VT - ACWI Vs. VT: Vanguard Wins Again

Summary

• In this article, I focus on the iShares MSCI ACWI ETF (ACWI) and how it compares to the Vanguard Total World Stock ETF (VT).
• For the longest common period since June 2008, ACWI currently lags VT by about 14%-points or 41 basis points per year.
• The performance gap mostly comes from different underlying indices. VT tracks an index with >9,400 stocks whereas ACWI ignores small caps and "only" holds about 2,800 positions.
• ACWI is thus farther away from the academic idea of truly passive investing (holding a market-cap weighted portfolio of all investable stocks).
• ACWI also comes with higher fees (0.32% TER vs. 0.07% for VT). For investors who seek passive exposure to global stock markets, VT therefore seems the better choice.

After focusing on the logic and idea behind global diversification in my last article on the Vanguard Total World Stock ETF ( VT ), this one will be more about concrete investment products. Specifically, I will focus on the iShares MSCI ACWI ETF ( ACWI ) and how it compares to VT. Similar to the large index funds for the US market (see my articles on VOO here , VTI here , and ITOT here ), this is again a battle of the two giants, iShares a.k.a. BlackRock and Vanguard.

I will not repeat the details about passive investing at this point and refer interested readers to the aforementioned articles. My working definition of truly passive investing for the rest of this article is to simply hold the global market portfolio. Applied to equities, this is the market-cap weighted portfolio of all investable stocks in the world.

ACWI vs. VT - Performance since June 2008

I think the best way to start is a simple performance comparison of the two ETFs for the longest common period since June 2008 (ACWI was incepted on March 26, 2008, VT on June 24). The following chart shows cumulative total returns.

First observation: ACWI is somewhat behind VT (142.2% vs. 156.3%, respectively). Over the almost 15 year period since June 2008, this translates into an annualized underperformance of about 41 basis points (6.66% return p.a. for VT vs. 6.25% for ACWI). Given that both funds attempt to deliver exposure to global equities, this is quite surprising. So let's see what may caused the difference.

The first and obvious point are the underlying indices. As I mentioned in my last article, VT tracks the FTSE Global All Cap Index which consists of more than 9,400 stocks from both emerging and developed markets. In contrast, ACWI, as the ticker suggests, tracks the MSCI ACWI Index. This index also captures stocks from both emerging and developed countries but without small caps. As a consequence, it "only" consists of about 2,800 stocks.

These are of course about 6,500 stocks less, but given that both indices are market-cap weighted, the difference should be less severe than you might think. Because of their low market-cap, the weight of most small-caps in the FTSE index is probably just in the range of a few basis points.

Unfortunately, I don't have direct access to the time series of the FTSE Index. So I cannot compare the indices directly and even the factsheets don't help because MSCI reports Net Total Returns and FTSE Gross Total Returns. To still isolate the effect of the different index construction, I compare the MSCI ACWI Index with the MSCI ACWI IMI Index. "IMI" stands for Investable Market Index and in this category, MSCI also considers small caps.

As a consequence, the MSCI ACWI IMI Index is more similar to the FTSE Global All Cap Index as it also covers more than 9,000 stocks. Using MSCI's end of day index data search , we can see that the IMI version was indeed slightly better than the "normal" MSCI ACWI. Over the period from June 30, 2008 to December 30, 2022, the difference amounts to 3.6%-points or 11 basis points per year. This result indicates that the more comprehensive index construction of FTSE was historically beneficial and explains some parts of ACWI‘s underperformance versus VT.

As you can imagine, I think this is a nice result. Why? Because the FTSE Index is closer to the theoretical ideal of the global market portfolio. 9,400 stocks is undeniably closer to "all investable stocks" than 2,800. Although the performance difference is certainly not statistically significant in this case, I like it when the data supports what theory tells us to do.

Some more details about ACWI

So the fact that VT also invests in small caps already explains some of the differences. But what about the rest? The next point is fees. Fees are deducted from performance, so all else equal, the cheaper ETF should outperform the more expensive one. VT comes with a total expense ratio of just 0.07%. ACWI, in contrast, charges with 0.32% more than 4-times more. In theory, this is another 25 basis points performance gap. So together with the 11 basis points from before, we would be already close to our observed underperformance of ACWI.

Unfortunately, it is not that simple. Although total expense ratios are definitely important and you should monitor them carefully, for ETFs it is ultimately important to track the underlying index as close as possible. Suppose you have two ETFs on the same index. The first one charges 0.1% and trails the index by 10 basis points, the second one charges 0.2% but trails the index by only 5 basis points. Although the first one is cheaper, the second one is still better with respect to its mandate of tracking the index. So costs are important but not everything.

Looking at the official website of ACWI, we can see that the ETF delivered 5.74% p.a. since inception which is actually better than the stated benchmark return of 5.58% p.a. (as of December 31, 2022; Website accessed on January 29, 2023). This kills the higher expense ratio as reason for the underperformance. If anything, iShares apparently found ways to slightly outperform the underlying index in the past. This also means that the ACWI ETF over-delivered on its mandate of tracking the MSCI ACWI Index which deserves credit!

On the other hand, it also means that the historical performance gap between ACWI and VT seems to come entirely from different underlying indices. Given that we already looked at the more comprehensive MSCI ACWI IMI Index, there thus seem to be some more differences in the index construction of MSCI and FTSE. However, at this point we already have enough arguments for what is probably the better choice (see Conclusion below), so digging deeper into such details doesn't seems worthwhile.

Conclusion

After using the acronym ACWI 23 times throughout this article, I think it is time to wrap it up. For the longest common period since June 2008, ACWI underperformed VT by 14.1%-points or 40 basis points per year. Although ACWI is considerably more expensive than VT (0.32% TER vs. 0.07%), the underperformance mostly came from different underlying indices. VT's FTSE Global All Cap Index consists of more than 9,400 stocks whereas the MSCI ACWI Index ignores small caps and "only" captures about 2,800 positions.

Which one is better? If you read my previous articles, you probably know what's coming. I think it is fair to assume that investors who buy either ETF want to have passive exposure to global stock markets. The (in my opinion) easiest (and only logical) way to decide between the two ETFs is to simply follow the theory. If we want to be passive we should take the ETF that is closer to the theoretical ideal of the global market portfolio. And with about 6,500 stocks more than ACWI, this is clearly VT. The fact that it is also considerably cheaper and that it delivered better performance over the last 15 years are nice add-ons, but I think the first argument is much more important. You should always pick the ETF with a process that's closer to what you are actually trying to achieve. For global passive investors, this is VT.

For further details see: