Pedro01

I am curious about one thing. When you copied the Jigsaw Trading products, why did you not even change the color scheme? You have it all - even down to the yellow/white highlighting.... I think it's a great compliment to Jigsaw that you were afraid to even change the colors, lest your might miss something!

He did say 'could' He could have said.... Big sell imbalance here, market will head down unless this is a trap for new shorts in which case we will see new highs. Or to put it another way. Could go down, could go up. Jus tryin' to inject a bit of humour....:haha:

Probably explain why the market had been heading down in the 11 minutes before your post. :doh:

My thoughts are - there are a lot of players and it's hard to imagine there's any collective opinion between them. What you can be certain is that there are a lot of short term players and they are alltrying to make money off everyone else. As such, there are points at which you can predict people will be in pain and points at which people will be 'twitchy'. I think that a good perspective of where those pain points could be, combined with tape reading when you get there can help. I am also sure that younger traders would not find tape reading as exhausting and so could do it for longer. But Automated Tape reading? Buy/Sell signals programmatically? I don't think I'd trust it.

The market is like a bucket. Lots of people put money in to the bucket. All of them want to take more money out of the bucket than they put in. This is physically impossible. Hence, whether you realise it or not, you are competing with people for that money. Unlike a bucket, the market also has brokers, back office staff, sales people, lap dancers, marketing etc. etc. etc pulling money out too. Why do 90% lose? How could it be any other way?

good point - many indicators can be put on a lower timeframe and the parameters tweaked to emulate a higher timeframe. Not all indicators can though - it depends on the type of calculations being used.

There's an overhead both in the extra time it takes to write/test the indicators as well as at execution time and the overhead is hughe. Saying that, I only use strategy code for testing a strategy, for real time execution, I use code with embedded macros which allows you to build in some redundancy. For instance, with macro code you can place a stop order as well as code to monitor the current price vs stop price. If you go through the stop price, you can place a market order as a 'backup' to your stop order. It may seem like overkill but I prefer to have that extra degree of control you get with the macros.

As per my other, lengthy post on this topic (in your other thread). Tradestation strategy code won't do this properly. Real time & backtested results will be different because of a 'quirk' in Tradestation. You can buy JAM HT code for synthetic bars to help or spend CONSIDERABLE time coding up synthetic bars in your strategy and then the indicators that go on top of them. You risk spending a lot of time working on a strategy whose backtested results are not based on what you think the entries are based on.

You can do this 2 ways. First and easiest is to put the 3 timeframes on 1 chart, then you don't need to use Global Variables. The Global Variable option can only be used in real time, you can't look at past data using that method. This is because TS loads up chart by chart, not minute by minute. You can use volume, ticks or time bars. If you are testing for direction, you have another issue to overcome... The chart has the e-mini 1 min (shown) and the e-mini 10 min (hidden) The first CCi indicator is on the 1 min The second CCi indicator is on the 10 min The last #PD_MTF_TEST indicator shows the direction of the CCi,red = down, green = up. The first row of dots is the 1 min, the second row of dots is the 10 min. So you can see that 7:20-7:30 the CCi 10 min was up slightly but this isn't shown 'in the dots' until the bar closes at 7:30. The bottom row of red dots showing direction of the CCi from 7:30 to 7:40 is based on it's direction from 7:20 to 7:30. Likewise, the bottom row of red dots from 7:40 to 7:50 shows red which is based on the CCi direction from 7:30 to 7:40. Whilst this is all just sample code to show the concept, what I was trying to do generally is use higher timeframes to give direction and lower timeframes for entry. This isn't much use if you can't get the 10 min direction as it's changing. Now - in real time this is different - you will see the CCi direction moving for the current 10 min bar as the bar is in progress. Even after the first minute of the 10 minute bar, if the strength of the move is enough to change the direction of the 10 minute bar, you use that immediately. The answer is in the use of synthetic bars. The second attachment shows the same indicator with a 'synthetic' version overlayed and how you can see the higher time frame indicator plotted out closer to the way it does in real time. Whilst a lot of people don't seem to be able to wrap their heads around this issue, what it means is that you you use these indicators in multiple timeframes in a strategy, your strategy entries & real-time entries will be different & the results you get will not be based on the entry technique you think you are coding.... Pete

Thanks for the replies to this - it looks like it may be the way forward for me. Just a shame that I end up paying full platform fees for Tradestation as I won't be using them as a broker. I guess I can bear that as I don't really want to go to Tradestation 2000i. Cheers Pete

by TS2Ki - I mean Tradestation 2000. I like the idea of being able to number the trades & then deal with each in order, it seems sensible. Would I have the same issue I have now though - would I need to wait for cancellation of my stop loss order before issuing a new sell order or could I send them all at the same time ?

Hi Paul It's probably easier to explain this if I show an example of the trades: When I get a singal, I place a trade. Then I wait for the entry place & place a safety stop loss. Initial Buy : .PlaceOrder "Action='Buy', Account='SIMxxxxxxM', Symbol='MNT', SymbolCategory='Equity', OrderType='Market', Quantity=560, Duration='Day'" stop loss: .PlaceOrder "Action='Sell', Account='SIMxxxxxxM', Symbol='MNT', SymbolCategory='Equity', OrderType='Stop Market', StopPrice=30.43, Quantity=560, Duration='Day'" So I have my buy & I have my stop loss. Now let's say I'm really lucky & my target gets hit - maybe 20-30 minutes later. What I want to do now (and you will recognise this) is to sell 75% of my position and put a stop loss on the remaining 25% at entry +2c. The problem is that I can't sell 75% of my position because I have an outstanding sell order on the stock (the stop loss). So - as far as I know, I need to cancel the stop loss order, then issue the sell order for 75% of the position and then put back the stop loss on the remaining amount. Cancel Stop Loss : .CancelAllOrdersSymbolAccount MNT,SIMxxxxxxM Scale Out : .PlaceOrder "Action='Sell', Account='SIMxxxxxxM', Symbol='MNT', SymbolCategory='Equity', OrderType='Market', Quantity=420, Duration='Day'" New stop loss : .PlaceOrder "Action='Sell', Account='SIMxxxxxxM', Symbol='MNT', SymbolCategory='Equity', OrderType='Stop Market', StopPrice=30.79, Quantity=140, Duration='Day'" This occasionally fails because I send the orders to the TS servers using the PlaceOrder macro but the cancel stop loss has sometimes not yet been processed by the time it gets the scale out. I can't seem to see any order type that will let me do this. I could actually send the cancel stop loss and then wait for it to be processed but I don't see any way to either see if the stop loss is still there OR get confirmation it has been removed. Without either of these, my only option is an arbritrary pause in the code of let's say 10 seconds to ensure the cancel stop loss has plenty of time to get processed. That's not something I want to do though - who wants to wait any amount of time to get out ? Anyway - I'm a bit stumped and fearing the worst ! Cheers Pete

Thanks for the post. Am I right in presuming they use TS2Ki with IB ? I do like the features of TS 8.3 BUT if I use tradelink to another broker, this does mean I'll end up paying for Tradestation as I won't be using them as a brokerage. Saying that, I could port my code back to TS2Ki but it does worry me that TS2Ki seems a lot less reliable than the later versions, I hear from friends all the time that their 'global server' program falls over all the time. Ideally, a proper trading API (with confirmations back) to the Tradestation brokerage would be ideal.

Hi All I have gotten to the point in my trading strategy automation where it feels as if Tradestation is less of a fit that it should be and I am wondering if this is just me being dumb or if TS isn't the tool for me. So a little on my strategy (without spilling the beans). part 1 - a radar screen with a list of stocks being rated (feeding info to global vars) as well as calculating position size for each stock considering my basket size, my account size, risk per trade & some stock specific items part 2 - a mutli time frame look at an index which gives me buy & sell signals When I get a buy or sell signal, I go to my global vars, look at my ratings & then place buys or sells on the radar screen stocks. I also set target prices for my stop loss & scale-out of a profitable trade. Global variables are slightly painful but if you handle all that in functions you can keep it consistent & away from the body of your code. What is a pain is the fact that you can't use strategies for this kind of system as you can't attach a strategy to radar screen, nor can you trade a basket of stocks in a strategy from a single chart. Note that I wouldn't want my strategy on the radar screen as I want my trade issued based on the action of the index, not wait for the next tick on the symbol to issue my order. So, I am using order macros and I am having to do the following: On placing a trade Store number of shares traded in GVs, then keep checking for fills. As you get filled, place stop loss orders at the appropriate levels as well as storing the average price. On scaling out Cancel your stop loss order Place a sell order for 75% of your position Put back the stop loss on the remaining 25% Of course, this doesn't actually work as the order to cancel the stop loss takes a while & if you send your sell order straight after that, it gets rejected as the stop loss is still there. You can't seem to queue orders and have them executed in sequence, nor is there any way to automatically scale out which to me is a fairly common thing to do. It seems the more I try to put in smart trade management into a strategy, the more I have to do myself. Note that I am obviously more of a coder than a trader and so it is entirely possible I have missed something obvious such as an order type that does exactly what I want. I guess there are 2 ways forward - scrap TS in favour of something with the features I need (and I have no idea what that is) or write my own/buy a better trade management API than the one provided with TS. Has anyone else encountered similar issues and resolved them ? Cheers Pete