Home
Search Forums
Keywords:
Search Titles Only
User Name:
Exact Match
Show Results As:
Advanced Options
Reply
 
Thread Tools Search this Thread
  #80051  
Old Nov 14, 2010 11:13pm
joelcf's Avatar
My gun control is a steady hand.
 
Member Since Jun 2009
More than 10 Vouchers  732 Posts
Default Strategies against Architecture.

I had dinner with a friend on Friday who works as a dealer for a CFD house. So, in between copious amounts of sake and shochu, we got to talking about trading, and the stupid things we have seen people do. On the way home (edit: the next day. late, late the next day), thanks to weather delays, I had plenty of time to think about things and got to typing.

This year has easily been my most profitable ever. The part I devote to straight PA trades has had a total return of 112% over the past 12 months. My entries and exits in other areas were much, much better than my old 'I want to buy xyz, hit buy button and ignore short term market conditions'. Part of that is overall market conditions, part is some big calls that worked out well... but a big part is just because I've stopped doing so many stupid things that I didnt even know I was doing, or didnt realise how stupid they were.

In fact, the biggest revelations about my own dumbassedness came from teaching someone else, and having them question me on why I do or think certain things. So it isnt even like I realised it on my own.

I'm taking a break for a little while, so I thought I'd leave you with some positive, supportive thoughts that came out of it.... in rough descending order of dumbness


Why you still suck at trading.

1. Your money management blows

You wont admit to it, but you stake way too much of your account on each trade. Whether its 100%, 50%, 25%, 10% or 5%, you know you do it. You might claim to only trade 2% in the thread, but when it comes down to it, you throw the rulebook out the window.

Why do you do this when you know it is a bad idea? It's usually one of three reasons, from what I can tell.

a) You are undercapitalised, and therefore look at your account in absolute terms - you dont risk 2%, you risk $20. After all, its only $20... and you will never get rich trading 2%! Guess what - you wont get rich trading 50% either, you will just lose your account faster and have to reload. You racka disciprine. So stop it.

b) Even worse, you know you shouldnt trade 25% of your account, but this trade is so good and you just know it is going to work out. What you dont understand is that you arent entering into the trade with a specific outcome, there is a distribution of outcomes possible. Some will win you money, some will break even (or close to it), and some will lose you money. Even if a trade is an 80% probability to 'succeed' (see: 7), in some definition, you dont know the outcome of this particular trade, just the distribution over a large number of trades. The 2% rule is there to make sure you can trade that setup enough times that your 'edge' (a term i hate) is borne out. Trading 25% makes it almost certain that you will get your ass kicked by variance and end up broke.

c) You already lost half your account doing a) or b), and so now you are trying to 'get back to even', or you are pissed off, or just gambling. You know this is dumb. Go for a walk or something.

There have been dozens of posts on why the 2% rule is a good one, so I wont bother reiterating it. Just stick to it. If I had to guess, I'd say that following this simple, dead easy rule would cut down the number of blown accounts out there by at least 75%.

2. Your trade management is non existent or dumb

You dont manage your trades appropriately, because you are a greedy jerk. Or a scared little girl. Mike and Jarroo tell you, repeatedly, to look for areas where a price reaction is highly likely so you can plan ahead and manage your trade. But you dont.

Instead, you engage in the age old trading tradition of falling back onto stupid, flippant sayings instead of solid logic.

Like 'Im giving my trade room to run!'. No you arent. You are failing to manage your risk because you think you are going to hit a 20R runner everytime, and the only reason you arent as rich as Abramovich is because you get keep getting stopped out at breakeven or 0.5R or whatever. If only you had your stop 400 pips backwards, you would have totally caught that last runner!

But unless your 'room to move' is based on some solid foundation, like an ATR distribution curve or prior S/R, you are just closing your eyes and hoping. Like every other broke loser.

Same goes for 'I didnt want to close my trade for under 1:1 return!'. That's flawed logic. The market doesnt know where your stop is (well, most of the times). It doesnt know what your return is on this trade. It doesnt care. Basing your exits around an arbitary target, whether pips, %R or moon cycles, is retarded in this market. Again, this has been reiterated dozens of times in the thread - keep your MM and your trade management seperate.

So do it. And while it is better than nothing, dont just go 'well, there is the FTA, so I will just take full profit'. Put some damn effort into it. How strong is the FTA? What kind of market momentum is there? Are there big news announcements? Make a list of potential 'trouble areas', how you expect price to react there and WHY you expect that outcome. Then you can make your plan.

3. You are still trading low timeframes

Low timeframes are different from high timeframes. It's just how it is. Argue your fractal wave symbosis theory all you want, but you are wrong.

Why? Because there are different forces acting on each time period, with different market participants having different strategies and goals.

At the most basic level, there is more noise at lower timeframes. Single orders, pieces of news and random spacejunk can shake you out of your trade alot easier. And so not only are your probabilities for 'successful' vs 'failed' trades changed, but so is your distribution of returns. And therefore your entire profitability.

It isnt just a case of 'there arent enough bars on the daily, ima trade the M15 using the same strategy'. Because you can try that, and it wont work. I know, because I lost a bunch of money doing it before my dumb ass adapted. It's harder. Alot harder. And until you can trade the daily profitably, you are an idiot if you are trying to trade the M15 to try and double your stupid account.

Jim tells you to stick to higher timeframes to protect yourself. As a Texan, he knows about protecting yourself. Listen.


4. You overtrade like a crackhead on welfare day

There are two forms of this: the egregiously stupid version, and the more subtle version. You just cant help your damn self. ooh, there is a 'C+ trade', I'll just trade that because I'm bored!... Or the market is so dead, im going to make some action! buybuybuy!.

Once again, show some discipline. Stop taking shitty trades.

Or the more subtle version, which is easily missed. Assuming you arent one of the lazy brokeasses who refuse to plan their trade, you should have a distribution of possible outcomes, and what you will do in each case.

As we will talk about later, estimating things is dodgy. But even if you go for the simplest version of 'bail out at FTA', you should have a decent ballpark of how often you expect the trade to get to the FTA, and how often you expect to be stopped out. So calculate your EV, even roughly, and make sure it is positive. If it isnt, you are losing money. Even if it gets to your FTA and you bank pips, you are losing money. You just have a short term focus, and you dont realise it yet.

So now I hear you whining BUT MIKE DOESNT DO THIS! You know what? You arent Mike. Or Jarroo. Or anyone else. You dont have a good idea of what price is going to do, you dont have an intuitive idea of your EV, and you are still losing money. So quit whining.


5. You copy trades from the thread

This is just dumb. I know a bunch of you do it, including the lurkers and the 'plz post chart' people. You want to disown responsibility for your own account and just copy Jarroo. Only, by the time you see the chart, the trade is gone and you are chasing the tail lights. And you lose. So try using that brain and do your own work - otherwise, you may as well go download an EA and then bitch in 6 months when your account is gone.

6. You are still bar hunting

This one is pretty straight forward. I cant count the number of times people have said 'the bar is just a trigger, focus on location'. But you dont. You run your eyes over thirty charts, looking for a pinbar. Because PINBARS = PROFITS!@#. And then it fails, and you bitch about how pinbars are unreliable, and trot off to Jackos thread.

NOTHING works in isolation. Nothing. Not the best looking bars, not the cross of the 213EMA and 38.75SMA, not your superadvanced quant strategy that you made up with Barney the Dinosaur math.

Without posting a bunch of math and charts, your 'edge' is much, much smaller than you realise. Especially on your M5 chart that you lie about not trading (see: 3). If pinbars were so bulletproof, they wouldnt work. Despite what clueless authors with a book to sell try to tell you, fx isnt special and immune from market efficiency. It isnt like pins (or any other basic setup) are some super secret that only a couple of people on the internet know about. They are common knowledge. And if they really were so foolproof that you could close your eyes and mash the buy button for mad profit, then dont you think the instos would have pounded them into the ground by now?

(as an aside, this is also why the guys that come around every 3 months and ask 'how did i make pinbar EA?' are so misguided. I know people who do algo programing for investment banks. They are a hell of alot smarter than you, they have toys you couldnt dream of, and they have actual accounts to trade tiny, tiny edges. You arent going to win that battle, friend.)

Put simply: try reading the damn thread, not just grabbing the barsetups.pdf and trading your $84 account into the dirt.


7. You keep shitty records which makes you overestimate your profitablity and the probability of your setup 'working'

Here is the problem: despite people telling you to keep records, you dont. MY ACCOUNT BALANCE TELLS THE STORY LOL you scream, in cat macro form. But it really doesnt.

Or, slightly better, you keep some records. But they are patchy. And you dont always do them. Or they only have a few little bits and pieces.

Your records are really, really important. Why? Because you want to know if you are making a profit or not, and your account balance doesnt tell you this. It only shows you one possibly outcome, usually based on a small sample size. You need an accurate picture of your expected value, otherwise you could be on a long, slow slide to bankruptcy and you dont even know it.

Anyone who works in an applied statistics environment (banking, trading, economics, actuarial, hell even poker) will tell you that solid data is absolutely invaluable. In fact, I'm going to go and tell you that right now. With a strong dataset you can really analyse your trading, spot any holes in your methodology, adjust your course, and make some cool graphs.

It can be as complex or as simple as you want, the format and application you use dont matter. Just capture the key info - details of the trade setup itself, the market conditions, the positives/negatives of said trade, your trading plan (including FTA distance/strength, profit outcomes, etc), how the trade actually worked out, and anything else you think is important. If that's too much for now, just take a screenshot and do it when you understand things better.

You are trading probabilities. Data is your lifeline. Dont drown.


8. You think backtesting is awesome and demoing is for suckers

Jim tells you to demo. You decide you dont want to wait, so you will do a backtest instead. Your exhaustive backtest is dumb and isnt a real substitute for demoing.

a) It's biased. You are going to look at all the bars, claim you would have done x and won tons of pips, even though it isnt true. I'm not a psychologist, so I cant tell you whether it is pride, vanity, laziness or arrogance. But its useless.

b) Even if you treat it like a live situation, you are assuming you would make the correct decisions. With no money on the line, of course you would. With actual money? Well.

c) Disregarding the above, your statistical prowess is close to that of a lemming.

i) Your sampling methods are junk. You look at a hundred pinbars, find that 70% worked and therefore assumed that you can autowin $$$. The problem is that you have cherry picked examples - obvious pins, rather than all pins.

ii) Your stratification methods suck. You didnt stratify them to look at homogenous groups. Or if you did, you were biased when you did it because you used subjective criteria. The old 'thats a b+ pin'. So you are end up with either a non-homogenous sample or a biased sample.

iii) You use vague criteria. You dont define your experiment outcomes - what is a successful pin? What is a failed pin? What time constraints?

iv) Your analysis doesnt take into account your method. You need to know where your stop would be, where your FTA would be, and the action you would take. Otherwise, you dont know how profitable you would have been actually trading that set of pins.

v) Your basic statistical methods are lacking. You dont understand sample size, confidence intervals, r-squared, p-vals, error levels or what a calculator is. You havent controlled for overall market conditions. You havent taken into account outliers. The hugest mistake people make with their 'testing' is drawing conclusions from a small sample size.

Am I saying you shouldnt do any backtesting? No. Just that it isnt a substitute for demoing like Jim tells you to IN THE FIRST DAMN POST.


Finally, we are out of the 'dumb' mistakes and more into concepts that can help. The last two points are stuff that you dont have to do, but they might help you... if your brain is so wired. Some people instincitvely understand this stuff. Some people dont care, and are happy with their results because it works for them. SOme people want all the 'edge' they can get. Some people overthink things and end up doing nothing. What works for you is up to you, basically.


9. You still dont see the whole market

This first one ties up alot of other amateurish behaviour - not managing trades, overestimating probabilities, dumb money management. It ties into all the stuff that you read Darkstar say but didnt understand. It ties into the posts about 'the big boys', and 'order flow', and other things most people confuse and misapply.

If you really want to progress, stop looking at 'the market' as a series of static, discrete bars. The number of 'what timezone is best?!' questions just shows that you still dont get it. Time periods are arbitary distinctions we make - in most cases, they are pretty meaningless. You are looking at something that works like a continuous process, so treat it as one.

Too esoteric? Fine.

When you see a bunch of strong moves up, and then a pin, does that mean the market is going to reverse? No. Draw it as a line chart and see how awesome your setup looks.

Want to be able to trade your super IPBs without begging Ghous to give his blessing before you hit the buy button? Then look at the bar in context and discen its meaning - what is the market doing, why is it doing that, who is making it do it, so what are they doing, and so what should happen if they keep doing it? You dont have to drill down to this level of detail, but if you ever want to move beyond ZOMGPINBARBUYBUYBUY! you need to start thinking 'why'.

Get to know market flow - how each timeframe 'moves'. Think about who is operating in each time period, and how it affects your chosen time period - are the instos trading off H4 bars? No? Then who cares what close you have. No one else is looking at it. Lots of people watching the hourly? Then look at your setup in the context of the hourly bars.

10. Stop being so damn myopic

All I need is charts, charts. All I need is charts. Or so the Beatles song goes (although my memory is a bit rusty).

You might have read that the charts tell the whole story, and so reading the news is pointless. Or that price reflects all the news, and so therefore you dont need to know anything else. It's those kinds of silly generalisations and misapplications of EMH that get technical analysts laughed at.

Note: the good ones dont do this. And the bad ones like to rely on this part of the EMH, and then ignore the part that says there is no predictive power in price. Make your mind up, boys.

Here is the thing: you dont need to know economics, finance, current affairs or even what century it is to be profitable. But can it help? Absolutely.

Firstly, as we talked about a few weeks ago, knowing when the news can come and kick your ass is a good thing. You wont get stopped out all the time and be left wondering why - as an informed trader, you managed your risk beforehand.

But going further, think of all that dumbass simplistic advice you have heard. The trend is your friend, follow the pack, buylowsellhigh, buyhighsellhigher, whatever. The point is, as a price taker (ie, you cant affect price because you are a baby fish in a bigass tank), you want to be going with the pack, not against it. But how do you know where the pack is going? They seem to gloss over that most of the time

At the most basic, simple level - you want a direction. Up or down. And the absolute best way I know of to get this direction is to understand some basic macroeconomics and finance theory. What drives currency flows? What drives specific currencies? What are some basic models of exchange rate movements? Of interest rate movements? How do I judge the relative 'strength' of economies? What macro risks are there in the trades I am carrying, and how will I react?

It sounds alot harder than it is. The hard stuff doesnt really come into it until you drill down further into modelling, and try to quantify things. But you dont need that. All you need to do to improve your trading by a fair whack, once you have all the other stupid mistakes from 1-8 fixed, is to do a bit of basic reading. It probably wont even cost you anything.

And, when combined with the kind of smart PA analysis that the guys here are trying to teach you, you will really start behaving like a professional.

For completeness, here are ten stupid things I still do, but I promise I'm working on

1. Recognise when you are wrong and adapt. Dont be so stubborn. A single instance of this single mistake cost me $9k this year. And its retarded and I should know better. I suspect there are many people in the same basket here.

2. Be flexible in your beliefs - dont just dismiss ideas that dont agree with your view of the world.

3. Stop being so arrogant. You arent smarter than the market.

4. Voting machine, weighing machine. Recognise the difference. Linked to (1)

5. Dont get lazy about record keeping. Especially when you have a memory as terrible as mine.

6. Some people are too misguided, too invested in their incorrect ideas or just plain dumb. Dont bother arguing with them.

7. Take every advantage you can get, but focus on the big picture. Your broker, your trading engine, the commissions and your execution speed and quality make a huge difference.... but in different areas. Worrying about 100ms on a 4 month trade is stupid, so stop micromanaging everything.

8. Trade something that suits you, and stop trying to force things that dont.

9. All work and no play make Jack chase his wife and kid around an abandoned haunted mansion with an axe and freeze to death. Dont do that.

10. There are way more than 10 stupid things I do, so stop pretending you recognise them all.


Take care guys, end 2010 in the money.
__________________
When you have to shoot, shoot. Dont talk.

Last edited by joelcf, Nov 15, 2010 1:01am
Reply With Quote
Reply

43 Traders Viewing This Thread (16 are members)
Adilius07, HanSolo, Bethel, RandomGamble, triger88990, fxtsun, jwm888, HxMark, plund, lafemo, Pinbar, ppfx, Kika, alexpop, crespoh, doogal
Thread Tools Search this Thread
Search this Thread:

Advanced Search


Similar Threads
Thread Thread Starter Forum Replies Last Post
PDF of james16 thread? Gasmeter Rookie Discussion 4 Jun 4, 2010 8:07am
rock my chair chart thread forex3e Trading Journals 48 Oct 30, 2007 11:21pm