Currency Attribution

A US-based fund buys a French stock that returns 8% in euros. Over the same period the euro weakens 5% against the dollar. In dollar terms — the currency the US client actually cares about — the position returned roughly 3%, not 8%. Was that a bad investment, or a good investment undone by a currency move the manager wasn't trying to make? Without splitting the two apart, a client has no way to tell whether their international manager is good at picking stocks, good at picking currencies, or neither, and those are genuinely separate skills usually assigned to different people or processes.

Splitting the two returns

The exact relationship, for a single foreign holding, is:

$$1 + R_{d} = (1 + R_{local}) \times (1 + R_{FX})$$

In words: the dollar return an investor actually earns equals the local-market return compounded with the currency's return against the dollar, not simply added to it, because the currency move applies to the already-grown local value, not to the original investment. Rearranged, the pure currency contribution is roughly $R_{FX} + R_{local} \times R_{FX}$, but for small moves the cross term is tiny and analysts often use the simpler approximation $R_{d} \approx R_{local} + R_{FX}$.

At the portfolio level, the manager's currency decision is usually isolated further by comparing the fund's actual currency exposure (which may be partially or fully hedged) against the benchmark's currency exposure, so a "currency effect" line captures the manager's active hedging or unhedging choice, not just the passive fact of holding foreign assets.

Worked example

The French stock above returns 8.00% in euros. EURUSD moves from 1.10 to 1.045, a 5.00% decline in euro value.

1. Compounded dollar return. $1.08 \times 0.95 - 1 = 1.026 - 1 = 2.6\%$. That's the actual dollar return an unhedged US investor earned.
2. Approximation check. $8.00\% + (-5.00\%) = 3.00\%$, close to the exact 2.6% but not exact — the gap is the cross term, $8.00\% \times (-5.00\%) = -0.40$ percentage points, which the simple sum ignores.
3. Attribution split. Local return contributes 8.00%, currency contributes approximately $2.6\% - 8.00\% = -5.4\%$ once the interaction is folded in. A client sees: strong local stock-picking, more than offset by adverse currency, netting to a modest 2.6% gain.

What this means in practice

Global equity and fixed income shops report currency attribution separately because many institutional mandates explicitly separate the two decisions: a manager may be told to fully hedge currency exposure and be judged purely on local-market stock selection, or explicitly permitted to run currency views as a separate source of active return. A pension fund that discovers its "stock picker" was actually losing money on stock selection and making it back on unauthorized currency bets has a real governance problem, not just a reporting curiosity.