Fully Automated Trading Journal - $50k per month goal

F

frostengine

Background
I have been designing automated trading systems for about 20 years. Most of my systems fail when transitioning from dev to production. However, occasionally I get a winner that continues to work with real money. I believe this system is one of those "winners" and will survive the transition. I spent most of Dec working out the kinks in the execution engine, and now I am ready to fully turn it on starting tomorrow.

This journal will document the ups and downs of bringing a new fully automated system live. Based on testing, I am expecting this system to generate roughly $50k per month depending on how many stocks/indexes I turn on.

Strategy
The system actually uses two distinct strategies.

Strategy 1 - trades a basket of roughly a dozen ETFs. The strategy shorts near dated options expiring in the next 10 to 17 days. There are 4 exit scenarios. If the shorted option has less than 30% of its original value, if the trade has been active for 5 trading days, or if a profit target/stop loss has been hit.

Strategy 2 - trades a basket of a few dozen stocks/ETFs. This strategy goes long options expiring in the next 21-28 days. This strategy has 3 exit scenarios. Similar to Strategy 1, there is a pre-set profit target/stop loss. In addition, the trade will always exit at 2:30 Central time the day after its entered if the other 2 conditions have not yet been met.

Both Strategy 1 and Strategy 2 have been trained on approximately 15 years of data. Each were designed independently with the outcome in mind. Meaning, Strategy 1 was trained/designed to identify opportunities where short options would be profitable while Strategy 2 was trained/designed to identify opportunities to go long options.

Live Trades
The point of having an automated system is to be able to get away from the computer every now and then and enjoy life. As such, an easy way for me to keep track of what positions my bot currently has open and real time P/L is to have my bot post to twitter. This allows me to easily check in on the bot from my phone.

Therefore, what started as a simple way to help me monitor my bot in real time - also provides an easy way for you to see the results "live". @DayTradingJoe is the twitter handle. https://twitter.com/DayTradingJoe

I am planning to go live tomorrow - Monday January 4th.

I will update this thread daily with results, issues, and new learnings.
 
gkishot said:
What is your capital?
More...

I am trading this strategy in an account that runs other strategies as well. This account is a portfolio margin account with roughly $735k currently. However only about $250k is being allocated to this particular strategy.

This account always holds a basket of deep out the money puts as part of a “barbell” strategy. This significantly reduces margin requirements for this and other strategies run from this account.
 
frostengine said:
This account always holds a basket of deep out the money puts as part of a “barbell” strategy. This significantly reduces margin requirements for this and other strategies run from this account.
More...

What is a "barbell" strategy?
 
It's pretty much guaranteed that all the strategies you laid out won't work in live trading. Here is why: you use static parameterizations that don't adapt to changing market regimes, volatility levels, and the like. Your strategy 1 basket is a time bomb. When markets move sharply to the downside your basket will be eaten alive and probably gobbles up the entire sum of gains you previously made plus some. If you combine strategy 1 and 2 you pretty much sent a Christmas present to your broker. You are short basket vol in one strategy and long vol in another. Thanks for the commission, buddy. To make money off such trades (strategy 1 and 2 combined) one needs to really have a deep understanding of correlation trading and the risk dynamics of basket options. It's a lot more mathematical, probably above your pay grade.

frostengine said:
Background
I have been designing automated trading systems for about 20 years. Most of my systems fail when transitioning from dev to production. However, occasionally I get a winner that continues to work with real money. I believe this system is one of those "winners" and will survive the transition. I spent most of Dec working out the kinks in the execution engine, and now I am ready to fully turn it on starting tomorrow.

This journal will document the ups and downs of bringing a new fully automated system live. Based on testing, I am expecting this system to generate roughly $50k per month depending on how many stocks/indexes I turn on.

Strategy
The system actually uses two distinct strategies.

Strategy 1 - trades a basket of roughly a dozen ETFs. The strategy shorts near dated options expiring in the next 10 to 17 days. There are 4 exit scenarios. If the shorted option has less than 30% of its original value, if the trade has been active for 5 trading days, or if a profit target/stop loss has been hit.

Strategy 2 - trades a basket of a few dozen stocks/ETFs. This strategy goes long options expiring in the next 21-28 days. This strategy has 3 exit scenarios. Similar to Strategy 1, there is a pre-set profit target/stop loss. In addition, the trade will always exit at 2:30 Central time the day after its entered if the other 2 conditions have not yet been met.

Both Strategy 1 and Strategy 2 have been trained on approximately 15 years of data. Each were designed independently with the outcome in mind. Meaning, Strategy 1 was trained/designed to identify opportunities where short options would be profitable while Strategy 2 was trained/designed to identify opportunities to go long options.

Live Trades
The point of having an automated system is to be able to get away from the computer every now and then and enjoy life. As such, an easy way for me to keep track of what positions my bot currently has open and real time P/L is to have my bot post to twitter. This allows me to easily check in on the bot from my phone.

Therefore, what started as a simple way to help me monitor my bot in real time - also provides an easy way for you to see the results "live". @DayTradingJoe is the twitter handle. https://twitter.com/DayTradingJoe

I am planning to go live tomorrow - Monday January 4th.

I will update this thread daily with results, issues, and new learnings.
More...
 
DiceAreCast said:
It's pretty much guaranteed that all the strategies you laid out won't work in live trading. Here is why: you use static parameterizations that don't adapt to changing market regimes, volatility levels, and the like. Your strategy 1 basket is a time bomb. When markets move sharply to the downside your basket will be eaten alive and probably gobbles up the entire sum of gains you previously made plus some. If you combine strategy 1 and 2 you pretty much sent a Christmas present to your broker. You are short basket vol in one strategy and long vol in another. Thanks for the commission, buddy. To make money off such trades (strategy 1 and 2 combined) one needs to really have a deep understanding of correlation trading and the risk dynamics of basket options. It's a lot more mathematical, probably above your pay grade.
More...

I appreciate the feedback. This is not my first rodeo. To make money with options you need to either correctly predict volatility or predict direction. My two strategies are not "mechanical" sell X options every N days.

The strategies are designed to identify direction first, and volatility second. I admit they are somewhat weaker on the volatility predictions but are golden on direction. The one thing I have gone back and forth over is if I should trade the shares straight up instead of options (since the direction prediction is very accurate). It would be a very profitable strategy in its own right.

However, my goal is to smartly leverage the strategy up - and the way I intend to use options gives me the best shot.
 
frostengine said:
I am trading this strategy in an account that runs other strategies as well. This account is a portfolio margin account with roughly $735k currently. However only about $250k is being allocated to this particular strategy.

This account always holds a basket of deep out the money puts as part of a “barbell” strategy. This significantly reduces margin requirements for this and other strategies run from this account.
More...
I think you will be the first man on earth who would be pulling off 20% monthly profit in trading markets.
 
You must log in or register to reply here.
Back
Top