project no code day 18 - workshop sessions
first, I'm hosting a how to session 5June at 530PM PT. if you'd like to join, you can set a reminder here:
https://t.co/2uTF1fGcK8
this is a perfect place to pop in if you have questions, want to understand the project better and most importantly define a clear roadmap for yourself to get going.
second opportunity is joining one of the zoom workshops for my paid community, Outlier Pro. the Pro Plus tier meets 2x/mo for a 90 minute zoom workshop - after the launch of project no code, we are going to focus the workshops on each member building their own o/toc side by side.
i am opening a couple seats to others who'd like to participate. if you'd like to join in, simply shoot me an email with what you're wanting to research.
I will reply to the winners (it's drawn at random) by 12Jun at 5pm PT.
within 20 days, ive been able to stand up a complete end to end internal database, consisting of ~46 tables with 56 views over over 1Bn rows of data.
everything im doing now is expanding capabilities, building more tools, and tightening up the infrastructure. today, i spent time setting up my GPU to leverage the blackwell capability to 30x compute heavy workstreams.
I still am adding more data as well, not quite fully current yet. right now I'm pulling over 33,000 more tickers to capture literally the entire stock universe from 2000. this isn't just completionism - it's a perfect bedrock for testing survirship bias free ideas.
project no code and the o/toc have been an absolutely eye opening project for me. it started as scoping out something to share with the community and has genuinely morphed into what i actually will be using for my personal trading operations.
more to follow!
project no code day 17 - getting close to PNC wrapping and the O/TOC being fully built.
**I am hosting a zoom workshop on 15June at 5pm PT. I will work through project no code alongside you and a few others to answer questions, help with design planning, etc. if you are interested in joining, please shoot me an email.
today was hanging two larger projects:
1. ISR (military for intelligence, surveillance, recon) - this will be where im conducting live analysis of markets (vs historical research). things like breadth dashboards, momentum tracking, etc.
2. NEXUS this will be a repo level locator that will track and log ALL research on the roadmap, in flight, completed, or archived. it will also have an external outreach to conduct comprehensive research external from the O/TOC on the specific topic.
completed a massive re-org to streamline data ingest.
i am moving my own processes from my own db onto the O/TOC setup so we all will be working from a common language, space, setup, etc.
onward!
project no code day 17 - getting close to PNC wrapping and the O/TOC being fully built.
**I am hosting a zoom workshop on 15June at 5pm PT. I will work through project no code alongside you and a few others to answer questions, help with design planning, etc. if you are interested in joining, please shoot me an email.
today was hanging two larger projects:
1. ISR (military for intelligence, surveillance, recon) - this will be where im conducting live analysis of markets (vs historical research). things like breadth dashboards, momentum tracking, etc.
2. NEXUS this will be a repo level locator that will track and log ALL research on the roadmap, in flight, completed, or archived. it will also have an external outreach to conduct comprehensive research external from the O/TOC on the specific topic.
completed a massive re-org to streamline data ingest.
i am moving my own processes from my own db onto the O/TOC setup so we all will be working from a common language, space, setup, etc.
onward!
project no code day 16 - the cleanup
the O/TOC is built functional and ready for follow on actions. im incrementally improving - tying in a little ML, monte carlo piping, R, etc.
some fun initial research:
- do insiders position before FOMC? > yes but not enoigh of an effect to generate alpha (btw the majority of things you research will land like this - cool but not directionally actionable - remember we're digging for gold here)
- ML forward vol in SPY and feature decomposition (no shock, VIX with top SHAP value - it's literally SP vol)
- quick and dirty 2Y SPY .30 rolling delta short put strat
watch the video, check out the free guides and start digging - have fun!
https://t.co/dDCAdOomsC
https://t.co/9tTDU5sqSV
project no code day 16 - the cleanup
the O/TOC is built functional and ready for follow on actions. im incrementally improving - tying in a little ML, monte carlo piping, R, etc.
some fun initial research:
- do insiders position before FOMC? > yes but not enoigh of an effect to generate alpha (btw the majority of things you research will land like this - cool but not directionally actionable - remember we're digging for gold here)
- ML forward vol in SPY and feature decomposition (no shock, VIX with top SHAP value - it's literally SP vol)
- quick and dirty 2Y SPY .30 rolling delta short put strat
watch the video, check out the free guides and start digging - have fun!
https://t.co/dDCAdOomsC
https://t.co/9tTDU5sqSV
project no code day 15
grab your free quick start guide, zero strings attached - not email needed, not a lead magnet, etc. get to work
https://t.co/6GfIYlDqbP
Life insurance continues to be one of the strangest scams to me.
I can understand it for the middle - lower income family trying to get by.
Outside of that - waste.
Lot of chatter on this one. It's a nice example of something that can be reasoned through via derivatives replication.
Ie what's the cash flows of taking the annuity, shorting a perpetuity, and using the cash from the short to buy life insurance?
project no code day 15
grab your free quick start guide, zero strings attached - not email needed, not a lead magnet, etc. get to work
https://t.co/6GfIYlDqbP
project no code day 14
mostly housekeeping, the project is nearing it's natural end. that means im happy to announce the birth of the O/TOC (Outlier Trading Operations Center).
for those that missed the primer livestream, I highly recommend you check it out. this lays the foundation for the video launching tomorrow and prempts the companion guides
https://t.co/9tTDU5sqSV
project no code day 14
mostly housekeeping, the project is nearing it's natural end. that means im happy to announce the birth of the O/TOC (Outlier Trading Operations Center).
for those that missed the primer livestream, I highly recommend you check it out. this lays the foundation for the video launching tomorrow and prempts the companion guides
https://t.co/9tTDU5sqSV
project no code day 13 - quiet day [which is good]
AI primer livestream @ 1730PT: https://t.co/ZTb8yvIaea
> this livestream is a intro to using AI in a trading workflow, designed to lay a foundation for the project no code video release on Sunday (which has a 23pg quick start guide downloadable by everyone to start your own version of this project. no email wall, not a lead magnet, simply download from google drive no strings).
finalizing the pull for intraday options data for the covered universe. should finish in about an hour then will build tables.
ive been building out the research pipeline which will be a really cool next series to the segment.
This week on The Options Trench
@KrisAbdelmessih and I revisit GEX and DEX. What they are. How they MAY be used. Their significant limitations. And how to test some ideas using them as signals.
Tune in!
https://t.co/TFQ2F8i1WP
project no code day 13 - quiet day [which is good]
AI primer livestream @ 1730PT: https://t.co/ZTb8yvIaea
> this livestream is a intro to using AI in a trading workflow, designed to lay a foundation for the project no code video release on Sunday (which has a 23pg quick start guide downloadable by everyone to start your own version of this project. no email wall, not a lead magnet, simply download from google drive no strings).
finalizing the pull for intraday options data for the covered universe. should finish in about an hour then will build tables.
ive been building out the research pipeline which will be a really cool next series to the segment.
project no code day 12 - the [good] end is near.
- im working on documentation to share, there will be a few versions.
- the community quick start guide will land around 5 pages and be designed to help you start the project without overwhelming.
- for pro members im doing a few higher touch versions, these are more for convenience than anything.
- on a recent member call I reviewed a real piece of research where they used claude and it was absolutely awesome. a legitimate stepping stone piece of research that contains actual edge. this is the cool stuff.
- this AI project is a literal game changer. I finally get to start having true dialogue with the community on a technical level and everyone will be able to show up at the same level.
- for those wondering, so far on the project I've spent probably $6K on APIs, storage etc. im doing everything from scratch and building a full stack at a true professional level.
- from the start of the project on 18May Ive logged over 160 hours on it (easily averaging over 15 hours a day with several no sleep days).
- note this is NOT the cost it would take for you or the time. the entire point if i am testing TONS of things - im getting duplicative data from different sources to assess quality differences (different API vendors, bulk downloads, etc) and i am meticulously exploring the capability of all the different agents im using, infrastructure tools, etc. all of it is being logged into my SOP im over 4300 lines AFTER a recent cleanup.
i reordered my videos, the Project No Code launch video was recorded last night and will release this coming Sunday. too excited to wait.
for all the traders thinking your job is uniquely challenging or difficult.
a quick reminder, stop putting yourself on a pedestal. trading has difficult elements like any other job.
one of the biggest disservices you do to yourself is pretending like it some uniquely challenging thing. it's not.
be a professional and do the fucking work.
Ukrainian soldier records what he thinks will be his final video as his unit is surrounded by Russian forces.
Look at the eyes of the soldier, you will feel this video.
project no code day 12 - the [good] end is near.
- im working on documentation to share, there will be a few versions.
- the community quick start guide will land around 5 pages and be designed to help you start the project without overwhelming.
- for pro members im doing a few higher touch versions, these are more for convenience than anything.
- on a recent member call I reviewed a real piece of research where they used claude and it was absolutely awesome. a legitimate stepping stone piece of research that contains actual edge. this is the cool stuff.
- this AI project is a literal game changer. I finally get to start having true dialogue with the community on a technical level and everyone will be able to show up at the same level.
- for those wondering, so far on the project I've spent probably $6K on APIs, storage etc. im doing everything from scratch and building a full stack at a true professional level.
- from the start of the project on 18May Ive logged over 160 hours on it (easily averaging over 15 hours a day with several no sleep days).
- note this is NOT the cost it would take for you or the time. the entire point if i am testing TONS of things - im getting duplicative data from different sources to assess quality differences (different API vendors, bulk downloads, etc) and i am meticulously exploring the capability of all the different agents im using, infrastructure tools, etc. all of it is being logged into my SOP im over 4300 lines AFTER a recent cleanup.
i reordered my videos, the Project No Code launch video was recorded last night and will release this coming Sunday. too excited to wait.
project no code day 11
playing around with some basic studies to kick us off. all those times someone is calling for a crashing or pending doom - remember..
the market spends 67.6% within 10% of all time highs. 55% within 5%. 39.1% within 2%.
the cool thing? you will be able to do this EXACT research using ZERO code just like I did.
all built by AI based on a basic prompt.
AI intro video will launch next wednesday and the project no code video will launch the following wednesday.
im giving myself plenty of time to get far out ahead on the project so I can experience all the gotchas so you don't have to.
project no code day 11
playing around with some basic studies to kick us off. all those times someone is calling for a crashing or pending doom - remember..
the market spends 67.6% within 10% of all time highs. 55% within 5%. 39.1% within 2%.
the cool thing? you will be able to do this EXACT research using ZERO code just like I did.
all built by AI based on a basic prompt.
AI intro video will launch next wednesday and the project no code video will launch the following wednesday.
im giving myself plenty of time to get far out ahead on the project so I can experience all the gotchas so you don't have to.
project no code - day 10 2225PT
project update up top, some general reflections below for those interested.
updates.
- im piping in options intraday data back to 2015 with full greeks, sill WIP. it's a lot, over 45TB of data. takes some time.
- i reconfigured my data footprint to accommodate having a mirror backup
- now that ingest is largely in progress and stable, im simultaneously working on a lot of qol updates. built a new interactive command menu that provides a ton of project level info quickly. useful that we're almost at 1B rows of data
- i have a queue of community requested research, keep sending them in. I have an ingest scraper that pulls them into a roadmap that i review. while my goal is to help you build your own infrastructure to test ideas yourself, running tests for you guys hopefully will inspire you to take the leap.
reflections.
theres a really fun part of my brain where once i get into something it becomes my obsession. this project has hit that threshold for me. its awesome. i really enjoy learning and this has opened a completely different world for me.
as ive noted, i already built my own db long ago. but it was really cobbled together based on what i learned through self teaching - same for my coding.
i started this on a whim exploring the perspective of a trader with zero coding experience leveraging AI. fast forward and i recently made the decision to actually put my own db on ice (a unequivocally completely inefficient choice) to fully lean into turning my operation onto proj no code.
before, when i would explain something and reference "my db" now I can reference OUR db - because this is truly something EVERYONE now has direct access to because of AI.
im now at a point where i feel comfortable beginning to build the content to make a video diving into this entire thing, the goal being the first video being done before it's fully complete.
3June a video will launch introducing how you might consider leveraging AI in trading beyond the surface level shit of "automate trading" or "analyze XYZ" - it can do SO much more. im focusing on the RESEARCH angle to begin.
the week after will be the project no code guide.
my vision for pnc is for this to serve as a mutual foundation for us to explore the markets together. learn how to conduct real analysis that's meaningful - vs overfit bullshit.
if you cant tell, im stoked. im excited to continue learning myself but even moreso because im fully aware how impactful this will be to people. think of it like if you were introducing someone you cared about to the internet for the first time. that kinda vibe.
project no code - day 10 2225PT
project update up top, some general reflections below for those interested.
updates.
- im piping in options intraday data back to 2015 with full greeks, sill WIP. it's a lot, over 45TB of data. takes some time.
- i reconfigured my data footprint to accommodate having a mirror backup
- now that ingest is largely in progress and stable, im simultaneously working on a lot of qol updates. built a new interactive command menu that provides a ton of project level info quickly. useful that we're almost at 1B rows of data
- i have a queue of community requested research, keep sending them in. I have an ingest scraper that pulls them into a roadmap that i review. while my goal is to help you build your own infrastructure to test ideas yourself, running tests for you guys hopefully will inspire you to take the leap.
reflections.
theres a really fun part of my brain where once i get into something it becomes my obsession. this project has hit that threshold for me. its awesome. i really enjoy learning and this has opened a completely different world for me.
as ive noted, i already built my own db long ago. but it was really cobbled together based on what i learned through self teaching - same for my coding.
i started this on a whim exploring the perspective of a trader with zero coding experience leveraging AI. fast forward and i recently made the decision to actually put my own db on ice (a unequivocally completely inefficient choice) to fully lean into turning my operation onto proj no code.
before, when i would explain something and reference "my db" now I can reference OUR db - because this is truly something EVERYONE now has direct access to because of AI.
im now at a point where i feel comfortable beginning to build the content to make a video diving into this entire thing, the goal being the first video being done before it's fully complete.
3June a video will launch introducing how you might consider leveraging AI in trading beyond the surface level shit of "automate trading" or "analyze XYZ" - it can do SO much more. im focusing on the RESEARCH angle to begin.
the week after will be the project no code guide.
my vision for pnc is for this to serve as a mutual foundation for us to explore the markets together. learn how to conduct real analysis that's meaningful - vs overfit bullshit.
if you cant tell, im stoked. im excited to continue learning myself but even moreso because im fully aware how impactful this will be to people. think of it like if you were introducing someone you cared about to the internet for the first time. that kinda vibe.
project no code - day 9.
if you have any questions on the project in the meantime - let me know below. my biggest advice for anyone even remotely interested in trading (stocks or options) is to recreate a MVP version of this - it's really easy and worth your time.
the database has full EOD options coverage back to 2007, full greeks. we have 1min interval stock data (ohlcv) back to 2000.
with full economic, BLS, fed, insider, etc.
with base coverage built, i am now backfilling 5 minute interval intraday options data for core and tier 1 tickers in the universe.
we are at the point of refining and now the fun of generating test ideas.
project SOP is still being maintained and is now over 3500 lines. i intend to release a simplified version that everyone interested can follow.
it has never been easier to leverage AI to massively update your research game.
project no code - day 9.
if you have any questions on the project in the meantime - let me know below. my biggest advice for anyone even remotely interested in trading (stocks or options) is to recreate a MVP version of this - it's really easy and worth your time.
the database has full EOD options coverage back to 2007, full greeks. we have 1min interval stock data (ohlcv) back to 2000.
with full economic, BLS, fed, insider, etc.
with base coverage built, i am now backfilling 5 minute interval intraday options data for core and tier 1 tickers in the universe.
we are at the point of refining and now the fun of generating test ideas.
project SOP is still being maintained and is now over 3500 lines. i intend to release a simplified version that everyone interested can follow.
it has never been easier to leverage AI to massively update your research game.
project no code day 8
today was a bit more of admin that meaningful database building or research.
prior to launching pnc, i already ordered a new PC - i do this every few years as technology tends to rapidly change. today was me migrating from my old machine to new - which actually highlighted some cool elements of the project.
AI was smart enough to include some key wrappers that allow me to stop any ingest cleanly, then restart at any time. AI also made sure there was a clean sequence of events to complete the cutover.
I used AI to handle most of the lifting, recreating folders, etc and was able to get onto the new machine quickly. the photo in the last post was me adding more SSD to have a larger datapool for larger projects.
i built a home database some time ago and it lives on it's own machine - however for PNC im running it end to end with no shortcuts. that includes obtaining a lot of duplicative data and the requisite storage requirements.
the next couple days will be building out more options data for testing, i have the deep tail fully filled from 2007 to 2012.
im also wrapping up grabbing stock data in minute intervals for other general testing.
I'm tracking your submitted ideas in it's own markdown for me to iterate on as i get the db built out.
the daily log is maintained, along with ai turnover guide, sop, etc. that documentation will provide the framework for others to follow along if they want.
getting the pipelines full for the night and back at it tomorrow.