Vanguard just put out another paper on why diversification wins.
The numbers are hard to argue with. Looking at the Russell 3000 from 1990 to 2025, roughly 75% of stocks underperformed the index. More than half failed to beat even the risk-free rate. At the same time, a very small group of companies created the majority of the wealth. Just ten (10) stocks accounted for more than 30% of total wealth creation over that period. Alphabet, Microsoft, Apple, Eli Lilly, NVIDIA and a few others did an enormous amount of the work.
The paper’s conclusion follows the usual logic. Because returns are so heavily skewed, the biggest risk is missing the small number of winners. Broad diversification reduces the chance of that happening. Indexing, in their view, is the rational response for most investors.
I don’t disagree with the data. Most stocks are average or worse over long periods. A tiny fraction of companies generate almost all the long-term value. That part is clear.
Where I separate from the standard conclusion is what you should actually do with that information.
If a small number of companies drive most of the returns, then the entire exercise of investing changes. The hard part becomes identifying the companies that have a realistic chance of becoming one of those long-term compounders, and then owning them with enough size that it actually matters when they work.
Charlie Munger spent decades making this point. He never believed you needed a long list of stocks. He believed you needed a few outstanding businesses that you understood deeply and could just hold. The idea of owning a little bit of everything was, in his view, what people do when they don’t know how to distinguish the great from the average.
Stanley Druckenmiller has been more aggressive about it. He has said for years that the real money is made in concentrated positions when the conviction is high. Diversification is what you do when you lack edge. When you have edge, sizing up is the logical next step.
Bill Ackman runs a concentrated portfolio for the same reason. Seth Klarman has written extensively about the difference between genuine risk management and simply spreading capital across dozens of mediocre ideas so that nothing can hurt too much — or help too much either.
The Vanguard research is correct about the base rates. The odds of randomly picking the winners are low. Most people who try to select individual stocks will underperform because they will miss the small group that matters or they will sell them too early. That is a real and well-documented problem.
It does not follow that the correct response is to abandon the attempt entirely and index by default.
It follows that you have to be far more selective than most people are willing to be. You have to accept that you will be wrong some of the time. You have to be willing to look different from the market for long stretches. And you have to size the positions large enough that the winners can actually move the portfolio.
Concentration Risk
Concentration is not free. When a large part of your capital is tied up in a small number of names, any single mistake or external shock hits harder. A regulatory change, a failed clinical trial, a shift in technology, or simply a long period of underperformance can damage the portfolio in ways that a broad index would absorb more easily.
You give up the comfort of averaging and you accept that your results will be more volatile and more dependent on a few decisions. There will be periods where the portfolio looks stupid relative to the market.
The people who succeed with concentration are usually the ones who size positions according to conviction and who can live with the discomfort when the market disagrees with them for an extended period. Most investors are not built for that. For them, the Vanguard conclusion is the right one.
For those who are willing to accept the trade-off, the same data that shows how skewed returns are also shows why concentration can be powerful. A small number of companies create most of the value. The question is whether you want a tiny piece of all of them, or meaningful ownership in the few that actually matter.
Practical Takeaway
If you are going to concentrate, do it deliberately.
Limit the number of positions so that each one can actually matter. Force yourself to write down why each company has a realistic path to becoming category-defining over a long period. Be honest about the ways the thesis can break. And size the position large enough that success would move the overall portfolio, but not so large that a complete failure wipes you out.
The distribution of stock returns is extreme. That fact can be used to argue for owning everything. It can also be used to argue for owning only the few companies that have a chance of ending up in the extreme right tail. Both conclusions are logical. The one you choose should match how you actually invest and how much discomfort you can handle.
If you want to see exactly how I apply this in practice, I publish my full portfolio for paid members. Not a model or a list of ideas I’m watching. The actual positions I own with my own capital, sized the way I size them, with the same concentration and the same risks I take.
You can see the current holdings here: