[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: [EquisMetaStock Group] Strange result - Monte Carlo Simulation



PureBytes Links

Trading Reference Links



Super,
 
Excellent post.
 
 


 
2009/4/29 superfragalist <no_reply@xxxxxxxxxxxxxxx>


Ticiano, the real question is can the strategy be traded live? One of the missing pieces in MS is a simulation module that would allow traders to trade their strategies live using a simulator that generates real time slippage and tracks the results from each trade from buy to exit.

A lot of strategies test out well, but are very hard to execute in live trading.

I track my trades manually to see how my actual trading results compare to my predicted results. If my actual results start to decline, I try to figure out why, and if it's something beyond my control like market conditions, I'll go into a holding mode and stop buying or selling for awhile.

Volume is one of the two variables that most TA is based on. Price and Volume are all the data available for most indicators, so volume is a factor.

While I understand system expectation calculations and the importance of them, I also look at system consistency. Consistency wins the game over the long run. Volatility in one's capital account is a real killer. The majority of systems and strategies fall apart when analyzed for consistency. Big gains and big losses create a lot of emotional intensity that causes mistakes.

I hope your live trading results turn out to be close to your test results.



Super

--- In equismetastock@xxxxxxxxxxxxxxx, Ticiano Rêgo <ticianorego@xxx> wrote:
>
> Super,
>
>
> What amazed me was the fact that with the same strategy but with different
> kind of securities (with volume filter) the result changes between these two
> different forms of order execution.
>
> Win ratio remains almost the same but the measure G that Ralph Vince
> describes in his book changes. Actually I found many systems with high win
> ratio but with lower G when one compare one system against the other (i.e
> the finishing capital is greater but the win ratio is lower).
>
> Kind regards,
>
> Ticiano
>
>
>
>
> 2009/4/29 superfragalist <no_reply@xxxxxxxxxxxxxxx>
>
> >
> >
> > Ticiano, your question is hard to answer because of the vague amount of
> > information in your post.
> >
> > First I don't rely on signals, indicators or chart patterns very much. I
> > use signals as a warning indicator, but I don't act on them. The trading
> > decision is based on my perception of the market, what's going to happen
> > over the next several days or weeks and what is happening with other similar
> > equities. If I get a signal, or the chart looks good, I'll buy if the other
> > variables look right to me. I have a group of market condition charts I look
> > at to help me understand what is going on, but again I don't rely solely on
> > them. I listen to cnbc during part of the day. I watch nightly business
> > shows after the close. I read the Wall Street Journal everyday, and use a
> > dozen other sources of info that factor into my buy decision.
> >
> > If you are day trading, the time of the day will impact what a system does.
> > Some day traders only trade in the last half hour, some trade mid day and
> > other like the open after the first five minutes. One indicator I got from
> > John Clayburg uses the first five minutes of trading data with the next
> > hours worth of trades to predict the direction of the price for the day as a
> > whole. It's 75% or so accurate.
> >
> > Based on the little info in your post, sure it's possible to have better
> > results in the last half hour. Sometimes computer trading kicks in and
> > sometimes it doesn't. Generally, you can see the volume swings when computer
> > trading kicks in. It is possible to make money from that if you are grinding
> > out day trades. You can also see volume swings and price adjusting when
> > lunch time is over in NYC. Since I buy for a longer time horizon, I know
> > which times of the day are likely get my orders filled at cheaper prices. I
> > also know the specialist is closing out there book late in the day so I'm
> > likely to get fills I wouldn't get earlier on thinly traded ETFs or closed
> > end funds. The price difference isn't huge but I always look for the best
> > price. Sometimes that strategy works and sometimes it doesn't. I've been
> > left partially filled at the end of the day, and the price move the next day
> > wiped out any price advantage from the day before.
> >
> > Volume is generally a good filter so it often improves results, but not all
> > of the time.
> >
> > Have fun!
> >
> > Super
> >
> > --- In equismetastock@xxxxxxxxxxxxxxx <equismetastock%40yahoogroups.com>,

> > Ticiano Rêgo <ticianorego@> wrote:
> > >
> > > I am testing one trading system with Monte Carlo Simulation and I have
> > found
> > > this:
> > >
> > >
> > > Test 01) Buy on the open if open <= ref(close,-1). Explaining better: Buy
> > > the next day of the signal if it opens less or equal yesterday´s close.
> > >
> > > Test 02) Buy on the close. On the same bar of the signal.
> > >
> > >
> > > With test 01 TradeSim gave me EXCELENT results with its performace much
> > > better than Test 02.
> > >
> > > The problem is when I include a liquidity filter such as mov(c*v,22,s)
> > > > = *certain
> > > value* than Test 02 becomes better.
> > >
> > > The results shouldn´t remain the same ? Is it possible that considering
> > > these securities (the most liquid ones) professionals are deciding on the
> > > close ?
> > >
> > > Super, if it is possible I would like to know your opinion.
> > >
> > >
> > > Kind regards,
> > >
> > >
> > > Ticiano Rêgo
> > >
> >
> >
> >
>




__._,_.___


Your email settings: Individual Email|Traditional
Change settings via the Web (Yahoo! ID required)
Change settings via email: Switch delivery to Daily Digest | Switch to Fully Featured
Visit Your Group | Yahoo! Groups Terms of Use | Unsubscribe

__,_._,___