Pedma

last time crypto had good general performance, or what old schoolers call "easy mode", was summer 2023 - q1 2024. that's it. it's kinda insane thinking it was 2 years ago. yes some periods here and there but very idiosyncratic like memes, perp dexes, btc, eth, and not much else.

such a bad take saying "why do I need to learn xyz topic if AI can do it for me".
yeh, you're getting a rude awakening pretty soon when you start working on real stuff.
speed of building stuff is not the same as understanding the thing you're building.

last time crypto had a good performance generally, was summer 2023 - q1 2024. that's it. it's kinda insane thinking it was 2 years ago. yes it had some periods here and there but very idiosyncratic like memes, perp dexes, btc, eth, and not much else.

ahhh getting sick after I spent the entire winter telling people that got the flu that I was immune and invencible to the flu.
seems like bragging about equity curve ath's.

didn't expect to like this book this much.
genuinely impressed.
might not be a topic for everyone, but personally, I find it quite interesting.

every day is a fight to avoid producing and consuming slop. slop in the form of work, slop on the form of content, slop everywhere. creating an environment that allows to focus on deep, good work, that you hold to a higher standard.
but while allowing a bit of randomness into the pursuit. because off that randomness, sometimes decent signal is found. just don't be consumed by slop.

I really have no respect for TV analysts, for the most part. arrogant, clueless, ignorant idiots with no real world achievement or risk. just bc you have a degree and use fancy words don't mean you're smart. I respect most professions, at least the ones I know, but not analysts.

every time I have a feel to sell my hype bag and move on with my life, hype pumps. lesson in there.

as I scale my trade infra, I increasingly become aware of how important it is to have your code documention in order across all processes such as:
> data fetching, cleaning, processing and storage
> signal processing
> backtester
> live execution engine
> live trade data cleaning and storage
> external risk controls
> performance benchmarking
with enough detail so it doesn't become a mess every time there's an error.
don't neglect the boring part of producing clean code that you understand.
do not outsource your peace of mind to an LLM because it will eventually come back to bite you, and prob

reminder to self that theres only 1 priority on the table right now.
research efficiency. how to go from idea to prod on an acceptable time period. how to maximize iteration of ideas on a weekly/monthly basis.
focus on that, ignore all else.

Imma start saying to every business owner in the EU that complains about taxes "skill issue".

I'm gonna switch from samsung to google pixel phone next.
don't even say apple pls, who would buy that crap anyways?

one thing I need to get better is answering DM's or private messages in a timely manner.
I often leave people on hold because I can't answer straight away and then something else comes up or I forget.
I don't like that. I need to to improve on it.

some days I have a dull task I need to get on with , and I procrastinate on it because I know how dull it is. but the more I procrastinate on it, the more I delay the important stuff I have to work on. just get it done with, spend the day on it, so you can carry on tomorrow.

registering an account in crypto exchanges under my company's name in Europe.
government: "oh you have to pay this €200 fee for this license for AML"
me: "ok, so what information you need?"
government: "no information, just sign here, pay and stfu."

sometimes I want to understand all the concepts and terms macro tradoors use to sound smart on the timeline while making no money.

how does one know if he really paid a fair price, even if good relative to mid, on an illiquid perp, where the underlying is closed? you're just paying whatever the market says is fair at that time, not the actual fair, even if your execution was great.
what I mean is, if I know btc is worth 90k today at major exchanges, and I pay 90k for it, I know I got a mid deal (pun intended) for it.
but if I'm trading AAPL perp and AAPL is not open like if it's weekend, and you're trying to get a good fill, you might get a good price relative to mid at that time, but that mid is irrelevant in the whole "

practice creating good, scalabale code, even on projects that are meant as a oneoff.
first, those oneoffs sometimes become good ideas and then its a mess to work with.
and finally, it helps us build the habit of creating good code that translates into the main projects.