Searching For The Market Portfolio
What is the “market portfolio?” When I first learned of the concept in business school it seemed so simple. You just combine all of the assets in the world and hold them in proportion to their market values. Why? Because academic theory says that’s the most efficient thing you can do, and trying to beat such a portfolio is an act of futility for most investors.
Moving on to the real world, I learned a few things:
1) Everyone has a different definition of “the market.”
2) Many try to beat it (regardless of what “it” is).
3) Even those that aren’t trying to beat it don’t own anything close to what fits the academic definition.
4) Buying all of the world’s assets, and doing so at a reasonable cost, is no easy task.
With the explosion of index ETF products and a sharp reduction in fees in recent years, the goal of approximating the market portfolio has become more attainable. But still, many questions remain:
- Are stocks and bonds enough (most advisors say “yes”)? If not, how many other asset classes should be included?
- Are U.S. assets enough (John Bogle says “yes”)? If not, how many other countries/regions should be included?
- If a broad bond index doesn’t include certain types of bonds (ex: high yield, bank loans, TIPS, municipal bonds), should you add those other bonds separately?
- If a broad equity index does not include certain stocks or countries, should you add those via other vehicles?
- How should you treat private equity/debt?
- What about MLPs, preferred stock, and convertible bonds?
- How should you treat commodities? Ignore them because futures contracts net to zero or calculate their weights based on their physical value? If on physical value is buying an ETF that buys futures contracts an accurate representation?
- Should Gold be treated as distinct from other commodities? How should one weight Gold, based on how much has been mined in all eternity or how much is currently held as an investment?
- Should real estate be weighted by investable asset size (including private real estate) or just the value of REITs?
- Does one simply ignore things like stamp collections, jewelry, fine art, expensive wine, vintage cars, and other collectibles?
- What about the value of private companies, private loans, etc.?
- Should the litmus test on whether to include an asset be whether it pays interest, a dividend, or has the potential to produce earnings/cash flow? If yes, how do you treat negative yielding bonds and commodities?
- For international exposure, should the weighting be currency hedged or unhedged?
- Is Bitcoin an asset?
The questions here are endless. The “market portfolio” in theory is a simple concept. In practice, it is exceedingly complex. But a decision still needs to be made: which assets do you choose to include and at what weights?
Let’s attempt to answer that question in stages, from the least expansive to the most expansion definition.
#1: The U.S. Stock-Bond Portfolio (aka the Bogle Portfolio)
The simplest and cheapest “market portfolio” can consist of just two funds: a total U.S. stock fund and an aggregate U.S. bond fund. John Bogle, the founder of Vanguard, advocates a 50/50 split between the two. As you can see in the following table, that’s a deviation from market weights which point to a 65/35 split with bonds representing the larger allocation.
The appeal of this portfolio is clear: simple (2 positions) and cheap (less than 4 bps blended). But is this really “the market?”
#2: The Global Stock-Bond Portfolio
While a U.S. only portfolio is the simplest and cheapest, it’s not “the market.” You can’t have a global market portfolio without including global markets. If we add international stocks and bonds, the picture changes quite a bit. The weighting to equities drops down to 28% (of which 46% is outside of the U.S.) and international bonds become the largest holding, at 42%.
How many investors would want to hold 42% of their portfolio in International Bonds with a current yield of 0.94%? My guess is not many, but the market portfolio does not discriminate based on yield or valuation – the market is the market and you must accept it or choose a more active path.
#3: The Global Stock-Bond-Real Estate Portfolio
Does it make sense to exclude the world’s largest asset class from the market portfolio? If you answered “no,” then real estate must be accounted for. While real estate is already included in the Vanguard Total World Stock ETF (VT), at a weighting of roughly 1% of that fund, it is highly underrepresented.
The total value of real estate globally is estimated to be $217 trillion, dwarfing the value of public equities and debt (see this Sevills report for data). Even if we only include the investable portion of residential and commercial properties – $73 trillion – it would drastically change the market portfolio. Global stock exposure drops down to 19% and real estate jumps to a total weighting of 31%.
#4: The “Almost Everything” Portfolio
Stocks, bonds, and real estate comprise the bulk of global market portfolio. Does that mean we should ignore everything else? That depends on how close to want to get to the true market portfolio.
If we include everything that can be closely replicated in an ETF (I chose the lowest cost ETF in each category), we arrive at the following portfolio:
The almost everything ETF portfolio:
- Fills in the gaps that the U.S. aggregate bond index excludes: high yield, bank loans, munis, money markets/cash, and TIPS.
- Adds Emerging Market Equities, Frontier Market Equities, and Emerging Market Bonds.
- Adds MLPs, Convertible Bonds, Preferred Stock, Mortgage REITs, and Private Equity.
- Adds Gold (at investable market size).
- Adds Commodities (at total commodity assets under management).
Is this the true market portfolio? No, not by a long shot. Our search for the market portfolio is seemingly unending.
But we have to start somewhere, for replicating the theoretical market portfolio is impossible and therefore not an attainable goal. The goal should be to replicate the most expansive investable portfolio possible that can be achieved in a low-cost fashion. With a weighted average cost of 16 bps, the low-cost objective is achieved with the “almost everything” portfolio. Whether it is expansive enough is still open for debate. But with more and more data freely available, new ETF products coming to market on a daily basis, and lower and lower fees, the gap between the theoretical and actual market portfolio is closing. This evolution can only be good news for long-term investors.
One thing stands out here, regardless of whether you agree with the asset classes or weights in the “almost everything” portfolio:
The standard 60/40 U.S. investor is making a very active bet as they are 1) vastly overweight the U.S. (home bias), 2) overweight equities, 3) significantly underweight real estate, and 4) effectively ignoring every non-mainstream asset class.
Running a 60/40 U.S. portfolio may be a good choice for many investors, but it’s a very active choice nonetheless. It would be hard to argue otherwise as 60/40 does not resemble anything close to the investable market portfolio.
This will be an ongoing project for me and I welcome all feedback and debate. What is the market portfolio to you? How close is your portfolio tracking it? If you are deviating from it do you consider that an active bet? Do you have a sustainable edge to warrant such a deviation?
Disclaimer: At Pension Partners, we use Bonds as our defensive position in our absolute return strategies for all of the above reasons. Bonds have provided a more ...
more