- Joined
- 28 December 2013
- Posts
- 6,392
- Reactions
- 24,319
HI Skate (and all) Im just getting back into trading after a 20 year break and spent hours reading this thread on the weekend - very interesting, thanks. Its probably been mentioned, but when backtesting, what 'group' of ASX stocks are you using? eg ASX300, mid caps etc. I dont see any top 100 stocks in your lists (BHP, CBA etc), so interested to know your grouping/s of stocks you consider to trade and back test cheers Marty
I would like to offer my many thanks to Skate for the effort he has put into this thread. I had been hearing Skate "banging on"about mechanical systems long enough that I thought I should at least check it out. The biggest one being stock selection. I too trade the All Ords universe and of my now 38 scanned stocks NOT ONE matches Skates selection. No doubt we will have different exit strategies. My biggest problem now is what to do with my newly acquired free time if I continue this journey.
If you don't assign a value to PositionScore in AB, my understanding is that the positions will be chosen at random (but I'm not 100% with that).
1. I would like to offer my many thanks to Skate for the effort he has put into this thread.
2. Starting with a clean slate, I found inspiration within the excellent contributions of "ducati916" in the "Trading the Bounce" thread. Thank You.
3. Being a discretionary trader, I put together a weekly system that I was happy with.
4. My testing went well and showed positive expectancy. My major requirement.
5. Several interesting facts have emerged.
The biggest one being stock selection.
I too trade the All Ords universe and of my now 38 scanned stocks NOT ONE matches Skates selection.
6. I do not rely too much on indicators preferring to focus on price action. (Thanks to Captain Black).
7. I think that with a positive expectancy trading plan you could almost throw darts to make selections and be successful.
Technical noteIf you don't assign a value to PositionScore in AB, my understanding is that the positions will be chosen at random (but I'm not 100% with that).
You can make your PositionScore unique if you want. If its a long term uptrend you could make it the ROC (rate of change) over a certain period, or you could do linear regression value, if some people rely on RSI for signals, they can even do the RSI as the sole ranking factor. If you have an indicator, or anything that includes a number, it can become a factor in the PositionScore (i.e. RSI * 200day MA).
The skys the limit (but therein lies the problem, as well. With so many options to chose from, finding the most effective can be difficult at times).
Forget my post, @Lone Wolf gave the answer alreadyTechnical note
Position scoring is definitively one of my weakness, as i understand it, if not defined, AB is not random but chooses the packet with the more units(shares).
I was looking at this a while back after trying to understand a backtest where the position score had been removed
It is not random by default
@Nina4, what a great post, it was a pleasure to read on many fronts, squeezing in a few "thank you's" along the way, a very nice touch. I've cherry-picked your post & condensed your words into a few & attached them above. Why? because it gives me a reference to respond.
"Banging on"
First off, I agree, I do "bang on" about most things & I do have a habit of not knowing when to stop.
Skate.
So after reading your post there were some interesting points raised, which I have responded to below.
1. I agree, this thread has become a valuable resource.
2. You're welcome.
3. As a fellow discretionary trader, what finally pushed you to mechanical?
4. A good start.
5. Potentially indicating good breadth in the market.
6. As indicators are derivatives of price, they actually contain very valuable information: specifically 'divergence' and 'comparative confirmation/denial'. I am always interested when an indicator indicates a divergence or a comparative indicator confirms/denies the information contained in price, as these tend to be the strongest signals. I have also found that there is a time lag in both (highly predictive) before that effect is realised in price, which allows one plenty of time to take advantage of it. How this could be coded is a further issue for Mr Skate to address.
7. Possibly a contentious statement. Care to elaborate?
jog on
duc
The percentage gains I mentioned yesterday show positive movement, which is pleasing but are for the most part meaningless until sell signals are generated and the money, if any, is in the bank.
I'm on my Soapbox
When I start one of my posts with the "Soapbox" graphic it's a way to warn you that I'm about to express an opinion or get a few things off my chest. For a better word, I'll just be "banging on".
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?
We use cookies and similar technologies for the following purposes:
Do you accept cookies and these technologies?