What are the questions you ask yourself in order to eliminate false positives?
- Thread starter rin4et
- Start date
My win rate is only 46% so by your definition I had a lot of false positives. But that comes with the territory and is a byproduct of the way I trade options.
I am actually not upset with my "false positives" so am not trying to eliminate them. Rather, I am looking at ways to increase my alpha and reduce my beta.
L
lovethetrade
Some people need to research the definition of a false positive.
Here, I'll help you
false positive
noun
How can a signal be a false positive when the conditions were True?
Here, I'll help you
false positive
noun
- a test result which wrongly indicates that a particular condition or attribute is present.
"stress testing, a routine diagnostic tool used in detecting heart disease, results in a significant number of false positives in women"
How can a signal be a false positive when the conditions were True?
Yep, a signal is a signal...some trades work and some don't.
Truth tables are essential for processing multiple variables in a dataset. The goal being one of producing a singular true/false result. There's a bit of semantic confusion in calling different distinctions by the same name.
If one tracks 10 variables, then the expression of those variables when evaluated as a set produces a single true/false result.
This is what I would define as a signal. Each variable in and of themselves are not. Lot's of confusion arises when interpreting variables as signals.
An effective methodology gives clear go/no-go signals, based upon a collection of a dataset.
The quality of the signal is confirmed after the entry.
The quality of the trade is confirmed after the exit.
The quality of the methodology is confirmed by the pnl.
One can manage an strong signal into a poor trade.
One can also manage a weak signal into an awesome trade.
What one cannot manage is a poor methodology that has no feedback structure to make it a better one.
If one tracks 10 variables, then the expression of those variables when evaluated as a set produces a single true/false result.
This is what I would define as a signal. Each variable in and of themselves are not. Lot's of confusion arises when interpreting variables as signals.
An effective methodology gives clear go/no-go signals, based upon a collection of a dataset.
The quality of the signal is confirmed after the entry.
The quality of the trade is confirmed after the exit.
The quality of the methodology is confirmed by the pnl.
One can manage an strong signal into a poor trade.
One can also manage a weak signal into an awesome trade.
What one cannot manage is a poor methodology that has no feedback structure to make it a better one.
Personally I have four rules-based conditions that need to be satisfied before I enter a trade. I do use a highly refined TA indicator package but the other three rules are by design and intent not data sampling oriented. This is the point which I have arrived at over the course of a couple decades of continuous work.
The most common mistake I see is the gross negligent misapplication of technical studies. To be more specific - TA 'group think'. It is a mistake to confirm one technical study signal with another study of the same genre - you are confirming the study and not the market price action behavior. It's akin to tolerance stacking in manufacturing. I've seen guys confirm an oscillator with another type of oscillator - yep, an RSI and a MACD are both oscillators (and both momentum studies as well). You'd be silly to confirm a Stochastic study with S/R levels. You would be much more correct to confirm, let's say, that MACD study with for example a volatility based study like a Bollinger Band or a Keltner Channel. You are sampling price data with completely different methodologies and therefore the confirmation would be more legitimate in terms of price action.
I wish everyone good fortune with their trading! There is no one correct way to trade markets. If it works and it's consistent over a protracted period of time then ride that horse until it falls dead.
The most common mistake I see is the gross negligent misapplication of technical studies. To be more specific - TA 'group think'. It is a mistake to confirm one technical study signal with another study of the same genre - you are confirming the study and not the market price action behavior. It's akin to tolerance stacking in manufacturing. I've seen guys confirm an oscillator with another type of oscillator - yep, an RSI and a MACD are both oscillators (and both momentum studies as well). You'd be silly to confirm a Stochastic study with S/R levels. You would be much more correct to confirm, let's say, that MACD study with for example a volatility based study like a Bollinger Band or a Keltner Channel. You are sampling price data with completely different methodologies and therefore the confirmation would be more legitimate in terms of price action.
I wish everyone good fortune with their trading! There is no one correct way to trade markets. If it works and it's consistent over a protracted period of time then ride that horse until it falls dead.
It seems majority of people here trade with one single signal, "BUY or SELL" lol.
Fine if it works for them, but i consider markets and trading quite more complex than just buy when one signal tells to do it. We are not robots, thanks God we got a brain, gut feeling and market experience that no robots can replace. Markets are not Chess or Black Jack.
So consider (or imagine) you got multiple signals. For example, one signal for valuation, one for momentum, one for price action and one for trend. Just an example.
If two signals are positive, and two negative, should I take the trade or not?
That s why I am waiting for a confirmation. Because most of the time, signals are mixed.
Of course sometimes all signals are positive or negative, this is called buying or selling opportunities. But markets are not that generous to frequently give you these high reward risks trades.
CM
Ahh ok I see what you are getting at now. I was confused why you wouldn't trade a signal but it is because you are evaluating more than one simultaneously. Thats where we are different. I prefer preparing as best I can beforehand so that the closer I get to the signal, the less Im thinking.
L
lovethetrade
I recommend you cut your losses before you embarrass yourself any further. if you have other external conditions or signals that affect your decision-making that should be embodied with your logic or processes to derive your single signal. You're clutching at straws now; accept when you're wrong, learn and move on.
Last edited by a moderator: