Mechanical versus discretionary swing trading

[Jim_Nasium]

Quote from galvinlee888:

I don't want to be rude, but this is my take on your so called mechanical system.

There is no edge in your system, e.g. It is completely random. The only rea son your account sink overtime is because of commission and slipage. In other words, you will be in break even if you no need to pay those "fees".

Market is completely random, your system will work in certain market conditions ( either sideway, trend or revert to mean). You can't find a system that will work in all conditions. The tricl is you have to know the market condition, but no one ever can predict it correctly in long term, even the "most evil" institutions in the world. Tips: the BS indicator such as MA will not help, neither do the "high probability" PA or pattern as thosw never exist (at least none of them ever pass any real statistical significant test)

That wasn't rude, we're all grown ups here (I assume)

What does help, in your opinion?

 

[Visaria]

I had a v brief look at your system, makes logical sense to me (i don't like MAs though, but that's me). I would ignore everyone here and commit yourself to rigourous back testing to see if it really has merit.

So don't trade it, spend time looking at charts and do a thorough back test BY HAND not machine.

Also be very hard nosed about costs, entries and exits. Consider this: say your bar is 10 ticks wide. However your entry is 1 tick beyond this and say your stop loss is 1 tick as well beyond the bar. Your risk is 12 ticks, not 10. And this is before any commissions or the (ludicrous) spreads applied by your SB company! Something similar with your profits! You may think in your mind you are looking at a trade which gives you 2:1 reward/risk, but really it might be 1:1!

To overcome this, you will need big targets and (probably wider stops) and low costs.

edit: test it on futures as well.

 

[Visaria]

Quote from bwolinsky:

Stocks mean revert. Indexes do not, they trend and build momentum.

Stock index components as part of indexes do this due to the nature of market structure within the index they're a part of.

Futures and indices are not that way and dependent on other components will trend more than they revert.

If you haven't noticed this or observed through years of trading at the top of performance lists of mechanical systems that's primarily how I can identify complete beginners from people who have actually looked at the data and drawn these conclusions quantitatively.

Your post doesn't make sense, his system (i've only looked at it briefly) is essentially following a trend on a dip. If you are right about indices being more trending then his system would be ideal for index futures etc.

 

[Jim_Nasium]

Quote from Visaria:

I had a v brief look at your system, makes logical sense to me (i don't like MAs though, but that's me). I would ignore everyone here and commit yourself to rigourous back testing to see if it really has merit.

So don't trade it, spend time looking at charts and do a thorough back test BY HAND not machine.

Also be very hard nosed about costs, entries and exits. Consider this: say your bar is 10 ticks wide. However your entry is 1 tick beyond this and say your stop loss is 1 tick as well beyond the bar. Your risk is 12 ticks, not 10. And this is before any commissions or the (ludicrous) spreads applied by your SB company! Something similar with your profits! You may think in your mind you are looking at a trade which gives you 2:1 reward/risk, but really it might be 1:1!

To overcome this, you will need big targets and (probably wider stops) and low costs.

edit: test it on futures as well.

Thank you

 

[wrbtrader]

Quote from Visaria:

I had a v brief look at your system, makes logical sense to me (i don't like MAs though, but that's me). I would ignore everyone here and commit yourself to rigourous back testing to see if it really has merit...

I have always assumed that when someone say's they're using a mechanical system or automated system that there was rigorous testing (backwards, forwards and anything else).

It's just tough for me to imagine someone using a mechanical system or automated system without testing it prior to any real money trading considering the testing is done by the computer and can easily be simulated traded prior to throwing hard earned cash at it in trading live.

Quote from Visaria:

...And this is before any commissions or the (ludicrous) spreads applied by your SB company! Something similar with your profits! You may think in your mind you are looking at a trade which gives you 2:1 reward/risk, but really it might be 1:1!

To overcome this, you will need big targets and (probably wider stops) and low costs...

I wish more traders will listen to this and apply it. Too many losing traders and profitable traders underestimate the real cost of trading until the end of the year or income tax time.

Gotta treat trading like a business and understand that there are business costs of trading that's sitting out there waiting to stick its hand in our pockets after the broker commissions and spreads.

 

[MarketAddict]

Quote from Jim_Nasium:

Hi,

I have been working on a trading approach which is largely mechanical and based on price action and volume, i.e. I run a screener which returns stocks where x, y and z conditions are met and I then take positions in those stocks. Sometimes I may override the screener if something really looks wrong but largely I let the screener pick the trades and I enter the orders. I have a fixed rule on when to exit too.

I have a colleague who I recently discovered trades futures, and this morning we had a water cooler moment discussing the over-extendedness of the S&P and how he was looking to go short, I said that I thought a correction was long over due but my system keeps on throwing up long signals so I have no choice but to be long as I don't second the guess the system. If I think back to all the times I have thought "this market can't go any lower/higher" most times I have been wrong.

He looked at me like I was being extremely naive, and said "It's really more of an art than a science", and that ended the conversation as I contemplated what he meant.

Now I am not making money, but I am not losing big time either, my account is just slowly decreasing overtime but I am concerned that there is something fundamentally wrong in my approach and maybe my colleague is on to something. He has been in this game way longer than me and uses a discretionary approach, he talks the talk but I have no idea if he actually makes money over the long term.

I can see how mechanical systems can be seen as naive, if it were as easy as "if x do y" everyone would be making their living running a trade bot while they sun themselves on a Caribbean island, which is unfortunate for me because I think my mindset is very much geared towards mechanical systems.

Anybody have any thoughts on this? Any long term traders out there successfully using mechanical systems? Or to have an edge in trading do you need to bring something extra to the table?

Mechanical systems are cool if your risk reward is good enough.. If i lose two or three times knowing that one win will bring me back to breakeven or show profits.. Then im cool with the mechanical system..

Here's an idea that most people don't do.. Trade two systems! A lot of people are just stuck on trading or relying on one system.. Remember you said "my system keeps on throwing up long signals so I have no choice"? Well now you have a choice..

 

[bashatrader]

Quote from wrbtrader:

Gotta treat trading like a business and understand that there are business costs of trading that's sitting out there waiting to stick its hand in our pockets after the broker commissions and spreads.

+1.

 

[drcha]

Quote from MarketAddict:


Here's an idea that most people don't do.. Trade two systems! A lot of people are just stuck on trading or relying on one system.. Remember you said "my system keeps on throwing up long signals so I have no choice"? Well now you have a choice..

You are right, it is not a bad idea to trade 2 or a few systems. But the OP is currently just trying to get 1 system that works well. Two suboptimal systems do not add up to anything other than mediocrity.

 

[Jim_Nasium]

Quote from drcha:

You are right, it is not a bad idea to trade 2 or a few systems. But the OP is currently just trying to get 1 system that works well. Two suboptimal systems do not add up to anything other than mediocrity.

That's what I was thinking

 

[MarketAddict]

Quote from drcha:

You are right, it is not a bad idea to trade 2 or a few systems. But the OP is currently just trying to get 1 system that works well. Two suboptimal systems do not add up to anything other than mediocrity.

I agree with you 100 percent... but my main goal was to have the OP think outside the box... Trading isn't black and white.. There are many ways to skin this cat..