How to develop "second-level thinking"

My experience its not like s&t guys are necessarily taught how to do that, just banks select s&t juniors based on their interest and ability for that type of analysis, and then they get to practice it every day (rather than fixing typos in powerpoints). So if you want to get better just practice more - e.g. for macro take one or two issues in the news that interest you/you ahve already read a decent amount about, big or small, e.g. consumer inflation, steel prices, or rural real estate, and spend 30 minutes writing down what you think is going to happen for each one on some timescale. Keep it decently short, like 5-6 sentences on each. Doesn't have to be perfect, and you dont have to fully justify everything, but make sure you know what you aren't justifying. You're trying to figure out what are the most important knowns and unknowns driving the process:

random example: city real estate prices - I know nothing about this just to show you what I mean:

• Assuming gas prices continue to remain high or rise (you might have read about this), driving a car will become more expensive. This will hit lower income rural workers who rely on cars to get to work the worst.
• Additionally, cost of living in rural areas has increased due to supply chain issues /inflationary pressure, with rural wages slower to catch up than city ones. 
• Combining these two together, I'd expect a large demand side pressure on major city real estate in the next few years as these workers (more numerous than the WFH white collar jobs) are forced to look for better paying jobs or jobs which do not require a car commute. 
• Construction work has been so disrupted by covid they will not fully recover for a few years, adding supply side upwards pressure on prices. 
• Hence I believe housing prices in major cities will continue to rise.

Point is you are writing down a prediction and some rationale for that. Do this every few weeks / when you have a new idea and a) you'll get better at making predictions b) when you read the news you will be synthesising it within the context of your previous predictions. Does this story confirm your prediction or not? If the head of real estate at XYZ bank is saying the opposite of what you predicted, you should figure out why.