If you want to make something wildly popular in Britain, hide it in a financial product and give it a reassuringly competent name. Strategic bond funds, absolute return vehicles, private credit strategies: the City is full of financial products that sound like sedatives and often behave like them too. Against this backdrop, gold is an oddity. Concrete, primal, and a bit basic, it relies on raw scarcity and fear of systemic breakdown to make its case. No wonder it tends to be discussed by investment professionals as a kind of lurid, speculative vice – the financial equivalent of raspberry-flavoured vapes, or dogging.
Yet gold was up over 50 per cent last year, completing its stately ascent without so much as a 10 per cent dip. It outperformed every major global equity market, save Spain and Korea. It crushed most baskets of AI-related stocks. That would typically bespeak the price action of a mania. Almost everyone thinks gold is in one. It has featured among the top three “most crowded trades” on Bank of America’s Fund Manager Survey for most of the last 12 months.
Yet gold may be a case of “nobody goes there – it’s too crowded.” In other words, everyone is bullish but no one really owns enough of it.
Consider central banks, which have been net buyers of gold every year since 2008 as a hedge against growing monetary and geopolitical instability, and which, since the Ukraine conflict in 2022, have markedly stepped up the pace. They now hoover up about a fifth of global gold demand vs half that in the preceding decade. Yet even for these glint-eyed institutions, gold is still under 30 per cent of official reserve assets, as against the 60 per cent level they were at in the early 1980s. Are they really at the feast or just politely nibbling? Put it another way: if you, a based central bank, are trying to hedge against the decline of the unipolar hegemon and the possibility of structurally higher inflation, why are your holdings of gold not much higher than they were during the 1990s Pax Americana, which was also a period of falling inflation? Oops. No surprise that a clear majority of the 73 central banks surveyed recently expect their holdings to rise over the next five years.
New year, new read. Save 40% off an annual subscription this January.
Other institutional investors are, if anything, even less at the party.
How do we know? Industry body World Gold Council and Coalition Greenwich did the obvious, slightly boring thing: they went and asked over 400 of the world’s pension funds, insurers and endowments what they actually hold. The results, in a 2022 study called “The Use of Gold in Institutional Portfolios”, show that only about 15 per cent of respondents had a specific allocation to gold at all; among that minority, the average allocation was roughly 4 per cent. Do the unforgiving arithmetic: if 15 per cent hold 4 per cent and 85 per cent hold nothing, the true average across will be well under 1 per cent. Gold has doubled since 2002, so this may now be 2 per cent. So much for an allegedly “crowded” trade.
Even family offices, which should have the capacity to dial up risk and frequently do, have only just begun loosening their cufflinks. According to UBS’s latest Global Family Office Report, the average allocation to gold has doubled…from 1 per cent. That is the allocator’s equivalent of moving the dial from “a touch won’t hurt” to “oh, go on then.” Europe and Asia are consistently more likely to be long than the US, where gold is thought to be a particularly unserious addition to portfolios. The average US family office allocation is closer to 0 per cent than to 1 per cent.
If you lined up the world’s great pools of capital and asked them to jingle what they have in their pockets, you’d hear mostly equities, bonds, private this and private that. Gold would be the lonely 1 pence coin at the back. Sovereign wealth funds? Under 1 per cent in commodities. US public pension plans? 2.9 per cent in commodities. The gnomic Swiss? 1 per cent in commodities, mostly gold. One per cent, two per cent, one per cent, two per cent. For institutional investors, that is really the new forty-forty-five, forty-forty-five.
What about retail investors? Total known ETF holdings of gold have indeed been rising all year, though are still below where they were in 2022.
In other words, when we say “everyone is long gold,” what we mean, in practice, is: a handful of central banks, a few excitable retail investors, and your very online friend who bought a Krugerrand after watching a documentary about Weimar.
Perhaps the institutions are right, and maybe this is just a blip. After all, the gold price did nothing between 2011 and 2021. Perhaps gold, which doesn’t have any cash flows and has a long term real return of 2 per cent, should indeed be a small eccentricity in the corner of a portfolio.
Yet that is not what academics conclude when trying to determine the optimal amount of gold in a portfolio. The exact “optimal” number depends on the period of historical data used, but typically the range is strikingly above current allocations. An allocation of roughly five and ten per cent in gold would have improved risk-adjusted returns – what professionals call the Sharpe ratio – over long periods compared with owning no gold. In other words, gold tends to produce “smaller drawdowns” or smoother journeys than the classic 60/40 mix of shares and bonds. One analysis even found that fully swapping bonds for gold in a 60-40 portfolio delivered similar returns but a slightly lower Sharpe ratio over 1973-2024.
This leaves a rather delicious gap. On one side: actual institutional allocations, often zero and rarely above one or two per cent, with a few honourable exceptions. On the other: the industry’s own models, blinking away on researcher’s laptops, serenely suggesting that something in the low single digits – sometimes more – would have made the last few decades noticeably less hair-raising.
Why the gap? Path dependence is the most likely reason. We have been in a benign period for growth and inflation and geopolitical risk. The 60/40 plus alternatives framework is now so entrenched as a starting point it might as well be liturgy. Gold never quite made it into the catechism. It was historically filed under “odd things you buy when markets crash”, not as a core building block.
Principal-agent problems are another. As an allocator, once you have painstakingly assembled your zoo of private equity, private credit, hedge funds and factor-tilted ESG smart-beta flapdoodle, it is administratively and politically tricky to carve out space for something as embarrassingly simple as a single metal. In fact, try this at your next trustee meeting: say, “we’re putting ten per cent into infrastructure,” and everyone nods and thinks of charismatic, photogenic mega offshore wind farms. Say, “We’re putting ten per cent into gold,” and half the room imagines you broadcasting from an undisclosed bunker. No one wants to be a “gold bug” – not even gold bugs.
Finally, time horizons. The virtues of gold – insulation in crises, smaller peak-to-trough losses, the ability to preserve purchasing power over long spans – reveal themselves over decades. The pain of holding an unfashionable asset reveals itself at every single quarterly performance review. If you are a CIO with a three-year contract, you are not, in practice, optimising over fifty years of data.
It wasn’t always this way. Maynard Keynes, in managing the King’s endowment, often had a third of his UK stock portfolio in mining stocks, dozens of points in excess of the benchmark. That was a lopsided world, but so is ours, and central banks have been the first to realise it. While the institutions that issue our money increasingly treat gold as a serious hedge, the institutions that look after our savings treat it as an optional eccentricity. If they decided to increase their holdings, just remember: all the gold in the world would fit into four Olympic swimming pools.
[Further reading: God loves Lily Phillips]
Content from our partners