No. of Recommendations: 1
for individual stocks, there has been some compelling chat around doing a rapid screen with just 2-3 public high conviction ratios.(justifying which would be an interesting discussion)
then spend minimum time reading on the biz, specifically looking for discard points. use AI last for pre-defined redflag checks, not for positive bias.
any further time spent assumes one has much greater skill at future forecasting than the best at such craft.
(the idea is that natural species adapt to changes coarsely, not finely. may be stretching the analogy here.)
note this is very different from a quant factor zoo approach working via large numbers in a portfolio.
anyone been using this a standard approach when in the bargain bin?
