Ask Dave: Index ETFs and Rebalancing (or lack therof)

I while ago I bought Vanguard Europe Pacific ETF (VEA) for the international portion of my portfolio and one reader had the following comment:

Just a quick question. With regards to balancing one’s portfolio, would ETFs like VEA (and I see there is now one that encompasses the whole world excluding the US), not pose a problem? If for example the European markets did well one year but the Pacific markets performed poorly, then one would be unable to rebalance by selling a European based ETF’s (like VGK for example) and buying more Pacific based ETF’s (VPL for example). With everything in one basket one could not take advantage of the gains to be made by selling high and buying low. Is this assumption correct? Or is it true that because VEA comprises 75% VGK and 25% VPL, that it would reflect any net changes made by owning a combination of both VGK and VPL?

Your assumption is correct, one could not take advantage of the gains to be made by selling high and buying low the stocks in one region vs. another. These rebalancing “bonuses” are small, but more importantly, they may completely disappear after trading costs are taken into account. VEA does not contain a fixed percentage of VGK and VPL underneath. It contains the market cap weighting of all its components. Market cap-weighted indexes have several advantages as investments:

Market value-weighted indexes have lower trading costs. If you made your own index, and the index never added or removed stocks the stocks would be bought once and never sold. As one stock goes up in value, it maintains the desired allocation in the index. This keeps trading costs low as one essentially rarely needs to make trades, except for when stocks are dropped or added from the index. Market value-weighted indexes have the lowest MERs, and other indexes like fundamental-weighted indexes, or fixed-weight indexes have higher MERs due to increased trading within the index.

With market value-weighted indexes, one never ends up holding on to dogs. Imagine it is the early 1900s and you own shares in a fixed sector-weighted index (did such a thing exist back then?). The index contains 25% financials, 25% railways, 25% consumer goods, and 25% manufacturing (or a fixed percentage of each individual stock, it’s the same thing). Every year the index is rebalanced. Pretty soon railways start to go out of style as air travel is invented and highways are built. The 1960s arrive and “tech” stocks are all the rage. The index rebalances religiously, which leads it to purchase a lot of railway stock which continues to do badly, meanwhile it has missed out on the rise in tech stocks. The worse railways do, the more you have to buy in order to keep your portfolio balanced (in addition to paying more commissions). You now wish the index had altered their weightings to contain less railways. One way to do it might have been to not fix the amount allocated to each sector and just use market cap-weighting instead. Or use market cap-weighting while taking into account large changes in what is going on in the market. For example, one could have envisioned a “smarter” version of the S&P 500 index that underweighted tech/IT stocks in around 2000 (selling high) and went back to their appropriate market cap-weight in 2002. Of course we only wish we had a crystal ball back in 2000 that could have warned us that tech stocks were about to fall. It seemed inevitable, just as it did in 1997. Unfortunately there is no foresight in the market and there was no telling if tech stocks would not have kept climbing after 2000. Furthermore, if you believe that markets are largely efficient, there is no such thing as “buying low” or “buying high” and it is not possible to “take advantage of the gains to be made by selling high and buying low” as there are essentially no “gains” to be made. There are only “losses” to be incurred through increased costs of trading and higher turnover.

In the scenario you described you said “If for example the European markets did well one year but the Pacific markets performed poorly, then one would be unable to rebalance by selling a European based ETF’s (like VGK for example) and buying more Pacific based ETF’s (VPL for example).” Personally I would get a bit nervous in buying more VPL. How do you know it is at a “low”. What if it actually at a “high” of even greater future “lows.” I was in an investment club from around 1999-2005 (the club ended in 2005). We consistently thought we were buying stocks at “lows.” We bought stocks like Nortel, Nokia, 360 Networks, Global Crossing, Lucent, etc… after sharp declines only to watch them decline even further (often with another purchase on the way down for good measure). So this is the problem I have with fixed allocations and rebalancing. If you bought 75% VPL and 25% VGK and held them in your portfolio. How do you know if 25% VGK is “just right”, “over-valued”, or “under-valued.” If it falls to 15% now how would you classify it? If it was over-valued before, now it might be just right. If it was just right before it might be considered under-valued now. But if it were truly under-valued (in the sense that the equities in VGK are worth far more than what the market is valuing them at) most likely other smart investors would have already taken advantage of it (as if you’d be the first to realize it!) and so it’s most likely that those equities are “just right.”

I blogged a lot about Equal Weight Indexes in the past:

• In Non-Market Cap Weighted Indexes: The Next Big Thing I ballyhooed equal-weighted indexes and lamented the lack of equal-weighted indexes in Canada. My opinion of equal-weighted indexes would soon change.
• In Equal-Weight S&P 500 Index I came to the full realization that RSP, the S&P 500 Equal Weighted index only performed better because of its higher concentration of mid-cap stocks.
• In Too Many Choices (or why I am ready to give up) I mentioned that “I have now come to the realization that RSP is very similar to the S&P 400 Midcap Index which is also available as an ETF (MDY). It is very highly correlated . . .”

Regrettably, I don’t think equal weight indexes are all that I initially hyped them to be. I did some searching on the web for more information on the advantages of market cap-based indexes over fixed weight indexes but had trouble finding any information at all, however, I think my reasoning above makes sense. As in many arguments over financial instruments and investing strategies, cost is again a huge factor.