Marketing communication
This page contains marketing information
Ultra processed portfolios. Is the diversification 'free lunch' still on the menu?
10 min read 26 Oct 26
The economist Harry Markowitz, who won a Nobel Prize for the mathematics of building portfolios, called diversification “the only free lunch in investing”. It is probably the most quoted line in portfolio construction. The principles behind it seem almost like common sense: don’t put all your eggs in one basket. Combine assets that do not move for the same reasons, and the whole carries less risk than the sum of its parts. No forecasting required. It is just arithmetic.
But it is worth being honest about the analogy, because the industry has leaned on it rather heavily. Just because a lunch is free does not mean it tastes good, and it certainly does not mean it is good for you.
Markowitz was making a claim about covariance – the statistical behaviour of investments held together. He was not making any claim about the quality of the ingredients. Somewhere along the way, ‘low cost, broad and technically uncorrelated on paper’ became accepted shorthand for ‘healthy’. Those are not the same thing, and the difference between them is now a source of a significant amount of portfolio risk.
Ultra-processed portfolios – the illusion of variety
An interesting and potentially powerful parallel comes from a current debate happening outside financial markets. I have been reading Ultra-Processed People, a book by doctor and broadcaster Dr. Chris van Tulleken,1 which argues that a large share of what fills our supermarkets is not really food, it is “an industrially produced edible substance”. It is engineered for shelf life, low cost and repeat consumption, and strongly implicated in obesity, chronic illness and environmental damage.
What is interesting is the difference between apparent variety and actual nutrition. A supermarket aisle offers dozens of brands, flavours and formats, which can all be reduced to the same handful of cheap engineered inputs: refined sugars, refined fats, salt and starch.
The variety is real at the level of the label, but largely illusory at the level of what anyone actually gets when they eat. And because these products are designed to be convenient, often low cost and moreish, they encourage you to keep consuming.
There is a parallel with modern portfolio construction. A top-down asset allocation populated with low-cost passive building blocks delivers genuine variety at the level of the label: thousands of holdings, dozens of countries, and multiple ‘asset classes’, each with a distinct and reassuring name.
For many years, this looked like an excellent investment solution, and in fairness it was a considerable improvement on the huge variety of less structured approaches that came before it.
But count the underlying portfolio nutrients rather than the labels, and the picture changes. A striking proportion of the return and risk in many modern portfolios actually reduces to a small number of common drivers – US mega-cap technology and the artificial-intelligence (AI) theme, duration, and the dollar. Many different names; the same few things actually creating the outcome.
If Dr. van Tulleken’s formulation is right about food, something similar can be said about the building blocks of the ultra-processed portfolio. A vehicle tracking a capitalisation-weighted index is not, by design or construction, a true economic exposure or a genuinely diversifying asset. It is an ‘industrially produced investible substance’ that is manufactured to be cheap, liquid, administrable and easy to allocate to – engineered for convenience and to support the system that distributes it rather than for the nourishment of the portfolio that consumes it.
That is not an argument against ever using one; it is simply an argument for knowing what you are eating. I like a can of soda, crisps, ice cream and chocolate as much as anyone, but I know that is not a diet that is going to help my body become strong or resilient.
When the map started running the territory
How did this happen? Largely through the quiet conversion of a filing system into a philosophy.
The taxonomy of asset classes we all use – equities and bonds, developed and emerging markets, investment grade and high yield – grew out of an entirely reasonable need. Global markets are complex to the point of illegibility, and institutions have to govern their participation in them: to report, to benchmark, to set risk budgets, to hold managers to account and to explain to a board what has been done with the money. Grouping exposures under names made that possible. It was, and remains, a sensible administrative solution to a genuine problem.
The difficulty is that a classification system designed for governance has gradually been treated as a description of economic reality. Investors no longer merely use the taxonomy to describe what they own; increasingly they allocate to the taxonomy, letting the categories drive the decisions. The map, having been drawn for convenience, has started running the territory.
And critically, the map has not kept pace with the ground it claims to represent. ‘US equity’ once described a broad exposure to American companies and the US economy; today its contours are established by a handful of companies pursuing a single technological theme with global, not domestic, revenue bases. The US label essentially describes where an asset sits on a map, not what actually drives its return or its risk.
When broad becomes narrow
There is a reinforcing mechanism at work here: a capitalisation-weighted index allocates the most capital to whatever has already risen most, regardless of whether that rise reflects improving fundamentals or simply narrative, capital flow and momentum.
Index weight attracts flow, flow supports price, price increases index weight. Within that loop, buying ‘the market’ is not a neutral, diversified stance at all: it is a position that systematically rewards price momentum rather than the underlying economic drivers of growth at a company or market level, and it concentrates further precisely when concentration is already rising.
The numbers make the point more powerfully. It is now a well-known fact that something like half the entire value of the S&P 500 sits in around twenty companies.2 Widen out to the global index and roughly ten names – every one US-listed, eight of them in the same artificial-intelligence and technology cluster – make up close to a quarter of the whole world’s stock market, with about two-thirds of the index tied to a single country.3
"Half the entire value of the S&P now sits in around twenty companies"
It is fair to say that these weights are not arbitrary. Market capitalisations reflect the collective judgement of millions of investors about what each company is worth, and that judgement may well be right. But it is a judgement about the value of each separately.
No one is actively voting on whether a quarter of the world's equity market should depend on a single technological theme. That outcome is the sum of thousands of separate pricing decisions, not a decision about diversification, and correct prices can still add up to a concentrated portfolio.
None of this is an argument against index investing, which has given savers low-cost access to markets and remains a perfectly sensible building block when used deliberately. It is an argument against mistaking an administrative convenience for an assurance of diversification.
Rethinking the equity/bond partnership
The same question applies to the relationship between equities and bonds – what is it really driven by? For most of the past forty years, the relationship has created a well used construct for diversification. The two broad ‘asset classes’ complemented each other well: when share prices fell in a growth scare, government bond prices usually rose, because the market expected central banks to cut interest rates to cushion the blow.
Believing this relationship would hold particularly in tough times, people bought bonds if they were expecting a crisis or just perceiving a crisis, making the relationship expectation almost self fulfilling.
But the pairing depends on a specific condition: the central bank having room to cut. An inflation shock removes it – with prices already rising too fast, policymakers cannot come to the rescue. In 2022 we saw the result, as equities and bonds fell together.
That was not a freak occurrence. For much of financial history, the two have moved together; the reliably offsetting relationship of recent decades was itself the exception. The lesson is not that bonds have stopped being useful, but that the relationship was never a property of the labels – it was a property of the policy environment, which we can argue has changed.
Nutrition, not calories: what genuinely active investing means
Put all this together and the conclusion is encouraging rather than gloomy. Diversification has not stopped working. But it can no longer be inherited from a set of labels – it has to be built actively and deliberately, from the bottom up, by understanding what drives the fortunes and ultimately the price of each holding.
There is an important distinction to be made. A great deal of what is sold as active investment today is index-relative decision making. It starts from benchmark index weights – most commonly cap-weighted indices – and then takes modest positions around them.
While this type of active investing is fundamental to ensuring efficient indices, the approach cannot solve the problem of diversification described here – because it inherits the very concentration it claims to manage. If the benchmark is a levered bet on a handful of names, then a portfolio built as a small deviation from that benchmark is also a bet on those names. To borrow the analogy once more, reformulating an ultra-processed product does not make it real food.
Genuinely active investing starts from a different place: not with the index weights, but with an understanding of each asset you intend to invest in. It asks what a company actually earns and why, what a borrower actually depends on and whether they can repay, and what a private asset’s cash flows are genuinely tied to. It builds the portfolio where each holding earns its place and builds aggregate exposures on purpose rather than starting with a price-related index.
That is hard work. It requires genuine research depth across equities, credit and private assets, and requires building sufficient conviction to hold positions that you believe will deliver superior returns rather than match some kind of index. It is the work that determines whether a portfolio is ‘nutritionally’ diverse rather than just varied by names and labels.
Building a nutritious portfolio
Looking through the equity index
If passive ownership of global stock markets has quietly become a bet on a few large names, genuine equity diversification has to be constructed on purpose. That means selecting companies on the basis of what they actually do and how good they are at generating earnings. Once you know this, you can start to have a genuine view about how they may behave in different conditions.
Encouragingly, the raw material for this within equity markets is abundant. In the S&P 500 Index, the so-called Mag 7 grab the attention and their market capitalisation continues to dominate. However, if we exclude these seven stocks, the dispersion between the best and worst performers has been increasing and is historically wide (estimated by deciles). Over the last decade, the top decile ex-Mag 7 has outperformed the Mag 7 in the majority of years. The companies that make up this decile in each year have changed often but this further demonstrates the opportunity for looking beyond the index.
Careful bond selection
Bonds are typically considered the ballast of a portfolio, and after years of uncertainty about whether they could still perform that role, they merit a fresh look: real, inflation-adjusted yields are positive for the first time in years.
But a bond index carries its own version of the same flaw, and arguably a worse one. The largest weightings go to whoever has borrowed the most, which common sense suggests is the opposite of how to actually diversify bond risk.
Active bond selection means identifying where the most attractive yield genuinely resides, and judging credit quality by company fundamentals and by their position in their market – and this is increasingly the same for a government or a corporation. The label ‘credit’ says nothing on its own about whether that judgement has been made.
As with equity, the story of an index or a rating bucket doesn’t characterise the full extent of the opportunity or the risk.
"Sourcing and originating specific assets is what produces genuinely differentiated exposure"
Broadening horizons to private markets
Some of the genuinely distinctive exposures available to investors sit outside public markets. Private credit, infrastructure and real assets are driven by things the stock market often is not: the rent on a specific building, the contracted cash flows of a toll road or power grid, the terms of an individual loan. Infrastructure frequently carries built-in inflation protection; in private credit, the close relationship between lender and borrower means any trouble tends to surface early.
This is also where the distinction between owning an exposure and originating one matters most. Buying broad, pooled access to private markets risks reproducing exactly the problem this article describes – a reassuring label covering a set of drivers you may not have actually examined. Sourcing and originating specific assets is what produces genuinely differentiated exposure, and it is a capability rather than an allocation.
A healthy lunch, not just a free one
Markowitz was right about the principle, and the mathematics have not changed. What has changed is that the industry built a system of names on top of it, and then began to mistake the names for the thing itself. A taxonomy created to make global investing governable has become the way investing is done. That taxonomy is also intrinsically self-reinforcing. Growth accrues to things that have grown the most. This investing behaviour has been quietly concentrating portfolios at the time when investors may need real diversification.
The remedy is not to abandon the idea of diversification, but to take it more seriously than just holding a portfolio of things with different names. It means looking through the names to the economic drivers underneath, building portfolios with specific exposures that each contribute to the desired outcome and bring something different.
This is harder than buying an index. But this kind of ‘free lunch’ is still available. As with food, the question is no longer whether the meal is free or even whether it offers plenty of choice. It is whether what is on the plate will actually nourish the portfolio – or make me feel really good while eating it, but sick and probably guilty when I’ve finished. That’s something to think about when you plan your diversification menu.
1 Chris van Tulleken, Ultra-Processed People: Why Do We All Eat Stuff That Isn’t Food… and Why Can’t We Stop? (Cornerstone Press, 2023).
2 BlackRock, ‘Mega-Cap Exposure: Manage Concentration Risk in the U.S. Market’, April 2026.
3 MSCI World Index, August 2026.
The views expressed on this webpage should not be taken as a recommendation, advice or forecast, nor a recommendation to purchase or sell any specific security.
The value of investments will fluctuate, which will cause prices to fall as well as rise and you may not get back the original amount you invested. Past performance is not a guide to future performance.
