[QUOTE=“Jezzode;486391”]
We don’t - but we do know that they are “robust” rules[/QUOTE]
How do we know they are robust though… When we are looking at profitability to make that determination.
[QUOTE=“Jezzode;486391”]
We don’t - but we do know that they are “robust” rules[/QUOTE]
How do we know they are robust though… When we are looking at profitability to make that determination.
Sometimes it feels that way, doesn’t it?
Aim: Not to be hit by car while crossing road
Rule 1: Have eyes open when crossing road
Rule 2: Close eyes and hope you don’t get hit
In both instances you could get hit by a crazy driver, but i’m sure we could all prove “Rule 1” to be more “robust” when it comes to staying alive.
[QUOTE=“Jezzode;486396”]
Aim: Not to be hit by car while crossing road
Rule 1: Have eyes open when crossing road
Rule 2: Close eyes and hope you don’t get hit
In both instances you could get hit by a crazy driver, but i’m sure we could all prove “Rule 1” to be more “robust” when it comes to staying alive.[/QUOTE]
What if backtests show that the survival rate of crossing the road with your eyes closed are higher then with your eyes open?
haha, then we live in a society that is more than backwards
[QUOTE=“Jezzode;486401”]haha, then we live in a society that is more than backwards ;)[/QUOTE]
Lol well we could think that rule one would be more robust… But in actuality we didn’t know that drivers respond more empathetically towards people walking with their eyes closed and are more careful around them…
Hence rule 2 is actually more robust… And the only way to make that determination was by looking at success rate.
oh dear, what has happened, seems we have a double helix of an answer… I can’t solve this i’m afraid.
Nothing can be quantified these days!!
[QUOTE=“Jezzode;486405”]oh dear, what has happened, seems we have a double helix of an answer… I can’t solve this i’m afraid.
Nothing can be quantified these days!![/QUOTE]
Well I’ll answer the original question…
5 months… I backtested a system over 5 years by hand, thinking it was enough data… Was happy with how consistent it was… Went to live forward testing… Made great profits the first 3 months and then lost it all and then some the following two… That’s when I stopped trading mechanicall systems.
Looking back… All I was essentially doing in the backtest was curve fitting inputs to result in a positive output… “If this indicator does this, and I reduce the SL by this, and rule out trades on this day… Then I can eliminate this historical loss! Woo hoo this is a great system!”
We all go through this stage, I have, and used curve fitting excessively. Which is why it all comes back to keeping the system simple, less variables to change as and when you want to fit the historical data in your favor.
To me there is a definite difference between robust and profitable.
Robust, to me, means a system that gives similar results even if you change the parameters slightly, and also a system that works on many or most timeframes and instruments.
Profitable just means it ends with higher equity than it started and it says nothing about the “quality” of the system.
Just my 2 cents
[QUOTE=“o990l6mh;486434”]To me there is a definite difference between robust and profitable.
Robust, to me, means a system that gives similar results even if you change the parameters slightly, and also a system that works on many or most timeframes and instruments.
Profitable just means it ends with higher equity than it started and it says nothing about the “quality” of the system.
Just my 2 cents[/QUOTE]
There we go… I can live with that definition.
However you still mentioned “and also if the system ‘works’ on many or most timeframes”… “Works” is the keyword … Which would be synonymous with “profitable”… So profitability still seems inherent… However there is at least a concrete expansion to the concept of mere profitability, which is profitability with broad application.
for me using a rigid set of parameters. 11 years, 6.5 in sample 4.5 out. Including live trading. there is 1 problem, it goes FLAT. So instead of re optimizing variables, it just goes through long flat periods where its basically “breaking even” or slowly losing until the market regime aligns positively with it again.
[QUOTE=“MeiHua;486457”]for me using a rigid set of parameters. 11 years, 6.5 in sample 4.5 out. Including live trading. there is 1 problem, it goes FLAT. So instead of re optimizing variables, it just goes through long flat periods where its basically “breaking even” or slowly losing until the market regime aligns positively with it again.[/QUOTE]
Does it work on multiple pairs?
I agree.
I will point out that one thing that hasn’t really changed is the trading sessions and their characteristics. The London and NY sessions see a lot more action at a consistent level than asian sessions.
O990L6MH~ IS THAT A BARON 58 ON YOUR AVATAR? How much is Avgas going for in Sweden? It’s close to $6\gal. here.
Works best on stock indices(SP500, Dow30, russel 2000) but it does perform on other instruments. It has 70+ system years.
i never stick to a particular system…sometime i trade on naked chart, the only system i believe profitable is your trading experience, so far i had no luck with other system…1 thing for sure is to accept lose while consistently profitable…
To be honest I have no idea what kind of airplane it is on the avatar, that’s MacGyver, my childhood TV-hero, in the avatar - not me. I’m not sure what Avgas is?
[QUOTE=“o990l6mh;486925”]
To be honest I have no idea what kind of airplane it is on the avatar, that’s MacGyver, my childhood TV-hero, in the avatar - not me. I’m not sure what Avgas is?[/QUOTE]
I learned from that guy how to make a welding torch out of a magnesium framed bicycle and rust scraped from a metal bar…
Darn…but you’re a dentist, right? Oh well…
Avgas is short for aviation gasoline…it’s usually 100LL (for low-lead), at least in the US. Thought it would be interesting to compare…well, what’s automotive gasoline going for over there?
…and for the thread…I think it’s safe to say it’s best to use ranging strategies generally in the Asian session, and Trend\breakout strategies during the London\NY sessions.
Am a dentist indeed.
Automotive gas… around $2.25 per liter. Something like 80% of the price is taxes… Lovely country eh
But actually it’s even worse in some other European countries. For instance, prices are generally even higher in the UK which surprised me. And the poor Greeks have the same prices for gasoline that we do in Sweden, just that the average Swede makes around 4-5 times more in monthly income after taxes… I couldn’t imagine what it must be like to have those prices combined with so little money.