Do commodities actually hedge inflation?

This is the most repeated claim in the asset class, and the honest answer is mixed, conditional, and different for each of the three things people lump together. Selling the tidy version of this story would be the easiest thing in this course to do and the least useful. So here is what the evidence actually supports.

Split the question into three

1. Energy — the strongest and least surprising link. Energy prices are inside the inflation basket. Energy is directly around 6-7% of the US consumer price index, and it feeds indirectly into transport, food and manufactured goods. A commodity that is a component of the index it is hedging will naturally co-move with it. This is close to a mechanical relationship rather than a market insight.

2. Broad commodity baskets — a positive but noisy relationship, mostly with the surprise. Research on long samples generally finds broad commodity exposure has been positively related to unexpected inflation — the part that neither bonds nor equities handle well, and the part that actually damages a conventional portfolio. But the relationship is statistically noisy, it is dominated by the energy weight in most indices, and there have been long stretches (much of the 2010s) when inflation was positive and broad commodities delivered poor real returns anyway.

3. Gold — the most popular hedge with the weakest medium-horizon evidence. It gets its own lesson next, because it barely behaves like a commodity at all.

The two distinctions that resolve most of the argument

• Level versus surprise. Commodities have historically responded to inflation shocks, not to the steady existence of inflation. In a year when inflation runs at an expected 2%, there is no reason for a commodity to do anything. This is why "commodities hedge inflation" and "commodities did badly during a decade of positive inflation" are both true statements.
• Horizon. Over one to two years, energy-heavy baskets have tracked inflation surprises reasonably. Over ten years, the commodity's own capex cycle overwhelms the price level entirely — you get the commodity cycle, not the inflation rate.

There is a third, practical distinction: the spot price is not the return. Commodity exposure is usually obtained through futures, whose return includes the effect of rolling contracts through the curve and the yield on collateral — which can differ substantially from the headline price change. That mechanism is the subject of this domain's curve course; note here only that quoting spot performance as if it were an investor's return overstates the case in some periods and understates it in others.

A worked example

Rounded, illustrative numbers for two years of a portfolio.

• Year A — inflation surprises 3 percentage points to the upside. A broad commodity basket returns +25%; equities return −15%; bonds return −10%. Commodities did the job precisely when the other two failed.
• Year B — inflation lands at 2%, exactly as expected. The same basket returns −20% on an oversupply cycle, while equities return +10%.

Average the two and the basket has returned roughly nothing while swinging violently. That is the honest shape of the evidence: not insurance, but a volatile exposure whose good years have tended to cluster in inflation surprises. Whether that trade-off suits any particular portfolio is not a question this academy answers — it is a question about that portfolio's objectives, and it belongs with a qualified adviser, not with a curriculum.

In the data

Both halves of this comparison exist, and neither is quite what people assume. The newest values of each are below: US consumer-price inflation, one figure a year, and the global all-commodities price index, monthly, set to 100 in 2016.

The index is a price index: it contains no roll and no collateral yield, and is therefore nobody's realised return. And the inflation series is the realised level, once a year (8.0% for 2022). The surprise this lesson turns on is not in it at all.

Try it now

1. Twenty years of both legs are below, lined up by us on 28 September 2026: US consumer-price inflation, and the year-on-year change in the annual all-commodities index. Remember that the second is a price index carrying no roll and no collateral yield.

2. Mark the periods where inflation surprised upward, and check the commodity's behaviour in those specific windows against its behaviour in the calm ones. The table holds no forecasts, so use the nearest honest proxy it does hold: the years where inflation jumped well above the year before.
3. Find one decade where inflation was positive and commodities were flat or falling. Being able to point at that stretch is what separates understanding the evidence from repeating the slogan.