Thanks for your detail, but normally (1) is really more reliable than (2), but it is hard for an individual to do (1) with multiple future contracts. Even with mini futures, normally it really needs like 10 mini futures in market at the same time to create a pair trading.
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.
Ernie Chan claims that statistical arbitrage is better than momentum like trend
- Thread starter GloriaBrown
- Start date
I think about it and have this conclusion at this moment:
Statistical arbitrage:
pro:
-normally and theoretically more reliable than momentum way, of course how much more reliable would be case by case.
con:
-Request multiple futures like 10 contracts to create a true pair trading. Even with mini futures, having at least 0.5 millions US dollars cash just for one strategy would be required.
-day trade would requests HFT speed, not possible for normal personal trader.
-double commission comparing with momentum way, basically impossible for retail personal trader to have profit with retail commission fee in day trade time frame.
-math is harder than momentum
Momentum:
pro:
-can happen with small amount of money no matter day trade or interdays.
-strategies can be easy or complicated. The maths with statistical arbitrage has to be kind of difficulty even with the easiest way.
-Having much more live trade stats with day trade, then can tell if a strategy works or not in real life with limited of real money.
-day trade can be profitable without HFT speed
con:
-not as reliable as stats arbit
Statistical arbitrage:
pro:
-normally and theoretically more reliable than momentum way, of course how much more reliable would be case by case.
con:
-Request multiple futures like 10 contracts to create a true pair trading. Even with mini futures, having at least 0.5 millions US dollars cash just for one strategy would be required.
-day trade would requests HFT speed, not possible for normal personal trader.
-double commission comparing with momentum way, basically impossible for retail personal trader to have profit with retail commission fee in day trade time frame.
-math is harder than momentum
Momentum:
pro:
-can happen with small amount of money no matter day trade or interdays.
-strategies can be easy or complicated. The maths with statistical arbitrage has to be kind of difficulty even with the easiest way.
-Having much more live trade stats with day trade, then can tell if a strategy works or not in real life with limited of real money.
-day trade can be profitable without HFT speed
con:
-not as reliable as stats arbit
Quote from gmst:
This is your problem man that you "think"....you should rather take up the data and "test" and "interpret" the results.....that will give you 10x more insight than random discussion here
talk and think about what is a possible way is a step before real back test right? Most of these like it takes so much money to have a stats arbit trade is true even without backtest it.
feels like a turing test
Stat arb is for big pockets not for retail accounts.
Quote from GloriaBrown:
talk and think about what is a possible way is a step before real back test right? Most of these like it takes so much money to have a stats arbit trade is true even without backtest it.
I precise you. Let's talk, why not?
You have many question, Mr. Sparkle. I send you premium
stats-armpits, answer question hundred percent!
You have chosen wisely.
For lucky best backtest, use Mr. Sparkle, fishbulb backtester, buy from
Matsumura Fishworks and Tamaribuchi Heavy Manufacturing Concern.
Quote from Equalizer:
I precise you. Let's talk, why not?
You have many question, Mr. Sparkle. I send you premium
stats-armpits, answer question hundred percent!
You have chosen wisely.
For lucky best backtest, use Mr. Sparkle, fishbulb backtester, buy from
Matsumura Fishworks and Tamaribuchi Heavy Manufacturing Concern.
Oh my.
Thank you so much for that. I needed the giggle after reading this entire thread.
This is your problem man that you "think"....you should rather take up the data and "test" and "interpret" the results.....that will give you 10x more insight than random discussion here
I think random discussion is helpful sometimes.