February 4, 2013
Slicing Up The Global Equity Market Pie
One of the key issues in structuring an asset allocation strategy is deciding how to divide up the world's equity markets. Everyone has an opinion, but it's usually best to start with the standard benchmark, otherwise known as relative market values based on capitalization. You may choose to second guess Mr. Market's equity allocation, but you should at least have an idea of what you're modifying, if only for perspective. Are you making a big bet in Asia vs. Europe? Is your US allocation hefty vs. foreign developed markets? Are you underweight Japan vs. the rest of Asia? Knowing the answers to these type of questions isn't a silver bullet, but as a general rule it's useful to know how your choices on risk factors compare before you start reshuffling the market portfolio. Indeed, the information on relative market caps may end up informing your decisions on how to customize a portfolio.
The good news is that finding relative market weights is easy. The tough part is deciding what's relevant in terms of defining the planet's various equity betas. There are no standard answers because every investor has a unique objective, risk tolerance, and time horizon. Assets under management also vary, of course, which may influence how finely you slice the list of betas too.
Where to begin? With something simple and intuitive, and proceed from there. What follows is hardly the last word on how to divide up the global equity pie, but it's a reasonable start. The assumption is that we're looking to hold a global equity allocation. For simplicity, I'm ignoring intra-market factors, such as small-cap vs. large-cap or value-vs.-growth. Those are certainly relevant, and I'll review the data on those fronts at a later date. For now, the first cut is thinking about equity beta through a developed-vs.-emerging lens, and then zooming in a bit on the regional definitions.
Keep in mind that there are ETF proxies for each of the market categories listed below, and so it's easy and cost efficient to construct a global equity portfolio with a wide array of risk-factor tilts. With that in mind, the following data presents relative market weights from several global perspectives, based on market caps in US dollar terms, as of February 1, 2013, via S&P Global BMI Indices.
Let's begin with a simple US-vs.-the-rest-of-the-world view:
Now let's divide the foreign component into developed and emerging market slices:
Breaking up the foreign markets into additional pieces introduces another layer of questions. One that frequently comes up: How to treat Canada? This market is left out of some of the older-generation foreign equity indices, but that's an oversight that's easily resolved in newer benchmarks. For perspective, let's break it out. Note too that Israel is usually considered a developed market, but one that doesn't always fit into the usual categories.
Here's another look at the world's markets, this time in terms of developed regions alone:
Let's also consider how emerging markets compare on their own terms:
Finally, a couple of loose ends that often pop up in discussions about global market equity allocations, namely: Japan and China. For different reasons, some investors treat these markets as separate entities. Here's how they compare in relative terms vs. the rest of Asia:
Update: Here are the market-value allocations for the global bond market.
Posted by jp at February 4, 2013 10:14 AM
Yes, there are several alternative ways to weight equities. Earnings, GDP, and so on. I'm familiar with the research, although I don't have particular views one way or another. Ultimately, it's about developing confidence that methodology X is superior to methodology Y. In order to reach that conclusion, you ultimately need faith in a better model. Then again, some argue to reject models outright and instead equal weight. Lots of intriguing ideas here. For now, one question: What methodology has an encouraging track record for more than a few years, in a real-world product? There are a few possibilities--fundamental weighting, for instance--but the jury's still out. In any case, start with a market-weight benchmark and proceed from there.
Posted by: JP at February 9, 2013 6:48 AM
As always, good work.
Question about your breakdown of the US Equity markets. I assume this is market caps based on company domicile (headquarters). Have you seen analysis of where earnings are sourced? Meaning for US Large Cap, xx% of earnings come from "abroad". Percentage would no doubt be less for small cap. But, I wonder how much international earnings are pulled through by just using cap weighted Russell 3000? No doubt Int'l large caps source some earnings from US, as well. I guess my point is that an alternative to this technique would be weight geographies by GDP and source public earnings from each region accordingly. If you believe that (loosely) earnings come from GDP (with dilution). Thoughts?
Posted by: KSA at February 7, 2013 1:02 PM