Need help on ExRem and Scale-Out

Hi

Can someone help me please on exrem and scale-out? What is correct way to combine them? Is it to use exrem first before scale-out or scale-out first after that do the exrem?

I see very different result if I put ExRem before and after the scale-out.

And how to fix trade size limit of 10% of entry bar volume has been hit xx times?

Thanks in advance.

Hi @newbietrader,

You're asking a couple of different questions in the one post (exrem and "entry bar volume has been hit xx times"). They may be related, but look like different problems.

It appears that you haven't thought about order-flow, the logic of your code, and how ExRem() and scale-in/-out work, and how they all interact, and at what point in time.

As I understand it, ExREm() provides a way to remove signals from the Buy and Sell variables (which control when to enter/exit) Long trades. Similarly for Short trades.

Is it appropriate, in your trading system, to remove extraneous signals before, or after you've adjusted your position size with scale-in/scale-out?

Please refer example 4 given on the the following link:-https://www.amibroker.com/guide/h_pyramid.html

ExRem removes consecutive signals of same type. If you are long and want to Scaleout, you need to allow the sell signal trigger more than once. So ExRem not to be used.

The second question is not well understood, at least by me.

Hi

Thank you @sinecospi and @rvdesh for replying.

Is it appropriate, in your trading system, to remove extraneous signals before, or after you've adjusted your position size with scale-in/scale-out ?

I am still looking for the right combination, currently I already have a trading system (not many buy signals) and I am looking to improve it. After testing scale-out, it seems that the performance worst than without scale out. Not really sure if the placement of exrem before or after scale-out will play a role or not and there are many posts in this forum regarding exrem and scalein/scaleout not being answered too. It seems scaling out trade is not good for trading system?

scaleout-noscaleout

@newbietrader,

In regard to:

there are many posts in this forum regarding exrem and scalein/scaleout not being answered too

Exrem and scale-in/out need to be considered in relation to the strategy, which is probably why there aren't simple answers on how to use them.

There are lots of very simple/complex trading systems around, but developing one that has an "edge" takes time, and a good understanding of how orders flow through the market. Developing a trading strategy in AmiBroker, or any other software, means you need to embed the logic of order flow into your strategy.

Backtesting (as you you seem to be doing) is great, because it gives you a sense of how a strategy might perform, but getting it "wrong", or getting the logic out of sync, means that you end up fooling yourself.

In regard to the strategy you've already got, some questions for you to consider:

  • What's wrong with the performance of the current strategy that makes you want to (force) trade it more often? Superficially, it looks like you've already got a reasonable strategy, but you haven't indicated why it's unacceptable?
  • Why do you want to increase the number of trades?
    • When you do that, you might be introducing unintended conditions which affect the strategy's performance? For example, drawdown might be unacceptable.
    • Is it to increase the overall size of the profit?
      • Why not just increase the position size, or,
      • Why not apply it to other tradeable/s which behave similarly to the ones you've already applied the strategy to.
  • What are you willing to sacrifice for those additional trades? eg your pic already shows what happens when you go for quantity over quality.

Looking at the stats in your pic, suggests a couple of things:

  • "scale-out" causes an existing trade to exit earlier than it might otherwise, enabling subsequent entry signals to be executed:
    • have a look at your settings for entering a new trade when there's an existing trade
    • have a look at "scale-in" when there's a new entry signal whilst you already have an open position - this'll increase your position size, take advantage of all those extra signals you wanted to get rid of with "exrem", and in the process, "trade more often", but there'll be +/- consequences

Perhaps you need to work out what's important to you - size of profit, or reliability. What's "important" will then influence when/how you use exrem/scale.

5 Likes

Hi

Thank you for the great advice, @sinecospi . I feel like I miss a lot of opportunity when many other traders are showing profit and I am stuck with my position because I am holding too long sometime and I thought maybe by having more trades, I can improve my trading result. I have limited capital and sometimes it is already maxed out.

I tried to introduce tighter stoploss if you read my previous thread, where I tried to insert 2 trailing stops. As you said, I think scale-in (average up) is better than scale-out (take profit early or holding to position when price against my position hoping it will reverse) in my case.

That's Ok @newbietrader,

I'm sure there's plenty of people in the same situation as yourself - when you're just starting out, there's so much to learn, and you're itching to trade, as very dear old friend used to say "make haste slowly".

Do some research on stop losses - how to use them depends on your strategy, and the reason for using them. Keep "playing" with them in your backtesting, to understand how they affect the outcome.

Scale-in may be a way to get the "extra trades" that you're looking for, but test it out first - the performance may actually suffer.

Perhaps you should also have a look at the various performance metrics available within AB, to help you gauge what's good/better performance.

Hi

After I tested out scale-ins, it seems that if I do many scale-ins also not good for the backtest result, not sure maybe my money management (position sizing) is not right. And during the quest to find out about scale-ins/outs, I found this post from Mr Howard B.

from-howard-yahoo-group