Here is one suggestion. If we know where is the low of a particular stock e.g. Agilent(A) Close low on June 27th, 2022 and scan for all the lows for all the stocks. We can monitor where are the lows each day. If the lows stay further in the past, then the market is trending up. If more new lows are forming, then the market is trending down. This can be done for sectors and industries.
What's a reasonable way to approach the subject? There was a suggestion later in the thread to look at the Ag thread. Is this a good start or is there more of a higher level perspective to approach the subject matter with?
So your approach was bottom-up through observing PA? As an intraday/swing trader, do you have any data on Time of Day and Day of Week tendencies?
Look up Jake Bernstein. I don’t know if he still does but years ago he had piles of seasonal patterns he wrote about.
Do not dive deep into the Almanac, because it is not worth the effort. Most of it does not work out-of-sample, except the few things I already mentioned here. If you want to read more about it, I would look into academic papers. There are plenty of studies here about this topic. E.g. 'Are Monthly Seasonals Real? A Three Century Perspective', etc...and many more.