TheBramble
Legendary member
- Messages
- 8,394
- Likes
- 1,170
There are a whole bunch of reasons why this wouldn't work (and I've addressed a few of these issues toward the end), but I'd appreciate your comments.
Although Grey1 (our guru and forum founder) is still in hiding, the Pairs Trading and Directional Trading posts/trading room sessions etc. covered such a lot of new ground (for me) and introduced so many exciting trading possibilities that I'm keen to develop them further. When I say, 'I' of course mean 'we' - hence the title of this topic.
One of the biggest problems when live trading was to have suitable stock candidates identified in a timely enough manner for a trading decision, strategy and action to be taken. This included working out the risk:reward, ratio of shares (for pairs), stoploss etc. This is in addition to working out the other basic metrics of the stocks themselves, such as daily historical volume, price range, historical daily range etc.
By far the biggest of all of these was the impossible number of stocks to be monitored.
On other threads recently, there has been discussion of this data overload problem and there have been some useful suggestions, but all in their early stages of development.
Breaking down the work into chunks (and this is good point to wake up from my boring intro and be ready to pull this apart...) we need:-
1. A weekly review of all stocks that meet the basic criteria for consideration in trading. And these criteria are:-
A. Not massively high volatility - the 'screamers' - we don't need them.
B. Trade in the range of $20-60
C. Have a historical daily trading volume in excess of 1.5 million
D. Have a historical daily trading range in excess of $1
2. A daily over-ride on stocks 'in the news' (same category as the 'screamers' really).
For those stocks identified in (1) and excluded in (2) we need from the live datafeed and on-board scanner processing logic (whatever that may be), the volume, price, vwap, MPD+/- and 52wk HILO.
You already know all of that.
But how do we address the problem of dealing with the thousands (please jump in again if I'm way off beam here...would there be thousands meeting these criteria?) of stocks meeting our criteria?
As Trader333 pointed out recently, even if one individual had the processor horsepower to churn through all of them, the datafeed supplier may well object.
This is where my idea of a geographically distributed arcade comes in.
On any one day, each of us has conflicting time constraints and non-trading responsibilities. Even us full-time traders aren't going to trade all day of every day, or even all day of any day we do trade. However, on any given day, there will be a core of traders who will/can be available for the prime US trading periods (first couple and last couple of hours) on any given day.
My thought is to split the X number of stocks between the Y number of available traders each day and each scans just their own allocation. Obviously on days where the number of stocks is too large to manage for the number of active traders available - we simply go back to what we're doing at the moment. I'm guessing 100 stocks per trader would be max.
The scanner would need to be common to all (or as common as possible) in having a DDE link to say EXCEL or something like that. Someone would need the skills and time to write this. Those that couldn’t use whatever general scanner comes out of the exercise would need to be able to provide the same scanner facilities. We need not only the potential trade stock candidates, but the risk:reward, ratio (for pairs) and stoploss – these need to be automatically calculated and then put out for all the other traders.
For those traders collaborating on any given day, they are linked via a live trading room (or any other suitable mechanism?) and alert the other traders when stock candidates are being alerted for potential trading opportunities.
Each trader then pulls up the candidate(s) onto their own RT charting system and plays them (or not) they way they want (but I'm guessing along previous vwap etc lines).
Apart from using these vwap techniques again, we’d be experiencing and developing the strategy more fully. This seems like a good thing to do.
We all get to benefit from each others expertise in reading LII, reading the volume, timing, execution, reasons why we did/didn’t go with a specific trade etc.
Reasons why this wouldn't work? (and some possible solutions?)
1. Live Trading rooms require someone to divert their attention from trading to telling others about stock candidates.
In the past the show was very much Grey1's and his trading suffered as a result. By spreading the workload over N traders in any one day, there will be 1/Nth less workload required of each.
2. The flux of different traders on a daily basis and allocation of stocks to scan might be a big issue. I don't know how long it would take each of us to completely re-load our scanners with new data each day.
However, if using a common EXCEL based solution, this really should be no more than a simple cut & paste.
3. Dependence upon someone else 'doing their bit'.
No way round this one, but to date, I think most traders in the live trading rooms have been pretty much clued up and ready to work!
4. The management overhead in handling this exercise.
All it would take is one person to arrange a schedule of who’s up for trading the next day and on the actual trading day, making any last minutes changes before parcelling out the stocks for each to monitor.
5. Won't be able to concentrate on my current trading activities, systems etc.
If the action is so hot that this becomes true - why would you care - you'd be trading vwap? And on days where there is no action, you can still play your own strategies. And even on days in the past when there was some good vwap trading, there were an awful lot of gaps too. Plenty of time to do other stuff.
When all's said and done, even if the system didn’t work every day and had some days where there weren't enough players, it's still better than no advancement at all in developing these very valuable trading techniques introduced by Grey1.
What do you think?
Although Grey1 (our guru and forum founder) is still in hiding, the Pairs Trading and Directional Trading posts/trading room sessions etc. covered such a lot of new ground (for me) and introduced so many exciting trading possibilities that I'm keen to develop them further. When I say, 'I' of course mean 'we' - hence the title of this topic.
One of the biggest problems when live trading was to have suitable stock candidates identified in a timely enough manner for a trading decision, strategy and action to be taken. This included working out the risk:reward, ratio of shares (for pairs), stoploss etc. This is in addition to working out the other basic metrics of the stocks themselves, such as daily historical volume, price range, historical daily range etc.
By far the biggest of all of these was the impossible number of stocks to be monitored.
On other threads recently, there has been discussion of this data overload problem and there have been some useful suggestions, but all in their early stages of development.
Breaking down the work into chunks (and this is good point to wake up from my boring intro and be ready to pull this apart...) we need:-
1. A weekly review of all stocks that meet the basic criteria for consideration in trading. And these criteria are:-
A. Not massively high volatility - the 'screamers' - we don't need them.
B. Trade in the range of $20-60
C. Have a historical daily trading volume in excess of 1.5 million
D. Have a historical daily trading range in excess of $1
2. A daily over-ride on stocks 'in the news' (same category as the 'screamers' really).
For those stocks identified in (1) and excluded in (2) we need from the live datafeed and on-board scanner processing logic (whatever that may be), the volume, price, vwap, MPD+/- and 52wk HILO.
You already know all of that.
But how do we address the problem of dealing with the thousands (please jump in again if I'm way off beam here...would there be thousands meeting these criteria?) of stocks meeting our criteria?
As Trader333 pointed out recently, even if one individual had the processor horsepower to churn through all of them, the datafeed supplier may well object.
This is where my idea of a geographically distributed arcade comes in.
On any one day, each of us has conflicting time constraints and non-trading responsibilities. Even us full-time traders aren't going to trade all day of every day, or even all day of any day we do trade. However, on any given day, there will be a core of traders who will/can be available for the prime US trading periods (first couple and last couple of hours) on any given day.
My thought is to split the X number of stocks between the Y number of available traders each day and each scans just their own allocation. Obviously on days where the number of stocks is too large to manage for the number of active traders available - we simply go back to what we're doing at the moment. I'm guessing 100 stocks per trader would be max.
The scanner would need to be common to all (or as common as possible) in having a DDE link to say EXCEL or something like that. Someone would need the skills and time to write this. Those that couldn’t use whatever general scanner comes out of the exercise would need to be able to provide the same scanner facilities. We need not only the potential trade stock candidates, but the risk:reward, ratio (for pairs) and stoploss – these need to be automatically calculated and then put out for all the other traders.
For those traders collaborating on any given day, they are linked via a live trading room (or any other suitable mechanism?) and alert the other traders when stock candidates are being alerted for potential trading opportunities.
Each trader then pulls up the candidate(s) onto their own RT charting system and plays them (or not) they way they want (but I'm guessing along previous vwap etc lines).
Apart from using these vwap techniques again, we’d be experiencing and developing the strategy more fully. This seems like a good thing to do.
We all get to benefit from each others expertise in reading LII, reading the volume, timing, execution, reasons why we did/didn’t go with a specific trade etc.
Reasons why this wouldn't work? (and some possible solutions?)
1. Live Trading rooms require someone to divert their attention from trading to telling others about stock candidates.
In the past the show was very much Grey1's and his trading suffered as a result. By spreading the workload over N traders in any one day, there will be 1/Nth less workload required of each.
2. The flux of different traders on a daily basis and allocation of stocks to scan might be a big issue. I don't know how long it would take each of us to completely re-load our scanners with new data each day.
However, if using a common EXCEL based solution, this really should be no more than a simple cut & paste.
3. Dependence upon someone else 'doing their bit'.
No way round this one, but to date, I think most traders in the live trading rooms have been pretty much clued up and ready to work!
4. The management overhead in handling this exercise.
All it would take is one person to arrange a schedule of who’s up for trading the next day and on the actual trading day, making any last minutes changes before parcelling out the stocks for each to monitor.
5. Won't be able to concentrate on my current trading activities, systems etc.
If the action is so hot that this becomes true - why would you care - you'd be trading vwap? And on days where there is no action, you can still play your own strategies. And even on days in the past when there was some good vwap trading, there were an awful lot of gaps too. Plenty of time to do other stuff.
When all's said and done, even if the system didn’t work every day and had some days where there weren't enough players, it's still better than no advancement at all in developing these very valuable trading techniques introduced by Grey1.
What do you think?
