Half baked trading scheme (Market Depth)

[LifeChoices]

Re: Half baked trading scheme

Ok, this one then
.

Sorry, that one doesn't make it to my watchlist. I look at stocks that have either dropped a lot or gone up a lot - ie in the past 12 sessions my computer picks these. I do this daily and do different combinations and look at graphs and read stuff. After doing this all day during my dull day job, stocks just start to stand out. Getting my watchlist is a black art.

+-------+---------+
| code | change |
+-------+---------+
| CVN | -25.81% |
| ERA | -24.96% |
| YTC | -24.5% |
| RMS | -23.73% |
| RSG | -17.58% |
| SDL | -17.58% |
| GRY | -17.07% |
| GXY | -16.54% |
| BDR | -16% |
| PRR | -15.79% |
+-------+---------+

+-------+---------+
| code | change |
+-------+---------+
| MSB | 20.92% |
| TEL | 8.7% |
| CPA | 6.51% |
| NAB | 6.07% |
| SGM | 5.99% |
| TLS | 5.69% |
| FXL | 5.56% |
| ABY | 5.49% |
| KCN | 5.33% |
| EQN | 5.11% |
+-------+---------+

Besides that CEU hasn't really moved anywhere, it comes up at 124% - not as strong to me CVN, HST or MSB.

 

[Gringotts Bank]

Re: Half baked trading scheme

Everyone will be signing up for a trial of Turbo Trader. I better email him and ask for a free month!

Remember though, tools like this can be very handy, and systems are nice for confidence, but you the trader have to make it all work.

 

[LifeChoices]

Re: Half baked trading scheme

Everyone will be signing up for a trial of Turbo Trader. I better email him and ask for a free month!

Remember though, tools like this can be very handy, and systems are nice for confidence, but you the trader have to make it all work.

I can trust my own half baked system. I think ERA is a good candidate for a buy sometime soon. Earlier today the Buy/Sell ratio was -30%, which wouldn't give me confidence to buy into it just yet.

But I can see those negatives turning into positives very soon, could even be in the next few days/weeks. I'll post it here when I notice a turnaround - and we will all be rich beyond our wildest dreams.

Oh yeah - I forgot to say gringo - couldn't agree with your last sentence more.

 

[Gringotts Bank]

Re: Half baked trading scheme

I'll hop on ERA when you give the signal, if it feels right. Posting on here or ERA thread?

Just for interest, might be useful to scrape the number of buyers vs number of sellers as well as the total shares on either side of the MD. If you had heaps of sellers and only one HUGE buyer, with buyer vol > seller volume, I wouldn't touch that!

 

[LifeChoices]

Re: Half baked trading scheme

I'll hop on ERA when you give the signal, if it feels right. Posting on here or ERA thread?

Just for interest, might be useful to scrape the number of buyers vs number of sellers as well as the total shares on either side of the MD. If you had heaps of sellers and only one HUGE buyer, with buyer vol > seller volume, I wouldn't touch that!

I'll just post when it turns around. I think it will soon. With both HST and CVN leading up to the turn around, it was still very strong selling - low prices and strong number of sellers. Even though ERA went up today, it was weak and the stock now looks like it's at a 5 or 6 year low. I'm not sure if there is a reason for it - but just from a trend perspective it does look like a good candidate to test this half baked trading scheme.

 

[tech/a]

Re: Half baked trading scheme

Everyone will be signing up for a trial of Turbo Trader. I better email him and ask for a free month!

Remember though, tools like this can be very handy, and systems are nice for confidence, but you the trader have to make it all work.

Ah all is clear now.
I see where your mis understanding originates.

 

[Gringotts Bank]

Re: Half baked trading scheme

Do I have to play a guessing game or what?

 

[tech/a]

Re: Half baked trading scheme

Remember though, tools like this can be very handy, and systems are nice for confidence, but you the trader have to make it all work.

Other than following a system with positive expectancy and the blueprint you have from
valid testing---the trader need do little else.
Even a Duck can do it.

Perhaps this would be best commented in another thread on perhaps systems?

 

[skc]

Re: Half baked trading scheme

I wrote a script that auto logs on to my account and scrapes data for any stock I choose - usually my watchlists, I then just worked on different algorithms and watched them, I thought the buy/sell volume ratio thing was particularly interesting to watch.

The next step I want to do is to hook it up to a timer so say every couple of minutes it will fetch data from chosen stocks and send me an alert if some matched criterea is reached. ie CVN < 0, ERA > 0

I have another script that mines the ASX site every day for all data. From that and my own research I develop my watchlist, I like stocks with turnover above 1M with big falls gains in past week/month/year. This is limited to about 10 - 20 stocks.

Very impressive. But is Comsec market depth actually live? Or 20 min delayed?

You can add FXJ, APN and CSR to your watch list.

Also, go sign up to a trial account with SPARK. They have the best live market depth graphics imo. You can also replay market depth over time. I think that will be very valuable in changing your half baked idea into a possible system. I personally think your half baked idea will work better with large caps.

And lastly, have a look at SGT's depth. I can never understand what's going on. There are usually a few K shares on each price level, but there are often 2 levels of 600K+ shares in the bid, which then switches to the ask in 20 minutes, then switches back... with the price barely moving. You need to be careful about things like that which might trigger your filter for the wrong seasons.

 

[Gringotts Bank]

Re: Half baked trading scheme

Not sure where to continue it, but I don't think anyone will care too much. My stance on system vs. individual goes like this:

- A very robust system cannot be traded profitably by someone with the wrong attitude, even if every attempt is made to follow the system accurately.

- A "throw a dart at the financial pages" approach will be profitable for anyone with a good positive attitude. By good positive attitude I don't mean grinning like an idiot and saying "I am rich!! I am rich!", but having a positive expectancy that is not based in any testing. The expectancy is psychological.

- The ideal is a good system with a good attitude. In my view the system contributes 10% and the attitude contributes 90% to the profits (or losses).

I base this on my own extensive experiences (with very good systems and dart board approaches) and that of others. I know it's unusual, and "out there", but there you have it. I also know that you and most others in here disagree, That's fine we can hold different views.

 

[Synergy]

Re: Half baked trading scheme

I like this idea...

I think it could be refined a little, but it makes a lot of sense to me. The part I like is that you're looking for a double confirmation. The drop or rise in price leading to a price turning point, and the turning of the market depth.

Although the volume filter should help minimise the effect of large orders, I'd feel more comfortable if there was something to limit the effect of large orders - especially those well away from the price action. Could you restrict it orders within a set distance away from the price action. I like the idea of using number of orders also.

If you get the double confirmation, and buy in after the price has turned, the other advantage is that you have probably also got a price turning point on good volume, which may well act as support if the price does turn against you. Low risk entry, even though you're basically buying against the trend, or at least in the very early stages of a new one.

 

[LifeChoices]

Re: Half baked trading scheme

I like this idea...

I think it could be refined a little, but it makes a lot of sense to me. The part I like is that you're looking for a double confirmation. The drop or rise in price leading to a price turning point, and the turning of the market depth.

Although the volume filter should help minimise the effect of large orders, I'd feel more comfortable if there was something to limit the effect of large orders - especially those well away from the price action. Could you restrict it orders within a set distance away from the price action. I like the idea of using number of orders also.

If you get the double confirmation, and buy in after the price has turned, the other advantage is that you have probably also got a price turning point on good volume, which may well act as support if the price does turn against you. Low risk entry, even though you're basically buying against the trend, or at least in the very early stages of a new one.

Yeah, I think I'm on to something also. I love the despondent threads in CVN and HST for example ....

scroll down for hangeng https://www.aussiestockforums.com/forums/showthread.php?t=10119&page=3 - very inspiring And then there's another Garpal Gumnut https://www.aussiestockforums.com/forums/showthread.php?t=3597&page=65

Look at dates of posts and then graphs.

Experts everywhere.

 

[LifeChoices]

Re: Half baked trading scheme

But is Comsec market depth actually live? Or 20 min delayed?

AFAIK it's live, when I do trades I see them turn up in the depth - so I'm guessing it must be. I would hate to be scraping old data.

 

[Tanaka]

You may be able to apply your system rationale with tick charts. For example, a 1000 tick chart with premium data will display up tick and down tick volume after the bar has been completed. I use this for futures ES and SPI. The issue here is determining whether the up tick volume is created by buy stops being hit on short positions as the market goes up, this can be a sign that the price is reaching exhaustion and ready to switch directions, probably caused by professionals stop hunting.
I don’t know if Aussie stocks include up and down volume in tick charts because I don’t subscribe to that data. If someone could let me know it would be appreciated.

 

[Country Lad]

Life, your system is not half baked, just quarter finished. Our trading group first developed a momentum system about 8 years ago and thanks to a couple of them being programmers, the system and the algorithms were written and tweaked over about another 18 month period until about 80% reliable for consistent gains

A few tips from the parameters we use.

Anything in the buy or sell queues more than 20% away from the current price should be ignored. Our system works on the combination of the queues up to 10% and 20% away from the last price. The rest is ignored as not having any relevance to the momentum.

Look at using the weighted price. By using only the number of shares, smaller parcels can in some instances have a skewing effect. Also, the bots testing prices will then have a lesser input.

Include the number of parcels bought and sold in the parameters. If anything this has a greater input into the momentum.

Make some allowance for the spread.

As an example, MNW probably has a high ratio from your calculations. However, the wide spread is a risk, the ratio for the prices above 20% of the last price is very low and the 10% doesn’t register because the gap between the buy and the sell is more than 10%. Also there were more sell trades than buy.

Hope that helps.

Cheers
Country Lad

 

[LifeChoices]

Life, your system is not half baked, just quarter finished. Our trading group first developed a momentum system about 8 years ago and thanks to a couple of them being programmers, the system and the algorithms were written and tweaked over about another 18 month period until about 80% reliable for consistent gains

A few tips from the parameters we use.

Anything in the buy or sell queues more than 20% away from the current price should be ignored. Our system works on the combination of the queues up to 10% and 20% away from the last price. The rest is ignored as not having any relevance to the momentum.

Look at using the weighted price. By using only the number of shares, smaller parcels can in some instances have a skewing effect. Also, the bots testing prices will then have a lesser input.

Include the number of parcels bought and sold in the parameters. If anything this has a greater input into the momentum.

Make some allowance for the spread.

As an example, MNW probably has a high ratio from your calculations. However, the wide spread is a risk, the ratio for the prices above 20% of the last price is very low and the 10% doesn’t register because the gap between the buy and the sell is more than 10%. Also there were more sell trades than buy.

Hope that helps.

Cheers
Country Lad

Is it too late to rename my thread to a less than quarter baked trading scheme? Some of you experts are all over me.

 

[LifeChoices]

I've only just got my alerts working and it looks as though ERA has turned the corner - there are now 15% more buyers than sellers. Although I would wait to buy this until I see a many more buyers come into it - more like 80% more buyers.

The other stock which I'm also watching has been GOR - Like MSB has strong buying support and has been going north strongly for some time.

 

[Noddy]

I've only just got my alerts working and it looks as though ERA has turned the corner - there are now 15% more buyers than sellers. Although I would wait to buy this until I see a many more buyers come into it - more like 80% more buyers.

The other stock which I'm also watching has been GOR - Like MSB has strong buying support and has been going north strongly for some time.

According to Commsec ERA has -
380,000 buy,
444,000 sell,
Where to now ?

Market depth changes so rapidly, it's not a very reliable indicator IMO.

 

[Synergy]

I'm going to watch FXJ with this thread in mind. It's been falling pretty fast, but is showing signs of a possible turnaround.

 

[ginar]

Market depth = smoke and mirrors most of the time . i think a more reliable approach is to guage who comes in on more volume at market price re sellers or buyers , i think tech alluded to this in an earlier post . to identify these trends so more sophisticated software is needed as normal course of sales and market depth makes this difficult . many years ago i used quotetracker for just this purpose and it does the job nicely . quite often im a contrarian of market depth , when i see one side stacked taking the opposite position isnt any worse than taking stacked side and at the right moments its statistically better . its all pretty discretionary for me so i cant help with any mechanical systems but check out the "real" ratio of sellers to buyers by observing trades that come of the screen that never hit depth .........