About this template
“TooGTBTrue” - bot intro
This bot is an evolution of Lealwei’s and Diego’s outstanding work on 0DTE strategies and their generosity to share their ideas and expertise in the community.
This bot is a high risk strategy where a couple of bad days can wipe out all of the allocation.
The concept:
This bot aims to grab some of the ripe fruit at the end of 0DTE days in SPY if conditions are tasty. The idea is to leave everything behind that is below ‘average profit’, play the decay a bit longer and wait for higher profits towards the end of the trading day. Though profit taking becomes more strict after 1pm as often the volatility increases in the afternoon. Hanging in there longer obviously needs a focus on controlling losses to not get wiped out with a row of bad days. Inspired by 0TC from Tammy Chambless (https://www.youtube.com/channel/UCGzfu0Q8QtSJvfwJlJ0t-0g ) the OA feature (chance of ending ITM) is a core element of this strategy to monitor if a position is within the desired range or not.
The past performance
QQQ started mid September. Old version that has not received an update but shows what a lucky number of trading days can do and what some not so fortunate days can take away. Obviously these few days don’t give a meaningful picture of the bots overall performance. For the latest updates I don't even have a history of more than a couple of days yet.
An example of a perfect day
A horrible day Nov. 22th
Everything got stopped out within at around 30 to 35%.
A good day for the strategy
SPY comes back into the ‘sweet spot channel’ and bot closes positions
FED day Nov 2nd
Bot description
The Scanner:
Sells four close to the money Iron Condors’s of SPY at 9:40 on 0DTE days. And again four close to the money Iron Condors’s of SPY at 9:45 on 0DTE days. All of them have a -40% 1 minute stop loss which gives the positions room to swing and are only meant to protect from a black swan. Number one (out of each batch of four positions entered) has a 15% 1 minute profit exit, number two 25%, number three 35% and number four is the runner which hopefully makes it to the end to capture most of the decay. This narrow IC is much more easy to manage as the values for ‘chance of position ending In The Money’ are wider and not always in the high 90’s when using the classic IB’s.
Each position gets a 10% allocation of net liquid which totals to 80% and leaves some buffer to recover from bad days. The positions entered with smart pricing type ‘fast’ as ‘normal’ lead to lots of fills missed in live trading in the past.
Filters:
You may want to switch filters on although they have reduced the overall performance of the strategy in the past but have on the other side reduced wilder swings. The filter checks high VIX, RSI, gap from last close and intraday movement since open.
The Monitor:
The monitor tracks seven main variables in hourly settings. The values come from data collected for this specific type of option over the last months.
1. Absolute profit target for that hour.
Exit if the position has reached a ‘max profit’ which is significantly above the average for that timeframe.
2. Absolute stop loss.
Exit if the position has crossed a max loss value for that time of the day. Additionally a fixed 30% Exit from 1min checks.
3. Relative profit.
The settings for the ‘relative profit’ consist of two components.
A) A ‘min. profit’ the position needs to have at that time of trading day.
B) A higher than normal value of the ‘chance of ending In The Money’ for the short call OR short put side of the IC.
The initial idea behind this was - that if one leg already leans out of the channel (over a certain value at that trading hour), but there is still a profit that can be taken, why then take the risk of this forming into a trend and not take the position off with a small profit. Over time this morphed more into a full ‘chance of ending ITM’ checkpoint for that trading hour.
Otherwise (if none of the above criteria is met) - the position will stay in the ‘Hang in there’ channel trying to capture further decay.
4. Relative loss
Same concept as with relative profit catering to the idea that - If the chance of ending in the money is too high at that time of the day and the loss is still acceptable > rather call it the day than hoping for a turn around.
5. Two smart trailing stops
Number one catches some of the profit (if existing and guarded by a min threshold specific to the time of the day) when a position looks to turn sour. A second smart trailing stop kicks in after 11am for all time intervals thereafter which again builds on the ‘chance of ending ITM’ being uncomfortably high and return has lost min 10%.
6. Balance
Generally, when during the day, the price of the underlying ticker comes back to the level where the position was initially entered - this is a very good time to take profits as both sides are close to their sweet spot. To execute this:
A) the position needs to have a certain profit at that trading hour and
B) the ‘chances of ending ITM’ for BOTH short legs of the IC need to be under a certain threshold specific for that trading hour.
7. No knots
If none of the above is true the ‘No knot’ will send a 5 min. limit order which is a long shot and would simply be lucky if it got filled. Which actually happens more often than expected.
Regardless of the performance - all positions get closed at 3:45 and swiped again at 3:50 if close at 3:45 did not go through.
Obviously those components overlap heavily but decay throughout the trading day and ticker movement have also shown to be vastly different and uncorrelated at times when the Greeks start to move.
The bot was tested in many different versions. The version attached here was the most successful with regards to risk and return. All of them were built on data from SPY but until now have performed significantly better on QQQ. You want to give the bot a min allocation of 5k to open eight positions.
AGAIN this a 0DTE neutral strategy with a lot of potential but also high risk.
This bot was built on data collected throughout the last couple of months. The findings, settings and variables defined from this were done more on a trial and error basis than by a scientific approach like propper data analytics.
This bot has a lot of room for improvement and I hope that the OA crowd-intelligence will take it to the next level.
Credits go out to Larry from OA for his extremely helpful back and forth explaining in great detail and creativity how to fine-tune the position entry in OA system. And again Lealwai, Diego as ‘mentors’ for 0DTE strategies at OA. I am very grateful for all the expertise they shared in uncountable discussions, comments, explanations and their own creations and variations thereof. Without them I would probably still be sitting there frustrated, puzzled and locked - wondering why my 30DTE IC-bots are drip-bleeding to death day after day.
And as always - please rip this bot apart, try out different tickers, settings and option types and report back what you have found valuable to add or change.
Enjoy
