Exactly! To share some insight into how my algo and sizing engine works: for absolutely every system I have, and within that system every variation I trade of that system, I have quantified the strength of the edge (there are many ways to do this, I use one approach). From there I know exactly the quantified strength of the edge for that trigger as my algo is executing on it.
From there, I use this number as part of my sizing engine to calculate the number of shares I need to take based on the price of the stock at that moment, the equity of my account, and the strength of the edge. This is all done automatically and is fully systematic. For STAK, my first entry prior to the halt had about a 2.6 edge score, which made me size in about 1X number of shares for simplicity sake.
Then once it gapped up massively after that halt, it triggered a different system where the edge score was about 10 (!). As such, my system laddered into this mega high edge entry. But since I already had a position on the stock, my system had to deduct some shares from its risk budget into this new entry so it doesn't take it as a fully independent trade - but since the strength of the edge after the gap up was so high, it still sized in about 9X of what I did prior to the halt.
The key concept I strongly advocate for is to only enter into the same ticker on the same directional bet if the edge score your signal is firing off is greater than what your current position is, and only size in using a dedicated risk budget. Even if you do believe your systems are uncorrelated, if it's the same ticker, do NOT ladder in (and add add add) into the same ticker for your different systems triggering on the same stock unless the edge score is greater!
Mega day to close out the week! Almost 10% account growth just on STAK alone. Sized heavy on it once all the shorts blew up on the halt resumption gap up. OMH was a nice trade too!
Been super busy building out my algo, which I've had live trading on my DAS account in 1-share mode for a few weeks now - fixing bugs, gathering data, tweaking things - all while I have beeen discretionary trading on top of it in the same account.
Have nonstop been working on it with my agents shipping between 40-60 PRs per day, and I am getting very close to flipping the switch to full real size. Currently splitting my agents in half right now: one side doing nonstop research, the other half working on my algo. Once it's stable and trading real size, I'm going full research-maxxing with all my agents.
Also, praying for a usage reset on codex and claude code this weekend. Completely drained all my usage across all my 20x subs even after all the resets we have been getting, and can't wait to try out Opus 5 !
It might look like I blindly martingaled it, but this was a fully systematic short where I simply laddered into a much higher edge setup which is how you should trade these optimally, for which I have very good data which is a very different thing!
I trade almost purely systematic. Blindly martingaling every stock discretionarily when you have no idea what you are doing and no data is how you blow you. And if this is how I really traded all stocks, I would have already blown my account after all the 1000%+ moves the last few weeks/months.
It is crazy to think about. These criminal riggers have gone beyond trying to just squeeze shorts for quick profit. Their new motto is to not only squeeze shorts but also deliberately manipulate these stocks to heights where they deliberately get them T12 halted, so they can trap shorts to pay 500-1000% short interest fee.
My only question then is if that float was really only 400k, how come the locates were so cheap in after hours that day then?
Ending June with a mega 130% account growth this month AFTER all the absurd locate fees and many many mistakes from my side!
Biggest single month account growth for me, and this is even after I cut my size in half due to the madness earlier this month, which I dont plan on changing. If I ran optimal Kelly for my strategies, it could have easily been at 300+% account growth this month, but for me this is a marathon, not a sprint, so I prioritize reducing volatility in my equity curve over maximizing profits in the short term.
The last few weeks I've been building out the infra for my algo with about 30 different systems, all shorting small caps. Hoping to have that running in PROD in about 1-2 weeks. Really happy with how things are going here. Really doing a lot of crazy stuff here I doubt few are doing, and can't wait to have everything running fully automated soon! With that wrapping up, I'll rarely be sharing charts going forward.
Can someone who's used @TradeThePool explain to me how this isn't free money if you're a profitable short seller?
You don't pay any locate fee whatsoever, and you can swing short positions without paying any fee forever?
And if you ever do get caught in a pre-market 1000%-in-1-minute teleport candle or a ZJYL-type move, since it's not your money you don't lose anything?
Sign me up!
Sizing in that much on a single sample is very stupid if you treat trading like a business, as you should.
Optimally you follow something similar to the Kelly criterion. From my side I use a custom self made version of a Kelly criterion sizing engine running very conservatively for low variance stable account growth.