Thank you, Dennis. Yes that's also my major problem. Commissions are just one side of the medal. Slippage can really kill you because there is no guarantee for the maximum range. So high frequency trading amplifies this source of unpleasant behavior. And of course the lesson then is to know the stock you want to trade. This does not make things easier but in many occasions a hell of a lot more interesting ...
Regards, Ton. ----- Original Message ----- From: Dennis Brown To: [email protected] Sent: Sunday, May 27, 2007 9:13 PM Subject: Re: [amibroker] Re: Ideas for Swing Trading? Ton, I have run numerous simulations and actual high frequency trading (up to 80 trades a day). The commissions on even very discounted brokers can make a big difference. It pays to simulate, then adjust the commissions to some popular plans and see what it does to your equity curve. Also don't forget that the executed prices are usually at the top of the bar for buys and the bottom for sells. However, stocks like AAPL trade often at just a one cent spread. Others that trade thin can run the whole swing for a minute before you get executed. You have to look carefully at the stocks you want to trade and adjust the "commissions" to account for such slippage to give a real world equity curve. Dennis On May 27, 2007, at 8:27 AM, Ton Sieverding wrote: Thanks a lot for your very positive answer. It gives me hope again. Any idea if the examples in your PDF are before or after transaction costs ? Should be after of course but you never know. I understand that high frequency trading and more in particular lots of different stocks can give you a smooth equity curve. But my experience is that a high frequency system creates 'the most bang for my broker' in stead of for my bucket ... And finally, yes that's what your email says : "a few percent on good days". So I assume that the good days represent the famous 20%, 30% is acceptable and 50% of the trades create a Stop Loss ... Regards, Ton. ----- Original Message ----- From: Herman To: Ton Sieverding Sent: Sunday, May 27, 2007 11:36 AM Subject: Re: [amibroker] Re: Ideas for Swing Trading? Hi Ton, to be exact I wrote "a few percent on good days..." I found that trading smaller time frames and not staying in the market overnight reduces DDs significantly. To see a typical example of what increasing frequency does to your DD see http://www.aima.org/uploads/Omega64.pdf Most high-performance EOD trading systems try to predict price movement, this in my opinion, is virtually impossible. imo, Success here can be equated with a lot of luck and good money management. On the other hand trading pure short-term volatility or noise, which varies far slower or is more constant if you like, and which is almost completely trend-immune, you can create systems with very small DDs. Further diversifying such systems by trading many stocks (10-100) you can create equity lines that make your mouth water. Such systems would be totally useless to traders trading large amounts of money however they would be perfect to the small trader. Regarding "minimum code". About a dozen lines is it for most of my systems however automation code can easily run into 500-1000 lines of code. A simple answer to your simple question: Yes. Of course compounding is impossible or at best limited, for these systems, that is simply common sense. best regards, herman Sunday, May 27, 2007, 3:35:44 PM, you wrote: > Herman thanks for your short resume of the Trading world. Just a simple question. Do you really believe that group number 1 exists ? So Traders that do generate with a minimum of code on a consistent basis a daily return of 2,5% without losing their pants on a terrible outlier or drawdown that will take them out of business ? My experience is that only a very small group of about 5% of the '2,5%+ return Day Traders' is reaching for a relatively short period of time the above target ... Regards, Ton. ----- Original Message ----- From: Herman To: Howard B Cc: [email protected] Sent: Sunday, May 27, 2007 2:08 AM Subject: Re: [amibroker] Re: Ideas for Swing Trading? Every few years this type of discussion surfaces and it is great fun to read It always surprises me how two types of traders can be so oblivious to each others' way of thinking. Consider two types of traders (ignoring the many types in between): 1) Those who scan 100+ stocks in Real-Time and trade small lots of 100 shares (or whatever the market allows) 5-100 times a day, easily making up to a few percent on good days, using an automated trading system. 2) Those who trade portfolios with 1000-10000 shares/trade and must roll over millions of dollars trading for others, making, if they are lucky a few percent/month. We have both of these traders on this list but really they should have their own lists, perhaps AmiBroker-Fat and AmiBroker-Skinny their expectations are not and cannot be the same. In the first category volumes, market trends, market analysis, traditional TA, etc. play a minor role in system design. Their systems can be extremely simple and their trading rules may be expressed using only half a dozen lines of code while their automation code may easily exceed 1000 lines. Their trading screen may only display a lists of tickers with order status: no charts. They work hard to design and optimize code for maximum execution speed so that to can get their orders placed before the next quote comes in - speed translates in profits and 20-40 mSec execution is typical. Almost everything for the second category is reversed: they thrive on traditional TA using many colorful chart-layouts, perhaps totalling 1000s of lines of code. Their automation code, if they use it, may just be a a hundred lines long and aims to save them some typing - not to catch a trade. They use old (10-20 years!) techniques and statistical analysis that are rehashed over and over, they thrive on sophisticated analysis to squeeze out a fraction of a percent more per month (or reduce awful DDs). Code can be bloated with cosmetic stuff and its OK if it takes 5 minutes to execute. Traders from both categories ought to respect each others. best regards, herman Sunday, May 27, 2007, 5:27:22 AM, you wrote: > Hi Dennis -- Averages 2.5% per day!? That same $1,000 starting account becomes $294,000,000 in two years. (1.025) ^ 510 = 294,558 Please pass my email address on to your friend who gets 2.5% per day. howardbandy at gmail.com I have contacts who will reward him handsomely. When Larry Williams ran $10,000 to $1,000,000 in one year and became famous for it, that required a return of 1.84% per day. 2.5% per day turns $10,000 to $5,039,800 in one year. Help me understand -- Assume I can average 1% per day on, say, $100,000. Every month, I start with $100,000 and make $24,471 on that $100,000. Why would I pull my $24,471 profits out so that they can make 1% for the next month instead of continuing to trade them and making 24% for the next month? And, yes, trading in size affects the market. But if your friend is trading several times per day in markets with high liquidity and narrow bid-asked spreads, then $1,000,000 is still small size. QQQQ and IWM each regularly trade $5 billion dollars a day -- $1,000,000 is 5 seconds worth of trading. Pardon my skepticism -- Thanks, Howard www.quantitativetradingsystems.com On 5/26/07, Dennis Brown <[EMAIL PROTECTED]> wrote: I know of more than one 1% per day method, but of course it will not work to compound. That is not the way a true trader does it. I know a trader who averages 2.5% per day on about 5 trades per day on one ETF, and holds no position overnight. He pulls his profits out and lives on them or puts them to work in longer term investments. High rates of return only work for small investments and usually require a lot of personal attention and pattern recognition during the day. If it worked for large sums, or easy computer algorithms, the big boys (or hoards) would work that angle to death and the edge would get neutralized. Once you try to increase position sizes above a certain amount, you start to influence the market and you have no one to play against --it takes two to have a market. That is why large mutual funds must look to a fundamental value model. They can not trade the technicals quick enough without killing the market. A true trader will just work the market technicals to pull out a small amount of money at a consistent rate (no home runs). Over time, the results add up to a decent living. Dennis On May 26, 2007, at 4:02 PM, Howard B wrote: One percent a day. Yeah, right. Compound one percent a day for five years and a $1,000 trading account becomes $278,000,000. Start with real money and own Manhattan. (1.01) ^ 1260 = 278,567 Howard On 5/26/07, dralexchambers <[EMAIL PROTECTED]> wrote: T-ohrt - the thing you are missing is not your technical ability, but your BELIEF and your ATTITUDE to new things. You seem to mistrust my recommendation when in fact you nothing of me, my level of trading knowledge, this system or my involvement with it (my involvement is none other than my affiliate link - just to make that entirely clear). If you believe that 1% a month is all that is possible, that will be your reality, and you will discount ideas that make more as trickery. If you want trade lists, further explanations on the system I recommended - discuss it with David, the author. It is not my job to divulge a system that someone else owns. However, I will say that David's system is very credible and also very simple. I have recieved a lot of support from David and his system opened my eyes to swing trading. I also know of an individual who makes 1% A DAY - and publishes all his methods and indicators for free, online. Look for The Rumpled One at: www.kreslik.com. I am currently porting his work over to Amibroker on that site. And yes, once again - it is all FREE, and you definately won't find it in your "Beyond Technical Analysis" book. AC
