I keep coming back to the same idea when someone tells me they want exposure to oil but they hate the way futures behave. You do not always have to touch the barrel. A lot of the commodity risk you care about is already priced into a currency somewhere, and that currency often trades cleaner, deeper, and with tighter spreads than the front-month contract you were about to buy. The Australian dollar, the Canadian dollar, and the Norwegian krone are the three I watch most, because each one is close to a pure play on a specific export, and the link is mechanical enough that you can actually reason about it.
Why terms of trade drive these currencies
The connective tissue here is terms of trade, which is just the price of what a country sells divided by the price of what it buys. When Australia's iron ore and coal sell for more, Australia earns more foreign currency per shipment, and all else equal that pushes the Aussie up. Same logic for Canada and oil, and for Norway and its oil and gas. These are economies where a single category of export is large enough to move the national accounts, so the currency ends up carrying the commodity's signal whether the central bank likes it or not.
A few things follow from that. First, the relationship is a proxy, not an identity. The Aussie also responds to Chinese demand, to rate differentials, and to broad risk appetite, since it tends to sell off when the whole market is scared. The loonie carries oil but is also tied at the hip to the US economy next door. The krone is thinner and can be jumpy because the market for it is smaller. So you are never trading pure iron ore or pure oil. You are trading the commodity plus a basket of macro factors, and the art is knowing when the commodity part is the part that matters.
Second, the beta is not one. A large move in oil does not produce an equally large move in the loonie, because oil is only a slice of Canada's economy and the currency is buffered by everything else. Historically the currency moves a fraction of the commodity's percentage move, and that fraction wanders around depending on the regime. This is exactly why the pair is useful and also why it burns people who assume a fixed ratio.
Why a currency can beat the future
If you have traded crude futures you know the operational tax. Contracts expire, so you roll, and the roll costs you when the curve is in contango. Margin can get ugly on a fast day. Overnight sessions are thin. The FX version of the same view sidesteps a lot of that. Major currency pairs are among the most liquid instruments on earth, spreads on the Aussie and the loonie are usually tiny, and there is no expiry to manage because the position is just an open spot exposure with a small carry cost or credit depending on the rate differential.
The trade also lets you separate the parts of your thesis. If you think oil rises but you have no view on the dollar, you can express the oil part through the loonie and, if you want, hedge some of the pure dollar exposure through a second leg. You are decomposing a messy commodity bet into cleaner factor bets, which is usually where the better risk-adjusted return lives.
None of this makes the currency a perfect substitute. If your view is a sharp, dated supply shock in the physical, the future will track it more tightly and faster. The currency is the better vehicle when your horizon is weeks to months and your thesis is about the direction of terms of trade rather than a specific barrel arriving on a specific day.
The divergence signal
Here is the part I actually use. Because the currency and the commodity are chained together through terms of trade, they normally move in rough sympathy. When they stop, when oil is grinding higher and the loonie is flat or falling, that gap is information. One of the two markets is telling you something the other has not priced yet. Sometimes the currency is early, because FX traders are reacting to a rate story or a risk-off wave the commodity crowd has not absorbed. Sometimes the commodity is early, because a supply story is real and the currency is lagging. The gap does not tell you which side is wrong. It tells you to go find out.
A workflow I trust looks roughly like this:
- Track the currency and its anchor commodity as a ratio or a simple spread over a rolling window, so you are looking at their relationship rather than either one alone.
- Flag when that spread stretches beyond its normal range for the recent regime. Do not use a fixed threshold across all time; the normal range in a calm year is not the normal range in a volatile one.
- When it stretches, ask what else moved. Was there a rate decision, a risk-off day where the Aussie sold off with equities, a China data print? If a known macro factor explains the currency's move, the divergence is probably not a commodity mispricing, and you leave it alone.
- If nothing obvious explains the gap, that is your candidate. Structure the trade to bet on convergence, and size it small, because convergence trades are right often and wrong expensively.
Structuring it usually means taking the side you think is lagging and, where you can, hedging the factor you have no view on. If you believe the loonie is lagging a real oil move, you go long the loonie and you might short a broad dollar basket to strip out the pure dollar drift, so what remains is closer to your actual thesis. If you think the currency is right and the commodity is the one that overshot, you fade the commodity instead. The point is to isolate the disagreement rather than take both legs of noise.
Where this goes wrong
The failure mode I have watched most often is treating the correlation as a law. It is a tendency, and tendencies break exactly when the other factors take over. In a full risk-off panic the Aussie will fall regardless of what iron ore does, because at that moment it is trading as a risk asset and nobody cares about the ore. A rate divergence between two central banks can hold a currency away from its commodity for months, and if you keep adding to a convergence trade the whole way, the carry and the drawdown will grind you down before you are proven right. The krone punishes over-sizing because it is thinner than the other two and can gap.
So the discipline is to always name the non-commodity explanation before you put the trade on. If you cannot rule out that the divergence is really a rate story or a risk story, you do not have a commodity mispricing, you have a macro bet wearing a commodity costume. On the tooling side, watching the currency and its commodity on one cross-asset scorecard, the way we do inside Blockcircle, makes these gaps obvious instead of something you notice three days late. The signal is simple. Acting on it without checking why the two disagree is the expensive part.