Weekly Portfolio - ASX

@Newt : I agree. It should be a good opportunity to align the system a bit better. I don't have a lot of experience, however, so this will be trial an error and should help to improve my ability to look at charts. What I do want to see is rising volume matched with rising prices (based on some volume analysis I've been reading) which should be (relatively) easy to code into scoring. I'm still a novice at coding so we will see how long this takes me, haha.

And what do you mean by score low trades I'd struggle to reliably take? Do you mean, score unreliable trades with a low score and place them at the bottom of the list? I'm also wondering if something like a restriction on maximum trades per week would be beneficial.

For me, its hard to take trades in very volatile stocks that are bouncing around. This doesn't have to be overly complex programming, so proving say reliable >10% benefit consistently across many time intervals for something based off rising price momentum with a measure of overall smoothness might be enough. You've noted yourself less conditions is often more robust code/system in the end.

Personally I've never had much luck PositionScoring on Volume, but you might be smarter than me. I do believe it should be a buy condition in some form however.

Yes, by low score I mean you might be able to filter out trades you know would be psychologically harder to take. This is just a reversal of "what conditions do I think would give better trades on average"? I was always a bit sceptical of using PositionScore, but learning from Skate convinced me its important to get a clear cut list of buys and sells each week or on your chosen trading time interval. Too easy otherwise to try and "cherry pick" yourself and fall into all the psychological traps of under-trading, over-trading, missing good winners.

Trend trading is typically a coin flip whether you get it right or not, hopefully with a larger average win than loss. Its just so easy to waste emotional energy and trading accuracy trying to predict if the coin will come up heads or tails (i.e. win or loss), when really you should be concentrating on process and following the system - getting in as many coin flips as you can when your buy conditions are met

Thanks @Newt . I appreciate the reply. I've only started to take a closer look into volume. And with the help @Skate gave me with a very simple and basic PositionScore code I found my backtests were much better.

And I'm not trying to predict the market, which as you suggest is a waste of energy. As I explain to friends, I'm trying to find a statistically probably outcome in my favour. I expect only 30% of winners but I'm trying to cut losses as quickly as I can without doing so prematurely. Will never fully get it right, but not trying to. Trying to predict the market with a systematic approach I think would lead to some over-fitting and frustrated live-testing.

System 1 (EMA Cross strat):
Week 15:
Buy: PAN, NWH, AHY, DCG, MAH
Sell: None

NWH was sold for a 0.2R loss.

System 2 (MAP strat):
Week 2:
Buy: MYX, API, NWH, AQZ, PSI, HLO
Sell: None

(now at 19 open positions for system 2)

I am aware that 1 of my systems has sold one position while they both buy it a week later. I'm not worried as I have faith in my systems.

Boggo said:

One (of many) that I have read and thought worthwhile...

View attachment 98133

Click to expand...

I finished reading this today. Definitely a good book if you want to do system trading! I got a lot of good things from it, mostly about considerations and validity of systems you design.

System 1 (EMA Cross Strat):
Week 16:
Buy: AWC, RDC
Sell: None

Open positions: 12 (with 2 more to be opened on Monday)

The profit for this system still remains low (aprox. $800 from a $20,000 initial capital). EHE, RBL, IFN all remain my positions that are keeping me in the black! Only minor losses on my other positions but they add up. My Win% is 41% right now which is above the 30-35% expected. Expectancy is at 0.51. Having a look at the charts for my positions and I notice there are a lot of positions that appear to be consolidating. Looking at the charts as I trade this system and I notice that I get in on a position and then it begins to pull-back/consolidate, which I know is normal. It is a double edge sword to have a system that quickly jumps on a trend as it may do so a little early or have the position never breakout and continue the trend.

System 2 (MAP Strat):
Week 3:
Buy: LNG (more signals generated but only 1 open position available)
Sell: None.

System 2 had a starting capital of $50,000. It current has an unrealised PL of $1776. From the get go this system appears to be working well. I also worked on my code so when I do explorations within AB I can keep track of my metrics. I have noticed since I started paper trading system 1 that even basic things of 'what is my current trailing stop', and for system 2 is my index filter still indicating an 'up trend' (impacts my trailing stop). These metrics are in my code when I run a backtest but not displayed to me when I have to manually do everything. I also included my position score in a column to help me out, and working on HHV since buy, number of bars since buy, etc. These can be calculated but since my code for stale exit and trailing stop is an IFF statement and not a loop its not all working out just yet (or maybe I'm doing it wrong?). But a good skeleton has been formed and my coding skills have improved and my systems will become more consistent as a result. (It's all rather basic, but coding for a backtest is one thing, adding additional code to ensure it is carried on in this paper trading is a different thing).

-----------------------------------------------------------------

I continue to read books or posts within ASF to get new ideas or improve my current systems. I have learned from Davey's book that I definitely haven't done the best thing when developing my systems in regards to data. His book is good and gives a lot of insight into developing systems, the use of optimization, data, forward testing, and incubation (which I believe would include this paper trading). I think Davey is too stringent in some of his ideas, but he is also working on a whole different ball game and develops systems fulltime. Not everything of his can be implemented but it does give some good direction! Worth a read.

I was reading Clenow's book 'Following the Trend'. The book focuses on futures but I am liking his approach to diversification and risk. It is definitely reinforcing the want to add other systems (that aren't directly correlated to your current system) in order to smooth out equity curves. Clenow shows how simple things are and has a certain way of demystifying the big boys systems. The downside, especially for most people, is that you need a decent amount of capital to start one system, and obviously need even more if you have another. I'm a long way off from being able to do things on his level. He doesn't really deal with shares, but using his ideas are definitely transferable.

I have also had a number of ideas lately on future systems. I have been following @Skate 's CAM Hybrid and still toy with modifying it. I will continue to toy with possible modifications and employ some of the new system development ideas I got from Davey. Lately I have also been thinking about coding a system that looks for consolidation and potential breakout targets. Just using rising volume matched with higher low's over consecutive weeks. My idea is simple and shouldn't be hard to code/test (essentially will be looking at rising wedges/flags). I'm sure there is already code out there too, but I will be trading to code and test this myself to match exactly what I want.

Finished Clenow's "Following The Trend". The last parts were definitely interesting (he reverse engineers some trend systems from big funds and starts going over comparisons). If only I had $1,000,000 that he suggests is needed, haha. There is also a section where he shows that while he does not like investing in equities (shares), such a system could be aided with 30-40% used within futures trading to add to CAR without increasing risk. He also shows how things are a little less volatility if you are using a long only system (you do sacrifice some returns but max draw down also decreases).

I was on amazing and saw that Clenow's other book "Stocks on the Move: Beating the Market with Hedge Fund Momentum Strategies" on special for $2.55. Can't pass that up.

I think I need to start a thread with all of these books haha.

Clenow's books are great.
Following the trend really has some great insights into mechanisms of trend following.
If you are into podcasts check out top traders unplugged. They interview Clenow which is great but also have some other nice discussions for people interested in trend following.

Good to know @jjbinks . I will be driving accross the nullabour in a few months, perhaps I will download some podcasts for the drive? Could be interesting. And I am definitely a fan of Clenow's books. He has something against equity's as an asset class to trade in, but at least he doesn't completely dismiss it.

I think one of his main points particularly with futures is that it offers diversification and you can trade different commodities both up and down which are not correlated. But as you mention to take advantage of the ability to diversify you need at least 1mill capital!

I agree. And if his point is that then I definitely agree. I don't think he's against trading in equities but he definitely prefers futures. Futures could be commodities, bonds, currencies, etc., so I see his point. But I don't have a mil to do this ! haha. But obviously equities have a strong correlation to each other, though some sectors will do well over others, in a GFC like event it all starts to form correlation to each other. He also makes that point in "Following the Trend". As markets start going down, those that weren't previously correlated begin to in a strong bear market. I have wondered, would trading in multi countries be a solution for us? I.e. run a system in Aus, USA, euro markets, etc. One day I'm sure ..... haha

System 1 (EMA Cross Strat):
Week 17:
Buy: None
Sell: None

System 2 (MAP Strat):
Week 4:
Buy: RUL
Sell: None

Not much to report this week. It looks like almost every position I current have open took a hit to some degree. System 1 remains in the black while system 2 is in the red.

I finished reading Clenow's other book (it was short so didn't take long). I will likely post something separate about that. What I will say, though, is that it is a great book and I highly recommend it for any trend or momentum traders out there!

hi @Warr87

I was looking at Norgate subscriptions and noted that you can transfer your remaining premium data subcription prorata to norgate. (In case you are interested)

Cheers
j

Thanks! I may just do that. I had a look at the python plugin and it looks good! I hope it wont be too much hassle for the change if I do go through with it.

Thanks

Warr

Yep python plug in is great. Have had some issues but I haven't had a chance to code for the last 2 or so months and there has been a fair few updates
@Richard Dale has also made scripts based on Clenow's book available. Haven't had a chance to try them. (Let me know if you do!)

That's awesome! It'd be interested to see what those scripts would look like. I have some amibroker code for Clenow's 'Following the Trend' using his linear regression model for ranking. But I'd be interested in seeing what a representation of his systems would look like in python.

The plan would be to run a python script along side the amibroker version and see if it all matches. Not sure how much time I'll get for that right now, but it'll be a good project to eventually reach my goal of fully automating my systems.

If you want the Python code to Andreas' latest book adapted to use Norgate Data, please email support@norgatedata.com and I'll send you the link to the code.

Thanks @Richard Dale

System 1 (EMA Cross Strat):
Week 18 & 19:
Buy: VEA (2/12/19)
Sell: None

Not much going on here. Only 1 buy signal for the 2 weeks. Best performers remain RBL, IFN, and EHE.

Open Positions: 15

System 2 (MAP Strat):
Week 5 & 6:
Buy:MSB,JLG,INA,ADH,AEF,EML,OCL,BPT,AQG (9/12/19)
Sell: ERA, LNG, SWM, ISU, SNK, MVF, OFX, MYX, API, HLO (sold 2/12/19), PAN (to be sold 9/12/19)

The MAP strat was much busier in comparison. Plenty of sell signals. No shortage of buy positions to fill the gap.

Open Positions: 18