A PhD in finance corners you in a meeting and asks what I think of the efficient frontier and mean-variance optimization. Fair question. My honest answer is that the theory has earned its keep. It gave the industry a common language and a useful starting point. The problem starts when people begin treating the model as reality because that is usually when you have already lost the plot.
Correlations move. Relationships between asset classes that held for a decade can break in a quarter. The past was never really a preview of the future; it just has a funny way of looking like one after the fact.
I did not argue the point. I showed it instead that was a set of global asset-class charts that took real effort to put together, the kind that are hard to fake and even harder to argue with. A few good charts can sometimes save you from having to give a 20-minute lecture on mean-variance optimization which is probably good for everyone in the room.
Here is what I probably do not say out loud often enough. Most of what I sat through in a US business college more than 30 years ago covering monetary policy models, interest-rate theory, inflation frameworks and the rest has aged rather badly. Some of it is simply wrong. It sounds airtight in a textbook but markets have never shown much respect for textbooks once they decide to do something else.
The same meeting moved on to another thread. Without alternatives in the mix, trying to earn consistent absolute returns year after year not just in the good years is getting close to fantasy in this environment.
Diversification looks very convincing in a spreadsheet. It becomes considerably more interesting when everything you own suddenly decides to behave like the same asset class.
Later, a financial planner asked me what I thought was the one thing that mattered most. I told him it is not a formula, a factor model or another clever framework. It is accepting that this business rewards hard work above almost everything else in bull markets, in bear markets and, perhaps most importantly, in the in-between markets that nobody writes textbooks about.