Selecting a Benchmark for a Multinational Company’s Equity Beta

Summary

The document considers how to estimate the equity beta of a multinational company listed in the Netherlands, whose business spans several regions. A regression against the domestic AEX index is reported to produce a beta below one and a low R-squared, prompting the question of whether a different benchmark would better reflect the company’s risk exposures. The discussion offers two approaches: construct a bespoke index weighted by the company’s regional business shares, or use a broad global index such as a developed-market or all-country benchmark.

The weighted-index proposal aims to align the regression benchmark with the geographic composition of the company’s operations, while a global index provides a ready-made alternative. The answers do not compare these benchmarks empirically or establish that revenue or operating shares are the right weighting basis. Beta estimates also depend on return frequency, currency, market coverage, and the chosen estimation period. The suggestions are therefore starting points for benchmark selection rather than evidence that one index will produce a more reliable valuation input.

Key ideas

• A domestic index may poorly reflect a multinational company’s geographic risk exposures.
• A bespoke benchmark can weight markets according to the company’s regional business mix.
• Broad global indices offer an off-the-shelf alternative to a custom benchmark.
• The document does not establish which weighting basis or index yields the most reliable beta.

Tags

# What is the best benchmark index for computing the beta of a multinational company?


# What is the best benchmark index for computing the beta of a multinational company?












I'm running a valuation of a multinational company listed on the AEX (Amsterdam Eurononext). The company has operations in Europe (70%), US (25%) and other (5%). I have historic stock data until from 2000.

In order to determine a risk profile I ran a regression of its returns against the AEX. As expected this yielded a beta <1.0 and an R-squared of < 0.5.

What would be a better index to run my regression against? Data does need to run back until at least 2000.

## Answer by SolitonK (score 2, accepted)

https://quant.stackexchange.com/a/32213

Matthias, I would create a bespoke index weighted on the percentages you mentioned above.

In that way you can be sure of capturing risk at the appropriate levels without overexposing your valuation to only one or if you use a global index, multiple markets with zero importance.

## Answer by Comp_Warrior (score 2)

https://quant.stackexchange.com/a/32152

If you want to use a more global index for your regression, you should consider these two:

- MSCI World (Developed Markets)

- MSCI ACWI (Developed and Emerging Markets)

EDIT: You might consider using global indices by other providers, for example, the S&P Global 1200. You can get 10 years of data from the link.

Shown in full with attribution under the source's licence. Licence: CC BY-SA 4.0 (Stack Exchange)

This summary was written by Stratmill's research agent from the original; it is not a copy of the source.