'Expected Returns' is a gripping thriller of a book because it lives and breathes the unspoken mission (maybe subconscious?) of academic finance - to systematize and grill down, neatly, to a number of risk factors, what investors, whether fundamental or technical, have been doing for decades and have been attributing returns and beta as their alpha. After all, in the words of Antti - you can never fully replicate what's going on in the heads of a discretionary manager and there's always the room to take their word for it that their returns are their genius. But quant is unfairly called black box! Academia always has to lag practice in finance due to the inherent nature of things (investors have to rack up decades of returns before academics even have any data to work with!), but academics surely have done a hell of a job in explaining and systematizing many factors. I'd even call it a somewhat embarrassingly few number of factors (value investors losing their shit over this), and I even amusingly noted, from the book, that the alpha pool is even shrinking, just because academics come up with new sources of alternative beta every now and then, to discretionary managers' demise.
Antti Ilmanen lays it all out - markets might not be fully efficient, but they sure are fucking efficient. What's the source of excess returns? Don't try to source returns by outsmarting someone else (clean alpha), but instead get it from harvesting market rewards, by bearing certain micro-founded, empirically documented, peer-reviewed, anecdotally economically intuitive risk premia.
Out of the massive number of risk premia out there in the market (quants working hard in investment banks and hedge funds tryna p-hack more and more premia every day), Antti nicely boils it down to 3 sources - asset, style, factor. Couldn't have been a better way to visualize it.
In my view, quant can be way better than a black-box, insulated, esoteric, elitist field only reserved for the mathematically gifted - Antti has laid it out that quant can actually be a hella of an intuitive and systematic way to visualize return sources of all investors. Sure, many things are inconclusive in the literature - the rational vs irrational camps, systematic vs discretionary, and I'm personally agnostic. But it got me thinking - holding all factors and beta risk constant, did the legends (Buffett, Klarman, Schloss etc) even generate clean Jensen's alpha? Recall that Jensen's alpha is the intercept of the Security Characteristic Line.
My theory - top down allocation and bottom up analysis are 2 sides of the same coin. Instead of allocating to assets or managers, allocate to beta factors that are determined ex ante, to outperform. We have to be as prescient and intelligent as Ben Graham in the 40s, who somehow determined ex ante that value would outperform, and thus being 50 years ahead of his time in the process. Is the next breakthrough beyond Modern Portfolio Theory - portfolio beta management instead of old-school asset management? Antti has even hinted to this in the last chapter, which I arrived at independently.
However, as promising as bearing risk premia may sound, smart beta performance has been mixed so far. Passive indexing can take guts as well. That being said, the literature says that smart beta lacks sufficient conviction and exposes the investor to many unintended bets? The equity premium also still remains a puzzle.
Such is life