, flat bases, bullflags. and many other subjective patterns.
More objective way or one of the better way to do it is by taking stocks with price growth of less than 10 or 15% in last one month or 65 days. If we use mathematics we get more objective methods. Otherwise one man's cup and handle is another persons developing head and shoulder or some other esoteric pattern.
The reason to do this is one is not buying an overextended stock with this filter, plus my testing shows significant improvement in profit once this filter is added.
Pradeep
Is it possible on just this occasion that you confirm the following results for all of us who use Telechart?
Step 2: 144 (top 2% of ranked stock) or 72 (top 1%)
Step 3: BWLD HRSH IDSA MDII MTOX RBY SNHY
Step 4: IDSA MDII MTOX SNHY (top 1% of ranked stock)
Also, would does '3 days low' = MINC2?
This probably makes no sense to anyone but me, but I would call the method 'Son of Double Trouble'
If you take 1% only you will get
IDSA, HRSH as final output.
If you take 2% you will get :
MDII, HRSH, IDSA, MTOX, SNHY
I don't know what is your step 3.
3 day low will be minc3
Let us call it Son of DT
Sky
Just investigate if there is any problem in your scan.
1% gives 35 stocks
2% gives 107
The ranking functions in Telechart can give you varying results sometime, but your 1% figure (72)is way off.
Sky
In Australia there is someone who runs lots of seminar propagating similar method using ROAR.
You may want to study his approach.
Calculate relative rank using : 365.25 / 135 * ((C - AVGC135 ) / AVGC135)
since you are only looking at top 2% 365.25 / 135 is not necessary
You are right
Following your modified directions, I created Son of DblTrbl pcf scans 1 & 2 early this morning. Then created an EasyScan to incorporate both of your new scans. Scan #1 I set List Rank 97-99 (is that correct?). Then added Scan #2 for a result of zero. I note you said that 0-3 candidates could be the result. I just want to make sure I'm on the right track please, and setup my easyscan correctly.
97-99 will give you 8 candidates. Setting it up at 99 will give you 2
Weird. Not on mine.
Scan 1 365.25 / 135 * ((C - AVGC135 ) / AVGC135)
Gives me 104 results. Sitting bottom number on 99-99 results in 35 hits.
Scan 2 (100 * (C1 - C22) / C22) = 4 AND V >= 1000 AND V > V1
Only shows true for qualified stocks. And gives me zero results.
I am running these 2 scans in an easyscan on AllStocks. May I ask what am I doing wrong?
TIA
Art
Your second scan is wrong:
use this
(100 * (C1 - C22) / C22) < 10 AND ( 100 * (C - C1) / C1) >= 4 AND V >= 1000 AND V > V1
That was my problem that somehow I missed. Thank you for you patience.
Pradeep,
Before it hits 20%, do you move the stop to breakeven at some level?
Also, I assume the time stop of 5 days also applies to this?
Thanks,
Rahul
Yes and yes
Can you please tell the conditions under which to move stop to breakeven before 20%?
Rahul
after 5 days
To the person asking about alternatives to TeleChart.....
MSN Investor has the best free stock screener (IMHO) and a reasonable substitute for TeleChart. I've been using MSN Investor since 1999 (tried tradestation for a bit in 1999, though it was overkill for my needs). Main downside: no real-time data update so best to run screens in the evening (non-issue for this purpose though).
Stockcharts.com also has a pretty good web-based screener for a reasonable fee.
While I know you can choose Display Value or Display Market Rank in Telechart, does anyone know how Market Rank is calculated?
If there are total "n" number of values, to find percentile rank of "a" value.
(100*number of scores below "a")/"n"
So if you have 10 values
40 ,50, 65, 70, 72, 75, 78, 80, 82, 84
You want to find percentile rank of 70 then, there are 3 positions below 6 so
100*3/10 = 30 is the % rank.
If you take 82 then
100*8/10=80 is the rank
Pradeep:
Have you analyzed how much % of stock in double trouble list will make big profits? What's the average percentage of gain for those double trouble candidates? The inclusiveness of large gain in the candidate list and the percentage of them turning into superior win are two different issues.
Good question. One is not buying every stock in that 1% list only stock meeting a condition(4% breakout with 30 days weakness). Which gives you few signals. The question is what happens to them.
The average profitability on profitable trades, post that event is 28%. The loss on losing trades is capped by stop by 1% but in real trading tends to be around average 1.4%. The system produces more winning trades than losing trades and outperforms market for last 10 years of testing.In bear market it does not produce too many signals as few stocks have 4% plus breakouts during bearish phase.
One must do ones own testing before trading any system.Just because a stock is up a lot does not mean it will go down.
In fact anyone who does any amount of testing on market will discover immediately that relative strength based systems perform very well in markets. The problem is you need stomach to trade those system.
Post a Comment