Swensen argues that there are basically three sources of returns: asset allocation (which market(s) do you choose? Bonds, stocks, real estate?), market timing (when to sell and when to buy) and security selection (after you chose your market(s), which stocks, bonds, etc. do you pick?). The book is structured by this argument.

I skipped large portions of the book as I realized that I want to look at different markets as Swensen. I am more interested in the token economy than the "core asset classes" he suggests, also his critique on mutual funds was not interesting to me because I do not plan to invest there.

However, I liked the way Swensen substantiates his claims. His book has some careful selected and compiled tables that actually show that his points are not some sort of gut feeling. That is something I did not see that often in Personal Finance literature.

What I remember from this book is: Market Timing is hard. Portfolio Rebalancing (for example, pick ten stocks with the highest estimated Sharpe ratio on regular intervals) may work, performance chasing most likely not. At least my investment strategy should benchmark against a strategy that does very little market timing.