It's the timeframe...gaddammit

[They]

What about life span (instead of time frame) of trade assuming your bet is small with no stop loss (meaning you can let it go against for a wide max pain point)? If you trade both long and short positions, then at some point in time an opportunity will present itself to profit. it may be one day or 10 years

Ok, you have an idea, a feeling, a simple or complex indicator that tells you to enter a trade.
Whatever the reason is you can test its validity by checking MFE (maximum favorable excursion) and MAE (maximum adverse excursion) results. You can create almost unlimited time or volume segmentations to test over - 1 minute, 1 hour, 1 day, 1 week, 1 month, etc or via volume/shares traded.

Favorable moves vs adverse moves from your entrance need to be statistically significant across some time or volume traded period otherwise your entrance signal is just random.

 

[userque]

Fair enough, but I don't really care what anyone else says. Not because I want to be ignorant or want to be a jerk it's because I've verified what I am saying myself and seen the evidence.

So even if my theory is wrong and the markets aren't fractal, the end result is my signals work on every single time frame(seconds to monthlys), chart type(renko,range,kase,volume,tick,momentum), the only difference is the ATR adjustment.

So, overall I can concede in may be a debate, have fun debating it, but the overall concept is not a debate for me personally, since I have actual proof.

I simply just responded to your post, and like you, I'm also not interesting in debating it. I'm all for people believing whatever they want, and trading however they want. It's their mind and money.

My post was more for the observers, than you.

 

[trdes]

I simply just responded to your post, and like you, I'm also not interesting in debating it. I'm all for people believing whatever they want, and trading however they want. It's their mind and money.

My post was more for the observers, than you.

Maybe a moot point, so if so I'll let it go, but I just want to clarify I don't believe it, I know it to be a fact. Via 40 years of price action, along with multiple years of forward / live testing as well.

 

[Poljot]

Guys, would you be trading smaller TF against higher TF?
For example: current H1 bar is bullish - do you sell 15M bars?
I have not backtested it but seems like that would be a low probability bet because it means going against the trend.

 

[athlonmank8]

I day trade the Forex (and a few highly liquid Futures products) on the 15-min. I truly believed that the higher time frames (the 1h for instance) could help me "stay with the trend" and improve my trading results.

They did not, at all. Yes, they eliminated a few losing trades here and there, but I also missed a lot of big trends that way.

There are a lot of trading ideas that sound cool and logic, like using a higher time frame for confirmation or trend direction, but when you really put them to the test they fail miserably.

That's why I never take any "brilliant" trading idea for granted anymore.

In fact, my most profitable systems are totally counterintuitive, like using reversal strategies to ...stay with the trend! (Yes I know, sounds crazy )

Yup same. Doesn't sound crazy. Sounds like we trade the same way.

 

[userque]

Maybe a moot point, so if so I'll let it go, but I just want to clarify I don't believe it, I know it to be a fact. Via 40 years of price action, along with multiple years of forward / live testing as well.

I also want to clarify.

You started by making a blanket statement that a strategy should work on all time frames, because the markets are fractal.

I responded by quoting the man who made that hypothesis, and where he clarified how it doesn't apply. I also used logic and examples to show that the markets aren't fractals.

Now, as usual with these types of discussions, you moved the goal post and responded that you've been trading for decades.

My responses are, again:

1. The markets aren't fractals due to the gaps in time and days, and it's non continuous intraday distributions, that can't/don't repeat on daily bars. Same with daily to weekly bars, and so on. These 'seasonalities' are different in each time frame, preventing the markets from 'looking the same' in every time frame--as required in order to be fractal.

You seemed to agree by saying "fair enough."

2. A strategy that works in one time frame, may not work in another.

I gave an example. This point was to rebut your blanket statement that all strategies should do as all your strategies have done for 40 years, and work in all time frames.

Note: No where did I say you were, or weren't profitable. My responses had, and have nothing to do with your particular strategy. I have no idea what your strategy is.

Now, I'll add what is hopefully a final point, in response to your last post:

3. Even though the markets aren't fractal, that doesn't mean that people still can't believe that they are and still trade profitably. Some believe the Earth is flat, and still manage through life just fine.

 

[trdes]

I also want to clarify.

You started by making a blanket statement that a strategy should work on all time frames, because the markets are fractal.

I responded by quoting the man who made that hypothesis, and where he clarified how it doesn't apply. I also used logic and examples to show that the markets aren't fractals.

Now, as usual with these types of discussions, you moved the goal post and responded that you've been trading for decades.

My responses are, again:

1. The markets aren't fractals due to the gaps in time and days, and it's non continuous intraday distributions, that can't/don't repeat on daily bars. Same with daily to weekly bars, and so on. These 'seasonalities' are different in each time frame, preventing the markets from 'looking the same' in every time frame--as required in order to be fractal.

You seemed to agree by saying "fair enough."

2. A strategy that works in one time frame, may not work in another.

I gave an example. This point was to rebut your blanket statement that all strategies should do as all your strategies have done for 40 years, and work in all time frames.

Note: No where did I say you were, or weren't profitable. My responses had, and have nothing to do with your particular strategy. I have no idea what your strategy is.

Now, I'll add what is hopefully a final point, in response to your last post:

3. Even though the markets aren't fractal, that doesn't mean that people still can't believe that they are and still trade profitably. Some believe the Earth is flat, and still manage through life just fine.

I apologize if I have poorly chosen my word selection. But, I never claimed to have traded for decades. Do not think you are intentionally misusing my words or maybe you misunderstood, let's even say I was unclear but I never stated that. I was speaking of back testing data as well as forward / live testing.

Any signal I've ever seen that has an actual positive expectancy whether via brute force R/R, win rate or both. Works the same on every time frame. Again, that was my intention and main point to tell the OP time frame doesn't matter (other than his own personal bias or issues making him trade a certain TF better or worst).

This has been my experience. If you're telling me you know someone or you yourself has a strategy that only works on one time frame, that is pretty impressive and surprising.

For example if someone says there setup only works on a 5 minute chart, I than ask what about a 4 minute and 59 second chart, or a 4 minute and 58 second chart and etc. At what point does the strategy suddenly no longer become effective?

I certainly don't trade weekly, monthly charts and etc, that's a little unrealistic for me personally, so can't claim I've live traded those. None the less via backtesting the same concepts apply there as they do to lesser time frames.

 

[userque]

I apologize if I have poorly chosen my word selection. But, I never claimed to have traded for decades. Do not think you are intentionally misusing my words or maybe you misunderstood, let's even say I was unclear but I never stated that. I was speaking of back testing data as well as forward / live testing.

Any signal I've ever seen that has an actual positive expectancy whether via brute force R/R, win rate or both. Works the same on every time frame. Again, that was my intention and main point to tell the OP time frame doesn't matter (other than his own personal bias or issues making him trade a certain TF better or worst).

This has been my experience. If you're telling me you know someone or you yourself has a strategy that only works on one time frame, that is pretty impressive and surprising.

For example if someone says there setup only works on a 5 minute chart, I than ask what about a 4 minute and 59 second chart, or a 4 minute and 58 second chart and etc. At what point does the strategy suddenly no longer become effective?

I certainly don't trade weekly, monthly charts and etc, that's a little unrealistic for me personally, so can't claim I've live traded those. None the less via backtesting the same concepts apply there as they do to lesser time frames.

I apologize if I misunderstood, sounds like I did--but I can't get motivated enough to go check. I don't think that particular point matters to the crux of the discussion.

But I do want to clarify, I never wrote about strategies working in only one time frame. I said something like:

A strategy doesn't have to work in all time frames (1, 5, 10, 15, 30, 60 min., for example). IOW:

A strategy can work in more than one time frame, but there is no realistic requirement, or expectation that (good?) strategies should work in all those time frames due to fractals.

To be clearer:

I learned long ago about making blanket statements (always, never, etc.).

You made a blanket statement in response to a poster, claiming that their strategy should work in all time frames due to fractals; also insinuating that this assertion applied to "all" strategies.

I had/have two main rebuttals:
1. The markets aren't fractals
2. A strategy can work in one time frame, and not another.

I have not suggested:
3. That strategies work in only one time frame.

There is a huge difference between what item #2 is saying, and what item #3 is saying.

 

[trdes]

I apologize if I misunderstood, sounds like I did--but I can't get motivated enough to go check. I don't think that particular point matters to the crux of the discussion.

But I do want to clarify, I never wrote about strategies working in only one time frame. I said something like:

A strategy doesn't have to work in all time frames (1, 5, 10, 15, 30, 60 min., for example). IOW:

A strategy can work in more than one time frame, but there is no realistic requirement, or expectation that (good?) strategies should work in all those time frames due to fractals.

To be clearer:

I learned long ago about making blanket statements (always, never, etc.).

You made a blanket statement in response to a poster, claiming that their strategy should work in all time frames due to fractals; also insinuating that this assertion applied to "all" strategies.

I had/have two main rebuttals:
1. The markets aren't fractals
2. A strategy can work in one time frame, and not another.

I have not suggested:
3. That strategies work in only one time frame.

There is a huge difference between what item #2 is saying, and what item #3 is saying.

That is fine. Every strategy I have works the same on every time frame I trade, the only adjustment is the ATR. That's my personal experience and I haven't seen an example to the contrary.

I don't understand why you can't just give me an actual example. You're making an absolute statement now as well, but not clarifying an example.

Do you have a strategy that works on 5,10,15,20 and 25 minute time frame, but on the 30 minute time frame it doesn't work? Do you have an actual example that you can state through personal experience? Or are you just not willing to do that?

 

[userque]

That is fine. Every strategy I have works the same on every time frame I trade, the only adjustment is the ATR. That's my personal experience and I haven't seen an example to the contrary.

I don't understand why you can't just give me an actual example. You're making an absolute statement now as well, but not clarifying an example.

Do you have a strategy that works on 5,10,15,20 and 25 minute time frame, but on the 30 minute time frame it doesn't work? Do you have an actual example that you can state through personal experience? Or are you just not willing to do that?

I'm not making a blanket statement. A blanket statement is, again, using all, always, never, etc.

My statement is telling you that there are exceptions to your blanket statement.

I've seen strategies during testing where changing the time frame can cause a winning strategy to lose. This is extremely easy to do with NT, so I'm sure others have seen this as well.

No, I didn't file them away for times like this. And I can't get motivated enough to create an example for you. I refer you to earlier in our discussion where we both proclaimed how little we care about others believing us.

Finally, you've indicated that you make adjustments for the different time frames, so technically, you aren't running the same strategy. You've mentioned this previously, I didn't reply to it then, as I didn't think this discussion would continue as it has.

This is fine (ATR adjustments). But maybe you should have made that caveat known when you first made your blanket statement to the OP; maybe saying something similar to this, if not in these exact words:

"Hey OP, try adjusting the metrics in your strategy based on the ATR, since the ATR will change with different time frames--and affect you results. I've never seen a strategy that can't work on all time frames, as long as you adjust for the change in the ATR."

Here is my new response updated with the new goal post:

I had/have two main rebuttals:

1. The markets aren't fractals
2. A strategy can work in one time frame, and not another.

I have not suggested:

3. That strategies work in only one time frame.

I agree:

4. It is likely that a strategy can work in all time frames, if it is adjusted for each time frame.