travis, I'm not clear on what problems you see with backtesting here. Can't you backtest scaling-in/out pretty much as you would any system, by specifiying conditions for buying and selling and how many contracts to trade?
You are using an out of date browser. It may not display this or other websites correctly.
You should upgrade or use an alternative browser.
You should upgrade or use an alternative browser.
Auto Mechanics vs. Automated Traders
- Thread starter chaostheory
- Start date
travis, I'm not clear on what problems you see with backtesting here. Can't you backtest scaling-in/out pretty much as you would any system, by specifiying conditions for buying and selling and how many contracts to trade?
This reply is to everyone in response the most recent questions or discussion.
We have been focusing our most recent discussion on the trend following perspective of scaling in and scaling out of positions.
But please note that trend following is really a sideline system.
The primary system is the RTM (reversion to mean) which makes steady consistent profits.
Just think of any channeling strategy (aka RTM) that you ever tried. Dozens of them exist. You always find them profitable during flat, sideways, chop strategies, but they suffer draw downs during trendy periods.
With fixed positions, it's impossible to pick the exact point to enter for a channel play. Is 1 standard deviation better? Or 1.5? Or maybe it's 2 standard deviations?
The further you get out from the mean with fixed position size, the higher your win rate but the less frequently you get to trade. Still you do get stopped out some times.
The closer you trade to the mean, you get higher trade frequencies but more often stopped out.
Due to being stopped out, fixed position trading of channels gives you occasional losing streaks. In money management that means you have to severely restrict how much you risk and leave room for potential draw down periods.
The draw down periods of any length are unacceptable in the goal of having 100% monthly profitability. Besides, who enjoys draw down periods?
However, try those RTM type strategies again while applying position sizing and you find you can easily make them 90% profitable or better. It's because you aren't forced to pick the exact point to enter or exit the channel. By fading into digressions from the channel center, it averages your entry price much better. When price does revert to the mean, you profit very consistently.
THAT my friends is the cash cow. Not trend following.
Why do I use trend following? It's only because the sideways systems don't work well during trends which happen infrequently and I want to make money every single month.
So that forces us to deal with the period where the markets become inefficient, the crowd mentality takes over and things go haywire.
When a trend starts it invariably catches the RTM channel-type system on the wrong side of the market.
So the trend following system kicks in to counteract the risk and convert them to profits.
In this way, the combined portfolio makes money "round the clock" so to speak.
You may find it really hard and nigh impossible to trade separate strategies simultaneously on the same symbol on the platforms mentioned here.
THAT is the real power here--merging strategies of different types on the same symbol.
So back to the trend system. Sure it costs some small losses to find a trend, but who cares compared to the piles of cash generated at the same time from the RTM strategy.
However, when the RTM strategy hits a trend, the trend strategy becomes a life saver--to be sure.
Still, in this holistic approach the trend doesn't produce a windfall. No. It simply counteracts the losses from the RTM strategy with a little profits to boot. So it mainly servers to keep up the 100% monthly profitability.
We have been focusing our most recent discussion on the trend following perspective of scaling in and scaling out of positions.
But please note that trend following is really a sideline system.
The primary system is the RTM (reversion to mean) which makes steady consistent profits.
Just think of any channeling strategy (aka RTM) that you ever tried. Dozens of them exist. You always find them profitable during flat, sideways, chop strategies, but they suffer draw downs during trendy periods.
With fixed positions, it's impossible to pick the exact point to enter for a channel play. Is 1 standard deviation better? Or 1.5? Or maybe it's 2 standard deviations?
The further you get out from the mean with fixed position size, the higher your win rate but the less frequently you get to trade. Still you do get stopped out some times.
The closer you trade to the mean, you get higher trade frequencies but more often stopped out.
Due to being stopped out, fixed position trading of channels gives you occasional losing streaks. In money management that means you have to severely restrict how much you risk and leave room for potential draw down periods.
The draw down periods of any length are unacceptable in the goal of having 100% monthly profitability. Besides, who enjoys draw down periods?
However, try those RTM type strategies again while applying position sizing and you find you can easily make them 90% profitable or better. It's because you aren't forced to pick the exact point to enter or exit the channel. By fading into digressions from the channel center, it averages your entry price much better. When price does revert to the mean, you profit very consistently.
THAT my friends is the cash cow. Not trend following.
Why do I use trend following? It's only because the sideways systems don't work well during trends which happen infrequently and I want to make money every single month.
So that forces us to deal with the period where the markets become inefficient, the crowd mentality takes over and things go haywire.
When a trend starts it invariably catches the RTM channel-type system on the wrong side of the market.
So the trend following system kicks in to counteract the risk and convert them to profits.
In this way, the combined portfolio makes money "round the clock" so to speak.
You may find it really hard and nigh impossible to trade separate strategies simultaneously on the same symbol on the platforms mentioned here.
THAT is the real power here--merging strategies of different types on the same symbol.
So back to the trend system. Sure it costs some small losses to find a trend, but who cares compared to the piles of cash generated at the same time from the RTM strategy.
However, when the RTM strategy hits a trend, the trend strategy becomes a life saver--to be sure.
Still, in this holistic approach the trend doesn't produce a windfall. No. It simply counteracts the losses from the RTM strategy with a little profits to boot. So it mainly servers to keep up the 100% monthly profitability.
Oh gawd...
I'm getting tempted to go berserk on the OP regarding the content of the post. Anyways.... your interpretation regarding martingale is different from my understanding. Please post a link of the source you got your points.
greaterreturn, how's the TickZoom project going lately?
I'm getting tempted to go berserk on the OP regarding the content of the post. Anyways.... your interpretation regarding martingale is different from my understanding. Please post a link of the source you got your points.
greaterreturn, how's the TickZoom project going lately?
I think greaterreturn is blocked from posting anything about tickzoom until he becomes a paid sponsor. He said he'll be doing that in a couple weeks. I guess you can check the status at the tz forum by googling for tickzoom.
Quote from jjw:
in dice and roulette and other common gambling games (not black jack), the outcome of the next event is independent of the outcome of the current event. in trading, though one may feel that one's offer to buy or sell (the current event) is so insiginificant that it cannot affect the next event, it can and does. so the math that explains/preditcts common gambling behavior may not, and imho, is not really applicable to trading (the major assumtions about the interactions between events is different).
Most excellent point. That never occurred to me and seems often overlooked. It's even obvious when you think of very large traders who know they move the market when they trade.
But you're right. That must even apply to small traders while less so.
Quote from TSGannGalt:
Oh gawd...
I'm getting tempted to go berserk on the OP regarding the content of the post. Anyways.... your interpretation regarding martingale is different from my understanding. Please post a link of the source you got your points.
Wow. I don't recall the sources now. Somehow it seems I didn't get an email notification of your post and now I don't recall. I'm willing to concede the point about the definition of martingale. My post was mainly to discuss scaling in and out whatever you want to call it in mathematics.
Quote from WalterEdward:
However, as you point out, there are many different ways to do the scaling in and out, and I may need to work on those until I find the better way.
Correct. My scaling works differently for trends and then for counter trend or chop strategies.
In one case, I scale in and then gradually scale out also.
On another, it works better to jump in and then scale out. In others, it profits more to scale in and jump out.
Some others may have figured this out a much smarter way.
In fact, it may come down to simply measuring the probability and trading number of contracts based on that.
It seems that on a trend, a rally has low probability of being an actual trend so trade it with less contracts.
But if it has gone further than most rallies, the odds of a trend increase so add contracts.
My guess is that you stop adding when the trend gets to the "average" size of a trend?
Well, after some point in the trend, the odds of it continuing start to decrease.
So as it proceeds, it makes sense to start scaling out to lock in profits.
Mine is currently somewhat "hardcoded" to scale at different price levels.
But I'm realizing that may not be very portable to other markets than the one upon which I'm trading now.
Eventually, I probably need a tool which will look at new market over time and cull statistics about trend, chop and other periods that can be used to variably handle the levels for scaling in and out.
The point of my thread is simply to say that scaling into and out of positions can be a road to take an otherwise decent strategy and make it sing!
Quote from lolatency:
- Gay prostitutes and Automated Traders both need to assess risk. For Gay Prostitutes, it's the risk of AIDS; whereas for Automated Traders, it's the risk of the exchange pushing out bad packets with erroneous sequence numbers.
... i don't think you even need to analogize 'bad packets'.