Do you need to understand where the edge comes from?

I believe the answer is Yes and No.
I generate millions of strategies and would only trade those that I “understand”.
At minimum this is the only way to find bugs in the code and/or in the data. Once i cannot find anymore bugs I realize that some strategies are a type of statistical arbitrage, others are trend-following, others are scalping, others are based on volatility, etc.
But, for example, I’m not sure if there is anyone who can say they fully understand why and how MACD or trend-following works, and whether it will continue working tomorrow. Some people say that it works because everyone else is using it so they amplify the initial trend that might be showing up. Others call bs and say that trends work on their own merits. Others don’t believe in trends at all. While multi-billion dollar hedge funds trade them.
So even though some people may say they understand why their strategy works, others can easily dispute this and “prove” that they don’t understand shit.
 
Last edited:
guru said:
I believe the answer is Yes and No.
I generate millions of strategies and would only trade those that I “understand”.
At minimum this is the only way to find bugs in the code and/or in the data. Once i cannot find anymore bugs I realize that some strategies are a type of statistical arbitrage, others are trend-following, others are scalping, others are based on volatility, etc.
But, for example, I’m not sure if there is anyone who can say they fully understand why and how MACD or trend-following works, and whether it will continue working tomorrow. Some people say that it works because everyone else is using it so they amplify the initial trend that might be showing up. Others call bs and say that trends work on their own merits. Others don’t believe in trends at all. While multi-billion dollar hedge funds trade them.
So even though some people may say they understand why their strategy works, others can easily dispute this and “prove” that they don’t understand shit.
More...


MACD and indicators work up to viewing 'inside the spread.' They are derivatives of the movement of the basic granularity of price and volume. It's basic math. They can be transformed into 'leading indicators' but only as you suggest is based upon something one 'understands.'

That's why they lag market 'turning points' on faster timeframes whereas these micro-turning points are 'smoothed' for those trading on slower timeframes for they are perceived at that level of observation as 'noise.'

As for the differences of opinion, there are exclusive and inclusive views. One contains the other, not vis-a-versa, it is a one way gate. The inclusive pov comes from a fully differentiated spectrum developed in the mind and emotional body that allows perception greater than the pov based on exclusion. One can tell the operating point of another by the degree of argumentation and attachment to 'being right' at the cost of an open mind.

One can observe a 'close-minded' person existing in their 'pain-oriented' emotional body by emotional responses of 'bullshit!' and other forms of expressed profanity, lack of logical reasoning and lack of respectful dialogue. For them, information outside of their 'unconscious morass' is threatening to their attachment of their chosen identity.
 
Sprout said:
They can be transformed into 'leading indicators' but only as you suggest is based upon something one 'understands.'
More...

Ummmm, not to be contentious, Sprout, but short of sending 1.21 gigawatts into the ol' Flux Capacitor, there is nothing in construction or in theory that allows anything in T/A to be termed 'leading indicator' -- outside of our own, rich, imagination(s).


The closest we can get is 'Coincident', but even that is subject to no small amount of what we might call Zeno's Paradoxism (e.g., if you're always 'halfway there' then you're never able to be *there*.) "Solvitur Ambulando" muttered Diogenes The Cynic, who stood, walked over and touched the philosopher on the shoulder, [dropping his figurative microphone :wtf:], and walked out. :rolleyes:

Ahhhhh. Ya gotta love a Sunday. :D
 
Last edited:
tommcginnis,

I have a question for you...by the way thanks for your posts.

Can predictions be made?

If a pattern repeats itself over and over for the last 10 years and repeats 90% of the time...is that tradeable?...What I mean is High, Low & Close. I am over simplifying this but the idea is does a pattern repeat itself?

ES

P.S. You can simply look at a previous candlestick (s) and see the midpoint and throw a dart and make an long or short entry at MOO and exit at MOC or try to use data showing patterns that repeat.


tommcginnis said:
Ummmm, not to be contentious, Sprout, but short of sending 1.21 gigawatts into the ol' Flux Capacitor, there is nothing in construction or in theory that allows anything in T/A to be termed 'leading indicator' -- outside of our own, rich, imagination(s).


The closest we can get is 'Coincident', but even that is subject to no small amount of what we might call Zeno's Paradoxism (e.g., if you're always 'halfway there' then you're never able to be *there*.) "Solvitur Ambulando" muttered Diogenes The Cynic, who stood, walked over and touched the philosopher on the shoulder, [dropping his figurative microphone :wtf:], and walked out. :rolleyes:

Ahhhhh. Ya gotta love a Sunday. :D
More...
 
ElectricSavant said:
If a pattern repeats itself over and over for the last 10 years and repeats 90% of the time...is that tradeable?...What I mean is High, Low & Close. I am over simplifying this but the idea is does a pattern repeat itself?
More...
ES, I see you asking the same questions over and over about patterns. No two patterns are the same. Why have you not quantified your "patterns" and simply back tested them? Won't that answer all your questions, and hopefully allow you to move on too something more productive?
 
tommcginnis said:
Ummmm, not to be contentious, Sprout, but short of sending 1.21 gigawatts into the ol' Flux Capacitor, there is nothing in construction or in theory that allows anything in T/A to be termed 'leading indicator' -- outside of our own, rich, imagination(s).
More...

Perhaps not exactly correct, search on negative group delay. Can be used in TA indicators based on digital filters (i.e. lowpass, highpass, bandpass.)
 
I can pm Tom if that would be easier for you?


TheBigShort said:
ES, I see you asking the same questions over and over about patterns. No two patterns are the same. Why have you not quantified your "patterns" and simply back tested them? Won't that answer all your questions, and hopefully allow you to move on too something more productive?
More...
 
qlai said:
For people who develop and test their strategies, do you try to understand why it works? Fo example, your strategy shows great results and when you dig down you realize it was buying dips. So you say to yourself - great I have a good strategy to deploy in a strong bull market, but keep it on a shelf otherwise.
What I don't understand is how people use ML etc to get seemingly random strategies. How will they trade it without understanding how/why it works? How will they know when to stop? Are they just trading a basket of uncorrelated strategies? Would be interested in how people on ET approach this.
More...
Prudent Risk Management is the the only true edge in trading.
 
You must log in or register to reply here.
Back
Top