How to Exploit Randomness

Quote from Butterball:

The poster assumes there are tradable (exploitable) trends in random data. There aren't.

Try it in Excel. Throw up 1000 random time series and devise a trading system that systematically makes money from the random data. Post your results.

Utterly Brilliant.
 
Quote from Butterball:

The poster assumes there are tradable (exploitable) trends in random data. There aren't.

Try it in Excel. Throw up 1000 random time series and devise a trading system that systematically makes money from the random data. Post your results.

From a purely mechanical point of view I agree.

But from my personal experience a human trader with enough screen time can cherry pick events. After awhile you just know whats noise and whats not.

Or maybe I am not thinking truly random. Too much screen time for me to believe anything in the mkts truly is I guess.
 
Quote from Butterball:

The poster assumes there are tradable (exploitable) trends in random data. There aren't.

Try it in Excel. Throw up 1000 random time series and devise a trading system that systematically makes money from the random data. Post your results.

The poster is correct and you are wrong. Trends in random data exist because people use trend indicators and patterns. Let me put it another way:

Patterns exist because poeple trade as though there exist patterns.

What you are missing is that which most people who do not understand the markets believe, i.e. that there is an independent random process that creates prices and people just trade them.

I want to inform you that it is people who create prices. Intraday you may be partially correct in that there is a lot of noise due to robots. But on a daily bar basis, things are very predictable for those who have market experience. What prevents them from becoming very rich is either their risk aversion (that is me) or their gambling style.
 
Quote from Roark:

Suppose you have been ship wrecked on an island. Desiring to return to civilization, you build a raft with a primitive sail to take advantage of wind power. Unfortunately, the wind blows randomly in two different directions. The first direction will drive your raft toward civilization. The second direction drives the raft further out to sea. What to do?

When the wind blows in the first direction, put the sail up. When the wind blows in the other direction, reef the sail and wait for the wind to change. And so it is with trading.

Setup a broker firm and start earning commision, or a HFT firm to capture all those nice slippage and scalping with the beauty of pure random.
 
"What prevents them from becoming very rich is either their risk aversion (that is me) or their gambling style."

Then by definition, to follow your logic into actual practive we must conclude that in order to find success in random noise requires the ability to endure drawdowns beyond your current capabilities. So, if one can take an extraordinary amount of heat or drawdown in a random market strategy is he or she bound to become successful ?
 
Quote from Butterball:

The poster assumes there are tradable (exploitable) trends in random data. There aren't.

Try it in Excel. Throw up 1000 random time series and devise a trading system that systematically makes money from the random data. Post your results.

And you are assuming that random data means i.i.d. It is not necessarily the case.

As for your experiment, how do you want to generate the RW's?

With a bias like markets have? Then a flipping strategy where you go long/short when crossing the origin will give you positive expectancy with a growth rate exponential to the bias.

Standard distribution? Then a strangle at +/- 2STD will give you positive expectancy.

In both cases you systematically make money from random data.

Ninna
 
Quote from Crispy:

"... I think many people overlook this cliche and disregard it

I think traders skip over the simple stuff in their rush to "get the the good (complex) stuff", never knowing they stepped right over the treasure.
 
Quote from Butterball:

The poster assumes there are tradable (exploitable) trends in random data. There aren't.

Try it in Excel. Throw up 1000 random time series and devise a trading system that systematically makes money from the random data. Post your results.

I agree. Anyone who doesn't see that...good luck to you
 
Pretty simple break the sail down and use the parts to create something that captures the energy of the wind, in either direction and attach that onto a mechanical device that drives the ship...


Create orders that take advantage of move in the direction in the market of either up or down, then all you have to do is spot out places where there is a good probability of volatility....Then you are not trying to guess market direction, but that the market is going to be volatile...


Dont care which direction the wind is going, all i care is its going to be windy.
 
Back
Top