I swear I read this section three time and I pick up facts here and there but I’m having trouble tieing it all back together with the rest of the Porftolio theories. From what I gather…
APT is another multifactor model, just like Macroeconomic, Fundamental, etc. The difference is though it is NOT based on a regression, rather it is derived from an equilibrium pricing model. As compared to CAPM, it hypothesizes that there are multiple factors that explain an asset’s return, whereas CAPM just comes down to sensitivity to market risk premium. But here’s what I’m having trouble connecting…

That example you always see with 3 asset choices, and if you take some combination of 2, you get the “same” beta of the 3rd but with higher returns  what does that have to do with APT?

And then later when the curriculum goes into Active Portfolio management, specifically Pure Factor Porfolios, how does this tie with APT? Each "lambda’ in APT is the expected risk premium from a pure factor portfolio where the sensitivity of the portfolio = 1.0 to only a single risk factor, and all other sensitivities are 0 to remaining factors. Is APT then an example of a Pure Factor Portfolio?

I guess in general I’m having trouble “visualizing” how one would USE APT? CAPM and all the other mutliregression based estimates seem somewhat intuitive but for APT, no idea how one is supposed to use it in asset selection…
Thanks in advance!