Forex correlation: how bonds, commodities and equities move currencies

What you’ll learn

• How the four main asset classes transmit signals into FX
• Distinguishing risk-on from risk-off regimes
• Yield spreads and their role as a leading indicator
• Spotting when a durable correlation stops working
• Building a pre-trade correlation checklist
• Common beginner mistakes and how to sidestep them
• A worked example of a correlation-based trade, with position sizing

The four-market framework

Price action across bonds, commodities, equities and currencies is linked. Rate expectations set the direction for bonds. Those bond moves spill into currencies. Commodities, in turn, answer to the dollar, while equities respond to both rates and growth. Isolating any one market means discarding most of what the other three are saying.

All four of those markets are tradeable from one account through TabTrade’s Instruments list, which is what makes the linkages usable rather than academic.

A simple way to picture the flow:

When the Fed turns hawkish, US yields rise. The dollar strengthens. Commodities priced in dollars become more expensive for non‑USD buyers, so they often fall. Equities may sell off if higher rates threaten growth. The chain rarely breaks cleanly, but understanding the direction of travel stops you fading a move that is being reinforced by three other asset classes.

Risk-on / risk-off

Regime	Bid	Offered
Risk-on	AUD, NZD, CAD, equities, EM	JPY, CHF, gold
Risk-off	JPY, CHF, USD, gold, treasuries	AUD, NZD, equities

Establish the regime before anything else. Without that filter, a long AUDJPY that looks technically clean in a risk-off tape becomes a technically clean way to lose money.

How do you spot the regime? Watch three things at the same time:

1. Equity index futures (S&P 500, Nasdaq). A sustained move of more than 0.5% in the first hour of the European session is one of the readings traders take on the day’s risk appetite.
2. The VIX. Readings above 25 have historically coincided with risk-off flows, and readings below 15 with risk-on conditions.
3. JPY crosses. If USDJPY, EURJPY and AUDJPY are all moving in the same direction with conviction, the market is trading a single risk theme.

A common beginner mistake is to treat every down day in equities as a risk-off regime. A 0.3% dip in the S&P 500 while the VIX stays at 14 is noise, not a regime change. A single signal is weaker evidence than two of the three moving together.

Yield spreads lead

For a currency pair, the gap between two economies’ 2-year yields is among the most closely watched macro inputs. Historically the pair has tracked that gap, with the yield move arriving first.

Using dated illustrative inputs, suppose the US 2-year yield is 4.80% and the German 2-year yield is 2.60%. The spread is 2.20 percentage points in favour of the dollar. If US yields climb to 5.00% while German yields stay flat, the spread widens to 2.40 percentage points. That extra 0.20 percentage point raises the carry on dollars, which is the mechanism the correlation runs through.

The cost assumptions here follow TabTrade’s April 2026 spread measurement.

If a pair separates from its yield spread for more than a handful of sessions, one of two conditions applies. Either the market is pricing something the spread has not yet reflected, or a mean-reversion setup is forming. Decide which condition holds before entering the trade.

A practical way to track this is to overlay the 2-year yield spread on a daily EURUSD chart. When the spread makes a new high and EURUSD does not make a corresponding low within two days, the divergence is worth investigating. It often resolves with the currency catching up to the bond market, but not always. If the divergence appears during a central bank blackout period, the market may be front-running a policy surprise. In that case, waiting for the actual announcement is usually safer than fading the move.

Commodity currencies

• CAD follows crude oil because energy dominates Canadian export flows.
• AUD follows iron ore, copper and Chinese demand conditions.
• NZD responds to dairy prices and broad risk appetite.

The inverse link between USDCAD and oil is among the more persistent macro relationships on record. When it decouples, the usual explanation is that a rate decision has taken over as the driver.

The table below summarises the primary driver for each commodity currency and the typical direction of the correlation.

Currency	Primary driver	Typical correlation with driver	Notes
CAD	WTI crude oil	Negative (USDCAD falls when oil rises)	Strongest during steady rate environments
AUD	Iron ore, copper	Positive (AUDUSD rises with metals)	Chinese data releases can override
NZD	Whole milk powder, risk appetite	Positive (NZDUSD rises with dairy)	Less liquid; spreads can widen sharply

A worked example for USDCAD:

WTI crude rallies from $75.00 to $80.00. The trader expects USDCAD to fall. They sell USDCAD at 1.3650 with a stop 30 pips above at 1.3680 and a target 60 pips lower at 1.3590. Account size is $10,000. Risk per trade is 1%, or $100.

Pip value for a standard lot of USDCAD is roughly C$10, which varies with the exchange rate. For a quick estimate, use $7.30 per pip when USDCAD is near 1.3650. The stop distance is 30 pips, so the risk per standard lot is 30 × $7.30 = $219. To risk only $100, the trader uses 0.46 lots, rounded to 0.45 lots. The arithmetic:

• Risk per pip on 0.45 lots = 0.45 × $7.30 ≈ $3.29
• Stop loss: 30 pips × $3.29 = $98.70 (within the $100 limit)
• Target reward: 60 pips × $3.29 = $197.40

The trade uses the oil correlation as the thesis. If oil suddenly reverses, the trader exits regardless of whether the stop is hit, because the driver has changed.

Gold and the dollar

Because gold is quoted in dollars and carries no yield, it has traditionally moved inversely to both the dollar and real rates. That relationship has broken down repeatedly during central-bank buying cycles and periods of geopolitical stress. Correlations depend on the regime; they are not laws.

A trader who shorts gold solely because the dollar index is up 0.3% is ignoring the last five years of market behaviour. Between 2022 and 2024, gold rallied alongside a strong dollar as central banks diversified reserves. The old inverse correlation went dormant. Treating it as a permanent fixture would have produced a string of losses.

When you see gold and the dollar moving in the same direction for more than a week, ask what else is driving gold. Often it is real-rate expectations or safe-haven demand that has decoupled from the dollar. On TabTrade’s platform, gold is one of 14 commodities available as a CFD, and it can be traded against the dollar or cross-referenced with silver and copper to gauge whether the move is precious-metal specific or a broader commodity flow. Watching gold alongside the US 10-year TIPS yield adds a second layer of confirmation.

Checking for breakdown

Track a rolling 20-day correlation. If a relationship normally sits near −0.8 and drifts toward −0.2, something new is driving the market. Treat that shift as information rather than noise to be averaged away.

A simple way to calculate a rolling 20-day correlation in a spreadsheet:

1. Collect daily closing prices for the two instruments over at least 60 days.
2. In a new column, compute the daily percentage change for each.
3. Use the CORREL function on the last 20 pairs of daily changes.
4. Drag the formula down to create a rolling series.

When the 20-day correlation crosses above −0.5 from a deeper negative reading, the relationship has weakened materially. That does not mean the old correlation will never return. It means the current regime is different, and your trade needs a different thesis.

A mistake traders make is to average the correlation over a longer period to make it look stable. A 60-day correlation that still reads −0.7 might hide a 20-day reading of −0.1. The shorter window shows the regime shift earlier; the 60-day number describes the longer backdrop.

A practical correlation checklist

Before you enter any trade that leans on an intermarket relationship, run through five questions. If you cannot answer all five, the correlation is not solid enough to bet on.

1. What is the current risk regime? (risk-on / risk-off / mixed)
2. Which yield spread is driving the pair, and has it moved in my favour?
3. Is the relevant commodity confirming or contradicting?
4. Is gold behaving as expected for this regime?
5. Has the 20-day correlation weakened in the last week?

If questions 1, 2 and 5 all point the same way, the macro backdrop lines up with the trade. If question 5 shows a breakdown, the thesis the size was based on no longer holds.

Mistakes that cost traders money

• Treating correlation as causation. Just because AUDUSD and copper have moved together for months does not mean a copper rally will push AUDUSD higher today. Both may be reacting to a third factor, such as Chinese credit data.
• Ignoring time zones. The correlation between USDCAD and oil is strongest during North American hours. During the Asian session, the pair may drift on flows that have nothing to do with crude.
• Overcomplicating the chart. You do not need five correlation overlays. Pick one yield spread, one commodity and the risk regime. Three clean inputs beat ten noisy ones.
• Doubling down when the correlation breaks. If USDCAD rallies while oil rallies too, the market is telling you something. Adding to a losing position because “the correlation will come back” is a fast way to blow up an account. Reduce size and reassess.

Worked example: trading a yield-spread signal

A trader watches the US - German 2-year yield spread. Using the dated illustrative set above, the spread has widened from 2.20 to 2.40 percentage points while EURUSD has not yet moved lower, creating a divergence. The trader expects EURUSD to catch up.

Account balance: $5,000. Risk per trade: 1% ($50). Using illustrative prices, the trader sells EURUSD at 1.0850. Stop is 25 pips above at 1.0875. The first target is 25 pips below at 1.0825, and the final target is 50 pips below at 1.0800.

Pip value for one standard lot of EURUSD is $10. Risk per standard lot over 25 pips is $250. To risk $50, the position size is:

$50 ÷ $250 = 0.20 lots (two mini lots)

Gross reward if half exits at 1.0825 and half at 1.0800 is $25 + $50 = $75 before costs. Against $50 of planned risk, the blended risk:reward ratio is 1:1.5.

The trader checks the Edge account cost on TabTrade. EURUSD average spread is 0.03 pips, and the round-turn commission is $7 per lot. For 0.20 lots, the total cost is roughly:

0.20 × $7 = $1.40, plus 0.03 pips of spread worth $0.06. Total cost about $1.46, a fraction of the expected move.

The trade works because the yield spread provided a leading signal. The trader exits half the position at 1.0825, moves the stop to breakeven on the remainder, and exits the remainder at 1.0800. If the spread suddenly narrows before either target is reached, the trader closes the trade early because the macro reason for being in it has disappeared.

Rule of thumb: A correlation is a compass, not a GPS. It tells you the general direction, but you still need to watch the road.