Quote from marauder:
On the contrary.... and from my own experience... I cannot see how anyone (non-pro, this is) can make money consistently by swing trading. Even if it's intraday swing trading.
During three years, I've focused my efforts in building a custom ATS from scratch. From scratch means... well, from scratch: day zero was a blank screen, me in front of the computer, and a beer in my hand. I'm not talking about some clunky meta-language script-kiddie platform, but a "real", multithreaded C++ ATS coded by following the MPA (massive parallel application) paradigm, running all sorts of parallel calculations, one per core, and synchronizing them with the risk & executor management modules.
You name it, I've studied it first and implemented it later: multilayer neural networks, check; DWT/SWT realtime transforms and MRAs, check; DOM realtime handling, check; momentum time/frequency plane analizer, check. Each one being a plugin which I can enable or disable... and there are many, many more of these.
At the end of the day, I came out with a system which could project intraday absolute pivots (roof and floor) with an average CEP / deviation of +-0.25% . That meant for example that, one hour after DAX30's session had begun, you could know in advance that day's session high and low values with an average error of 13pips.
Pretty cool, uh... you'd thought you had got the edge. Let the dollars fall, then.
Hell no.
Most days, you felt like playing poker with marked cards. You knew in advance where the price would take a reversal. It looked like a win-win situation.
But then, sooner or later, came one of those days. Be it the rising tension between north-south Korea... the Brent spiking out of control... rising tensions in medium east... european debt skyrocketing... Bernake shoving all of you americans with suitcases full of grands (go Bear go, SP skyrocketing without any reason, and of course the rest of the indices would follow)... when a fundamental day (as I've come to call them) shows, there's no valid projection, no valid numeric-based ATS which will work even remotely as expected.
Greed and fear are just non-computable tasks as of yet.
Believe it or not, three or four fundamental days could waste a good trading month. Finally, the Big Money didn't arrive: there was always some fundamental days wich fucked the month. Hence I was consistently profitable month over month, but not with "happy money" as this post speaks of, but instead with "humble money".
After three years of market study, investigation and research; after a 25-year career as a computer engineer; after having thrown every esoteric math you can think of against the market (chaos theory... random walk... wavelets... time series analysis) I, respectfully, doubt greatly that someone which is not on the big money (thus having the ability to move prices more or less at will) can profit consistently from the market by doing swing trading, month after month.
Notice the word "consistently", since it's very important: to earn money from the market, anyone can do it... one day... two.... three...
... but to make a living of the market, that's a very different kind of animal. Error tolerance is minimal, and you must profit regular and consistently.
Now, if you talk me about HFT, microscalping and the like, then things change. With a good ATS , a good server and a good latency (global RT < 10 msec, and you're in the game) you can give many big players a run for their money. I enjoy the microscale-level game much, much more than I enjoyed my swing trading times.
but once again I, for one, cannot see the whole swing thing....
just my $0.02
