In the portfolios we review, a common pattern emerges. The problem is rarely that someone owned nothing interesting. It is that the whole portfolio was interesting — fifteen positions, each bought for a reason that made sense at the time, no single one large enough to matter, and collectively expensive, overlapping, and impossible to describe in a sentence.
The structure we prefer is deliberately less exciting. A large, cheap, broadly diversified core. A small number of satellite positions around it, each with a specific thesis and a specific size limit. Two different jobs, kept separate on purpose.
What the core is for
The core is designed to capture broad market returns at a low cost, with minimal maintenance. Broad equity exposure across US, developed international, and emerging markets. High-quality fixed income sized to your actual tolerance for volatility rather than a rule of thumb about your age.
Built with index funds and ETFs, a core like this can cost a few basis points a year. That matters more than it sounds. Cost is one of the few components of future return you know in advance — most everything else is an estimate. In our view, reducing the cost of the largest part of the portfolio is one of the more dependable improvements available to an investor.
The core should also be boring enough that you never feel the urge to tinker with it. If you find yourself checking one of your core holdings weekly, it may not really be a core holding.
If everything in a portfolio is working at the same time, it may be less diversified than it looks.
What satellites are for
Satellites are where a genuine view gets expressed. A sector you believe is structurally undervalued. A commodity or real asset exposure that hedges a risk specific to your situation. An individual company you understand well and want to own directly.
These positions can play a meaningful role, and there is nothing unsophisticated about holding them. They also carry more risk than the core, and they may underperform. But each one has to answer three questions before it earns a place:
- What is the thesis? Stated in a sentence or two, in advance, in writing. "Semiconductors are interesting right now" is not a thesis.
- How large can it get? A ceiling set before you buy, not after it has run. This is the difference between a satellite and an accident.
- What would prove it wrong? The condition under which you would exit. Without one, a losing position quietly becomes a permanent one.
A portfolio with four well-reasoned satellites at three to five percent each is a portfolio with a point of view. One with fourteen satellites at one percent each is a portfolio with a collection.
The expensive middle ground
The structure we try hardest to avoid is neither of these. It is a portfolio full of moderately active, moderately expensive funds that collectively may look a great deal like the index while charging considerably more. In that case, you can take on the cost of active management and receive something close to market returns before fees — and less after them.
This tends to happen gradually, often without anyone deciding it should. A fund gets added here, a recommendation acted on there, and nobody ever steps back to look at the whole thing at once. In our experience, it is one of the more common findings when we review a new client's existing accounts.
How to audit your own portfolio
You can do most of this yourself in an afternoon:
- Add up the cost. Weight each holding's expense ratio by its allocation. Anything meaningfully above roughly 0.30 percent for the whole portfolio deserves an explanation.
- Check for overlap. Look at the top ten holdings of each fund. If four funds all lead with the same handful of large US companies, you own one bet in four wrappers.
- Count what you cannot explain. Go position by position and try to say why you own it. Anything you cannot explain is either a satellite without a thesis or a leftover.
- Look at the losers. In a diversified portfolio, something is usually lagging. If nothing ever is, it is worth checking whether you are concentrated in a single factor or region.
The takeaway
Keep the core cheap, broad, and dull. Let satellites carry conviction, but make each one justify itself with a thesis, a size limit, and an exit condition. Be wary of the expensive middle.
The point of this structure is not only cost and discipline. It is that a portfolio built this way can be explained in a few minutes — which makes it easier to understand, and in our experience understanding helps people hold on through difficult years.