For this post, I will share my experience with stocks. Feel free to chime in and offer any advice or insight.
I’ve been listening to Rich, Emperor for awhile and Genesis mogul for extended period of time, awhile ago. I’ve been in and out of the stock game (day trading) never really clicked for me until today. My invested liquid is small and I plan to snow ball, so my daily gains aren’t as much as others, yet. Before the subs the concept and strategies were like another language for me, I’d lose money and get frustrated with the ups and downs. Today when I opened it, the information on the screen flowed like a beautiful symphony of notes, I knew when to buy and sell. Profit of $30. Again my capital is low but when I increase my shares that $30 will increase as long as retain the fundamentals. I am not a financial advisor but NU holdings is a solid stock with low risk level and price that is easy to learn the flows with.
To test it, and determine if it was just “luck” I did 3 trades. All profitable. Bar up down strategy, RSi, and BB Indicator. Buy on bar down sell on bar up
Indicators are important, helps take the guessing out of it. Don’t die and live buy them but they’re a game changer as the market becomes more readable
I’ve been dabbling as well… I’ve had a solid algo for trade entries and exits for months now, the part that I was stuck on was screening the stocks to be watching, though I’ve got that sorted out now I think.
I’ve been chipping away at my indicator code, trying to build a reliable ranging/consolidation filtering to weed out false signals. I go for high volatility trend following… non-trend trades are just throwing money away on commissions.
What’s up ice, sounds like you’re getting pretty deep into it. Have you noticed the subs effecting your trading ability? The vetting process is crucial for trades especially when starting with a only so much liquid to invest. Motley fool has never let me down, there is a free and paid subscription to stock recommendations with news and ect
I have RICH but TBH I haven’t listened to that in quite a while, this is just me applying pattern analysis to stocks… I don’t even look at fundamentals, it’s not all that relevant for scalping, other than seeing what the current price is relative to the 1W & 1M high/low/average to give me a sense of where it is on the macro scale. Everything else is is based on on-chart trend analysis, fully algorithmic…
My trade balance is only what I’m willing to risk, which is about $2k… That being said, it’s all just been paper trading so far. I’m not putting a dime into it until I get enough reliable data, especially if I’m just hitting go and letting it trade for me, haha.
I have a features list a mile long that I want to add to my indicator (then strategy), but some will need to be brought in-house, to some extent at least… I’ve got a semi-functional scraper built that grabs headlines from a Tradingview news feed (the market analysis recommendations one) that grabs any with certain keywords and passes the URL to another function that parses the headline and squirts it over to L370B on Groq for a quick sentiment analysis, and depending on that response, it opens that chart in a Selenium browser with my indicator on it… at least it’s supposed to… so far it opens a default TV window, not even in dark mode! with no indicators at all… so it’s a cookies/persistence issue of some sort, which makes sense… Opening an actual Chrome window on my PC would do it, but that’s not ideal… and I just haven’t bothered to look for any other workaround yet as the rest of the system is still tossing out too many false signals.
Anyway, I don’t mean to jump all over your journal - shoot me a PM if you’d like! I’m always up for chatting about trading.
You’ve got quite the set up. Also never a bother talking about this shit, I enjoy the conversation. Don’t get caught in the preparation trap though man, there’s money to be made. Play around with a cheap stock $10 or so. Get a feel for the tools
Still a work in progress, but the backtesting is looking not too bad. That max drawdown is cringe-y though… some of the pattern recognition for candles is clearly not working quite yet. TONS of noise… way too many in/out trades that just racked up commissions and drove down my averages, Sharpe ratio, etc…
But even so, it’s the net profit that matters… and 130+% in 2 months isn’t bad, IMHO.
Details: $2k starting balance, (paper) trading period May 1, 2024 through to today. Trading Dell on NYSE.
@BLACKICE I appreciate the energy, I do very much so enjoy stocks and I love enthusiastic conversation about them. You have a much more systemic approach than I do which is why I’m very interested to see the profit your accumulating. You’re doing it right, with back testing and applying all the principles. I use trading view and have about 3 strategies with one stock in my portfolio right now. However, I’ve increased profit margin by 60%. It’s finally starting to snowball and grow for more profit in each trade. I did take a hit on Monday I believe because I was distracted while on lunch break and did not read the chart correctly, bought in at the wrong time. That was completely user error, other than that I’m at a point where each trade is a profit. Which coming from someone who used to suck and feel out of control with trading feels great
They sure do. Most of where I’ve been stuck is accurately identifying signals during consolidation/ranging periods. Combined with higher-timeframe trend confirmations this should cut down on my in-out churn… but it’s more easily said than done, haha.
I’m all about systems, I don’t see any other way as being viable for me, at least. This is just the barest beginnings of what I want to build though, one of my earliest theories was that not only do different strategies work differently with different securities, they work differently at different times as well. And now that I’ve got the ability to backtest a strategy that works well with a variety of stocks, it’s very interesting to see how poorly it does on others that ought to be roughly equivalent.
I’ve got a ton of projects on the go at the moment, so I don’t make a ton of progress on any of them really… oops… but one of the to-do items on this project is to characterize the securities where this strategy works well and then find out why… and then test on other securities with similar attributes… is it price? is it some fundamental attribute (p/e, dividends, etc)?? Even looking at comparable stocks based on price in the same sector, the results vary wildly, and I don’t yet know why.
So my approach to that will be to pull in as much data as I can for the timeframe I’m looking at (3-min) for the ones that work… find correlations… find matches… pull in that data… backtest those… over and over until I find what I’m looking for or yahoo finance blocks the last of my proxies haha.
My hope is there is a fixed number of strategy parameters that match the majority of the securities I’d want to be trading (and I have a rather specific set of criteria, so there won’t be that many)… and when I want to trade one on any given day, I’d be able to just give the symbol to the script and it would grab the latest data, run through all permutations of the parameters, aka local multivariate backtesting… and then tell me within a few seconds what the ideal settings were at that moment. Sounds simpler than it is.
A much more ambitious approach is to fine-tune an LLM with stock data instead of written content… they excel at next token/chunk prediction based on what’s come previously, so what better choice to look at historical bars for any given security, compare it to the reams of historical data for that and other securities, and give a forecast for what will (probably) happen next?
