Just a post of some tabulated results today. You’ll recall that the CFA program spiraled me into a the more I learn, the less I feel I know paradox concerning individual stock analysis, so I basically decided to flee to asset allocation through indexing as a way to stay in the market while I caught up to speed on how to analyze individual companies properly and thoroughly.
Below is an overview of what I ended up settling on. For full disclosure, I did not sell all individual stocks and buy all indexes in one fell swoop. Instead I started steering my portfolio slowly over time towards the asset allocation defined below. As of year end though, 96% of my portfolio had been moved to the indexes and I ended up with an amazingly similar return to the one you’ll read about below.
The above allocation may look simple to implement, but my version wasn’t. That’s because I have further sub-classes to each pie slice above, and it can be hard to find ETFs that track all of these. For each of the 4 equity allocations (USA, Europe, Pacific, Emerging Markets) I wanted an equally weighted mix of the nine Morningstar equity style boxes pertaining to market cap (large, mid, small) and valuation (value, blend, growth).
Similarly, for the bond allocations I wanted an equal mix of duration, credit quality, and geographic location. For TIPS I wanted an equal mix of foreign and domestic. The same applied to real estate, along with a further equal mix of residential and commercial. For commodities I wanted an equal mix of metals, energy, and agriculture. You get the idea…
I also read that despite the academic studies concluding that asset allocation alone explains 90-99% of the variation in portfolio returns, some investment firms hire multiple managers to pick stocks from the same asset class to get manager diversification. When I had the freedom to do so, I tried to copy this by choosing ETFs from different firms. In some cases, this resulted in very different performance for the same asset class due to a different style of indexing (market cap weighted, equal weighted, fundamentally weighted).
Below are the details of all the portfolio positions.
20% USA Stocks: The USA allocation proved easy to implement, though there were a few wrinkles. Since 20% divided by 9 style boxes didn’t equal an integer percentage, I decided to add a 10th category which was simply another large-cap ETF, but one that was not market-cap weighted like all the others (RSP). As shown below, this 20% USA allocation returned a combined 33.2%.
20% Emerging Markets Stocks: Fewer style box options were to be found in emerging markets, but I was able to divide equally between large and small caps, and then between blend and value. For the large caps I was able to do a third division to get the above mentioned manager (or index) diversification. As shown below, this 20% Emerging Markets allocation returned a combined 78.7%.
20% Developed Europe Stocks: Same as above, but with another complication. I had trouble finding specifically European small caps so I had to go with ETFs that had a mixture of European and Pacific. SCZ is more like your traditional EAFE with a 2/3 weight in Europe and a 1/3 weight in Pacific. DLS is closer to a 50/50 split. Therefore here I only count 2/3 of the total 5% put into SCZ and 1/2 of the total 5% put into DLS to (imperfectly) account for only the European portion. As shown below, the resulting 20.8% Developed Europe allocation returned a combined 37.8%.
10% Developed Pacific Stocks: For the Developed Pacific portion, I targeted a smaller allocation of 10% just because so few countries comprise this region. For the standard MSCI Pacific index (VPL) Japan dominates, and therefore I diversified my large cap allocation across a second ex-Japan Pacific ETF (EPP). And similar to above, here we (imperfectly) account for the Pacific portion of the allocations to the small cap mixed indexes too. As shown below, the resulting 9.2% Developed Pacific allocation returned a combined 40.8%.
10% Inflation Hedge: Many would argue that this allocation isn’t really necessary, since stocks, real estate, and short-term bonds already provide sufficient inflation protection. But I know some Ivy League endowment managers use TIPS and I had worries about my government’s overuse of the printing press, so I defined my own specific inflation hedge asset class. Half is in commodities, and half is in TIPS. You have lots of choices for commodity ETFs but I tried to pick the one that wasn’t heavily weighted to any particular area (DJP). For the TIPS portion, I further divided equally across domestic and international for a sort of currency hedge. As shown below, the resulting 10% Inflation Hedge allocation returned a combined 16.2%.
10% Real Estate: Same drill, division of labor across foreign and domestic. As shown below, the resulting 10% REIT allocation returned a combined 33.5%.
10% Bonds: I won’t go through the myriad bond allocations., but if you look at the breakdown you’ll see a roughly even mix of foreign & domestic allocations to high, medium, and low credit quality for short, intermediate, and long terms. As shown below, the resulting 10% bond allocation returned a combined 14.8%.
So that’s it, and everything above combined to return 40.5% for 2009. Not bad at all. 😉
Obviously my mindset was to keep an equity tilt, but own a little bit of everything and not too much of anything, to hedge against Rumsfeld’s “unknown unknowns“. A more sophisticated modern portfolio theorist might compute correlations to settle on the least volatile weighted mix of all of the above. Many of the sub-asset classes may contribute nothing but expenses.
I believe my “equal division” approach is only theoretically optimal for asset classes that are independent and uncorrelated, but psychologically was something I had no trouble sticking to during the darker days of the financial crisis.