Different approaches to breakeven with Renko

[arbuthnot]

Hi everyone

I’m in the process of preparing for my first foray into automated trading. I’ve done a lot of searching in the Forum on this and related subjects but I’d really appreciate some clarification.

I'm sorry for the length of this but I wish to cover all the options.

Firstly, three important points:

- I use Renko bars to enter trades

- I don’t have a very powerful PC.

- My Renko Strategy (to enter trades) will be CalculateOnBarClose = true.

I plan to use a stop, of course, but not a trailing stop, but do intend to have a ‘breakeven’ condition. I’m looking for the most efficient way of achieving this. (Setting stops and profit levels in Ninja strategies is very easy, it’s only the breakeven condition that is problematic.)

From my research in the Forum, there are various approaches I could take:

1) I could of course use techniques as exemplified in ‘SampleAtmStrategy’ to enter trades within my Renko Strategy controlled by ATMs. Certainly one of an ATM’s functionalities is to set a breakeven. However, it seems to me that, as I won’t be using trailing stops, this might be an inefficient approach, maybe ‘a sledgehammer to crack a nut’.

2) There’s the multi 'chart type' frame method: this would entail having Renko bars in one panel, and maybe with low-value tick bars in another panel. Trades would be entered using the Renko panel but worked using the tick panel.

3) There’s the OnMarketData method, but I know very little about this. (I only found out about this recently from thread: http://www.ninjatrader.com/support/f...d.php?p=388904.

4) There’s the option of letting the breakeven conditions be operated by the ‘Position’ methods (with only the Renko bars on the chart). However, with Renko bars, this will always end up being be worse than breakeven, simply by the definition of Renko. Just say the profit trigger is 4 on a 4 Renko, then if the next bar reverses, the theoretical position is back to ‘zero’ but with commission and slippage, a couple of points may well be lost, whereas with an efficient ‘breakeven’, one would hope to have no less at all. (So this approach is far from ideal.)

Given my situation, without a modern, fast PC, which of the above four would I be the best to go for? Is there possibly another approach I haven't thought of?

I also have these questions:

a) With a SampleAtmStrategy-type strategy entering an ATM, does this ATM work on a tick-by-tick basis even if the Strategy has CoBC = true?

b) Assuming the answer to a) is ‘yes’, can I assume that an ATM working on a tick-by-tick basis is much less CPU-intensive than a Strategy with CoBC = false (esp. if it’s quite complex)?

Many thanks in advance for any suggestions as to the best way of proceeding.

[NinjaTrader_PatrickH]

Hello arbuthnot,

Thank you for your post.

Whether you are adding an additional series as tick or running CalculateOnBarClose = False, the tick by tick stream is used. However, using the additional data series to submit the orders to and using CalculateOnBarClose = True would be helpful in backtesting, but in real-time all you would need to do is CalculateOnBarClose to True as the orders are always going to be tick-by-tick in real-time. So you could use your orders or the ATM strategy methods, either way they will be tick-by-tick in real-time.

You may consider reducing the amount of printing via Print() or logging via Log() to improve performance. If you strategy is calling indicators that add drawing objects to the chart you can look at removing the indicators from the Add() in Initialize() or commenting out the drawing in the Indicator code.

For performance tips please take a look at the following link: http://www.ninjatrader.com/support/h...ance_tips2.htm

Please let me know if I may be of further assistance.