Portfolios are usually spread along the dimensions that are easy to label rather than the ones that determine whether two holdings fall together. This lesson is about telling the two apart.
11 min read · Level 2
Every one of these dimensions is easy to report and hard to actually achieve, which is why portfolio summaries are so often more diversified than the portfolios they describe. The one that deserves more attention than it gets is vintage, and the reason it is neglected is psychological rather than analytical: spreading entry across years means sitting on capital during periods when everything looks attractive, which feels like failure in real time and is invisible as a benefit afterwards. The contrarian claim worth taking seriously is that when you committed may matter more than what you chose. Nobody can prove that in advance for any individual portfolio, and it is a reason to treat the pace of deployment as a genuine decision rather than as whatever happens when opportunities arrive faster than judgement.
Beyond how much of each thing you own, a portfolio is spread along dimensions. Four are usually named: stage - how mature the underlying thing is; sector - what business it is in; geography - which economy and currency it lives in; and vintage - when you bought it.
The fourth is the one most often left out and is frequently the most powerful, because it is the only dimension over which you have complete control. You cannot make a sector behave independently of another sector, but you can absolutely decide not to commit all of your capital in a single eighteen-month window.
The purpose of examining all four is not to fill every box. It is to find out where your apparent spread is real and where it is only a labelling exercise.
Because labels are the cheapest possible source of false comfort. A portfolio described as 'across six sectors and three continents' can still be one bet if every position depends on the same customer budget, the same financing market or the same currency - and the description will keep sounding diversified right up until the moment all six sectors move together.
And because in private markets in particular, stage diversification is weaker than it appears. Earlier positions typically depend on being able to raise a further round; later positions typically depend on an exit market. Both of those are the availability and price of capital, which is one underlying force wearing two different labels, and it moves for the whole market at once.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
Three pieces of arithmetic, each showing a spread that is not what it looks like.
Sector. A portfolio of twenty positions with fourteen in one sector has 70% of its positions riding a single set of conditions. The remaining six are spread across everything else. Counted as line items it looks broad; counted by what would make things fall together it is close to a two-position portfolio, one of which is very large.
Geography. A holding abroad returns 20% in its own currency over a year, while that currency falls 15% against the one you spend. Your return is 1.20 times 0.85, which is 1.02 - a 2% gain. The asset did what you hoped and eighteen percentage points of the result were decided somewhere else entirely. Geographic spread is therefore two decisions bundled together, and the currency half is often taken by accident.
Vintage. Suppose entry prices in a market are 40% higher in one year than another. The same business, bought at 1.4 times the price, returns 1 divided by 1.4 of the multiple - about 71.4% of it. An outcome that would have been a 5 times return at the lower price is 5 divided by 1.4, or 3.57 times, at the higher one; a 3 times return becomes 2.14 times. Nothing about the business changed. Committing an entire allocation inside one window makes the entry price of that window a permanent feature of every position you hold.
The commonest mistake is counting categories rather than drivers. Six sector labels that all depend on discretionary consumer spending are one driver, and the six labels actively obscure that.
The second is treating stage as a diversifier within one asset class when it is mostly a maturity gradient along a single funding cycle. Holding both early and late positions changes when you find out and how much you make; it does not change what you are exposed to nearly as much as the two labels imply.
The third is unintentional currency exposure. Foreign holdings are frequently chosen for the asset and inherited for the currency, and the currency half is never examined until a year when it dominates the result.
The fourth is ignoring vintage entirely, which is the easiest of the four to spread and the one that receives the least attention, because spreading it requires deliberately not acting while you still have money and enthusiasm.
Not 'software' but 'corporate IT budgets'. Not 'property' but 'the cost of borrowing'. Two positions with different labels and the same sentence beside them are one exposure, and this is the exercise that reveals it.
Add up the weight behind each distinct driver rather than each distinct holding. The largest number is what your portfolio is actually a bet on, whatever the category breakdown says.
For every foreign holding, note the currency and what share of the portfolio sits in it. Then ask whether you would have chosen that currency exposure on its own. Very often the answer is that nobody chose it.
Ask what each private position needs in order to succeed - another round, or a buyer. Then notice that both answers depend on the same capital conditions, and treat the resulting spread as a timing difference rather than as genuine independence.
Note, for each position, roughly when it was bought and at what level of prices. A portfolio whose positions all share one window carries that window's pricing permanently, and knowing that is the difference between a known concentration and a hidden one.
When reviewing whether a portfolio's apparent breadth is real, when considering a new position that resembles several existing ones, and when a large allocation is about to be committed inside a short period.
This is not an argument for holding a position in every sector, stage and region. A portfolio assembled to fill boxes tends to contain holdings nobody understands, and an exposure you cannot follow is a worse problem than an exposure you have chosen knowingly.
A frequent finding when a portfolio's holdings are examined by what they depend on rather than by how they are categorised is that several apparently distinct sectors resolve to the same source of demand. Businesses labelled software, marketing services, recruitment and professional services can all be funded from the discretionary budgets of the same kind of corporate buyer, and those budgets are cut at the same time for the same reason. The sector breakdown in the portfolio summary remains entirely accurate throughout, which is what makes it dangerous rather than merely useless.
A category system built for describing businesses is not a category system built for predicting what falls together, and only one of those is a risk tool.
Of the four dimensions, vintage is the only one that nobody has an incentive to promote. Sector and geographic spread can be sold as a product; stage spread is a strategy anyone can describe; but spreading capital across time consists mainly of declining to invest while you are enthusiastic and money is available, which no intermediary earns anything from and no investor enjoys. The result is that it is widely acknowledged in principle and rarely built into a plan.
Where a discipline has no natural advocate, it has to be written into the plan by the person it protects, or it does not happen at all.
Create a free account to read the rest and apply it to your own investing.