Hi, I think superfragalist has underlined the
more important fact; a good education in the mathematics of trade sizing is the
key factor for the success of a trader.
I'm not an expert trader, but more I study the subject and more I understand that
the key is trade sizing: very simple and ROBUST trading system with successful
money management (severly tested).
A strong plus is to trade different and uncorrelated markets, so a good money
management has also to manage the intermarket risk (avoid concentration based
on reliable correlation model).
For this reason i would like to know what are your idea of the best software
for backtest trading system on multiple markets at the same time with complex
money management.
Just a question:does anyone use MATLAB to backtest?
Also can be very interesting create a topic just on the trade sizing to
evaluate the best trade sizing idea and the more reliable money management
system.
There is few information on the web and bring all our study togheter can help everyone
to proceed faster.
Also I want to suggest a method for the martingale test: you can export the
data with the indicator in excel and then trough a simple counter add the size
you want based on the number of BUY signal.
If you want I send you an exemple in excel.
But I think It's not a good idea a pure martingale: you have big probability to
win little and little prob to lose everything....but the probability lo lose
everything haven't to exist in a good money management!!!!
It's the opposite of the winning money management: usually less prob to win
big, more prob to lose little (ad ex: 0.46 prob win 2x the loss).
You have little risk to die, and more time to exploit your statistical
advantage.
If you lose everything It means that your trading system hasn't a statistical
advantage...so go to casino, it's the same!!!
Just to Know:martingale have 0 expected value, so it's not a profitable
strategy, the only positive thing is: if you have INFINITE wealth (if you have
why you play?;-) always exist a time T where you can go away with a win, but
more you play, more wealth you need!!
--- In equismetastock@xxxxxxxxxxxxxxx,
superfragalist <no_reply@xxx> wrote:
>
> You can't test your betting strategy in MS. You will have to use other
testing software.
>
> You are trying to use a martingale strategy. If you don't know what that
is look it up. Martingale is a losing strategy unless two things are present.
The odds of a winning bet have to be greater than 50% and the gambler has to
have infinite wealth. The exponential growth of the bet size across consecutive
losses cause a maximum adverse excursion that will bankrupt anyone without an
infinite size capital account, even with a 50% win/loss ratio.
>
> I've written about this in Roy's newsletters. Maybe you can get a copy.
It's cheaper than the wallet whipping you're going to take using a martingale
system.
>
> But if you insist, you can get an idea of how expensive martingale systems
are by running a few back tests using TradeSim. Take the maximum number of
losses in a row, compound up the bets across the series and that will give you
MAE. It will also give you the final bet size before a win.
>
> That's where the infinite part comes in. Traders don't have infinite size
capital accounts. Most systems will generate one or more sets of consecutive
losses that will completely deplete the capital account. If you still don't
believe it will happen to you, run a monte carlo simulation focused on sets of
losses and see how many times strings occur that cause bankruptcy.
>
> If you like that kind of betting, then you can apply Optimal f trade size
to your system. At least that has a chance of paying off. And you'll get just
as many emotional thrills from optimal f as you would from a pure martingale
system, and if you have the brass to stick it out, you'll eventually win before
you go broke.
>
> Rather than spend your time reinventing a wheel that has already been
tested many, many times, you might consider reading a Ralph Vince and get a
basic education in the mathematics of trade sizing and why some systems fail
and others work regardless of the back test results.
>
> Super
>
>
>
>
> --- In equismetastock@xxxxxxxxxxxxxxx,
"deadlysmile01" <deadlysmile01@> wrote:
> >
> > So many advices against my thought of doubling the bet?
> >
> > But none asked about whats the system. I guess every one here is a
professional trader except me. My bet is that system wont give me wrong ideas
it correctly catches the trend, atleast the long term trends. It fails in the
whipsaws. Where in i double the bet.
> >
> > But, i am just BACKTESTING. I have not yet put the money. If the
logic fails ( that is double the bet size ) i will not include it.
> >
> > I mean c'mmon, metastock does have a performance sheet thingy? Right?
> >
> > So, please, pretty please, kindly provde me the formula on how to do
the doubling logic in metastock system tester.
> >
> > Thanks all..you all been great. :)
> >
> > --- In equismetastock@xxxxxxxxxxxxxxx,
Code 2 <Code2@> wrote:
> > >
> > > Nope, failure is guaranteed. That's simply a mathematical fact.
> > > Assume that your system guarantees a loss of no more than 10% of
your
> > > bet size and you double your bet after each loss. Even with an
> > > initial bet size of just 2% of your account, you'll be out of
business
> > > after only six consecutive losses. With an initial bet size of
3%,
> > > you'll be dead after five consecutive losses. That doesn't even
allow
> > > for the occasional Black Swan where the market blows through
your stop
> > > and causes a loss of more than 10%.
> > >
> > > Even a great trading system will yield five or six consecutive
losses
> > > every once in a while.
> > >
> > > Martingale bet sizing does not work in Las Vegas and it does not
work
> > > in the markets. You have to get smaller after each loss.
> > >
> > >
> > >
> > > From: thomas12ng <thomas12ng@>
> > > To: equismetastock@xxxxxxxxxxxxxxx
> > > Date: Saturday, March 28, 2009, 3:21:20 AM
> > > Subject: [EquisMetaStock Group] Re: Hi all - Help Need .
> > >
> > > depends on your system dude! and it depends on your personality
and
> > > your perception about this system
> > >
> > > if your system loss 20% of 100% total trade then your method is
> > > perhaps will claim back all the losses
> > >
> > > but if the system keeps trigering at the down trend even the
father of
> > > super martingale can find him self sleeping at his neighbour
garage in
> > > the next morning :)).
> > >
> > > my suggestion is that
> > >
> > > 1. just cut and paste your system logic into indicator builder
and
> > > see whether the signal correct and suit your personality
> > >
> > > 2. avoid pyramiding and see how your system goes
> > >
> > > 3. just ask your self about your system. was the system signal
is
> > > purely from you or someone who only know a slight coding method
and
> > > admit him self as a "coach or even trading guru" offer
you this
> > > system. I did experiencing this condition before. when I ask him
> > > back whether you know what the system picks he was just
smurking. at
> > > that time I was so angry and knew has been deceived. felt like I
was
> > > going to smash his head by a base ball stick. since then I never
> > > trust anyone before my eyes see it
> > >
> > >
> > >
> > > --- In equismetastock@xxxxxxxxxxxxxxx,
"Patrick Butler" <pat494@> wrote:
> > >
> > > > Hi,
> > > > You are just about to make one of the worst mistakes in
trading imho
> > > > - that of doubling down !
> > > > You will get away with it a few times and feel pretty
confident but
> > > > when it doesn't work out as will be the case it will eat a
large
> > > > hole in your capital ! Accept this advice of an old trader
and save
> > > > yourself a blown up account
> > > > regards
> > > > Pat
> > >
> > >
> > >
> > > > ----- Original Message -----
> > > > From: deadlysmile01
> > > > To: equismetastock@xxxxxxxxxxxxxxx
> > > > Sent: Thursday, March 26, 2009 9:39 AM
> > > > Subject: [EquisMetaStock Group] Hi all - Help Need .
> > > >
> > > > Hi all,
> > > >
> > > > I am backtesting a system where whenever i go wrong ( i.e.
if the
> > > > trade results in a loss) i double the size of the bet.
> > > >
> > > > So i need to modify the entry size under the Buy order /
sell
> > > > order tab, i tried using Simulation.CurrentPositionPerformance,
and
> > > > Simulation.CurrentPositionSize which resulted in an error.
> > > >
> > > > Can someone please help in that.
> > > >
> > > > Thanks.
> > >
> > > --
> > >
> >
>