I’m in Chicago this week for the 65th meeting of the Performance Measurement Forum. On Thursday we’ll address the topic of multi-currency attribution, and I’ll do a brief presentation on WHY it’s necessary for firms that invest across currencies to include at least one currency effect in their attribution.
We have found that many asset managers who invest across different currencies avoid the currency effect: this is hugely problematic. Why? Well, we know that exchange rates fluctuate, meaning that as we move in and out of currencies, our performance will not just be from what happens in the securities they’re invested in, but also the change in FX rates.
How challenging is multi-currency attribution?
Multi-currency attribution probably sounds a bit scary, and perhaps with some justification. If you’re not only exposed to the change in FX rates’ impact on the securities held, but also invest in currency forwards, for example, you’ll want to distinguish between the contributions from both: this is what an attribution model such as Karnosky/Singer does. However, if you’re not investing in such derivatives, then a more naïve model may suffice.
The shortcomings of avoiding multi-currency attribution
To ignore the contribution from changes in FX rates means you’ll provide less complete and accurate results, which may cause the misallocation of praise or criticism. Such sensitivity should be mandatory if the portfolio holds assets that are denominated in different currencies or if you’re investing for individuals whose base currency is different than the local currency of their investments.
My talk this week will address currency’s impact on performance. I’ll address this topic in greater detail in this month’s newsletters.
At BI-SAM’s recent online panel discussion, I mentioned that a challenge that exists is the failure of too many firms to include the currency effect in their attribution model. I will continue to champion this topic in the coming months.