I am seeking a full-time position in (southern) Germany or Switzerland starting from October related to my skills as demonstrated on this blog. If you know of any opportunities please message me on LinkedIn.

Since I often have to explain my mixed feelings regarding ETFs, I thought I should write a post about it. So, here it is.


The main advantages of ETFs are twofold: they are cheap and they provide you with some diversification. Often the annual fee is only around 0.5% which is less than the fee charged by actively managed funds. For a retail investor who does not actively want to rebalance its portfolio, ETFs are nice, since they provide exposure to different market factors. Since an ETF is a basket of stuff it has less volatility than its underlyings (depending on their covariance of course).


My first objection to ETFs is that they track a market index. A market index was originally invented as a way to judge the health of the economy and not as an investment instrument. Thus, they often have weird weightings. The S&P 500 (for my international readers), as well as the german DAX (for my german readers) are weighted by capitalization. That means that the higher cap a company has, the more weight its shares have in the index. For the S&P that means that Apple, Amazon and Microsoft make up most of the value. The top 10 shares influence the price as much as the bottom 300. Equal weighting or random weightings outperform the S&P 500. Thus, you do not get as much diversification as expected. And you also do not get as much profits as you could get, since it is more unlikely that a blue chip doubles in value than one of the cheaper shares.

My second objection is that ETFs make the markets more stupid. If ETFs are significantly more actively traded than their underlyings, then the correct pricing of the underlyings becomes more inefficient. Relative performance changes between different companies would not have such a direct effect on the price of the shares, because in the ETF one always buys the same proportion of shares.

Let me make this more clear with an exaggerated example. Assume we have two companies AAAA and BBBB. Assume further we have an ETF which consists of 20% AAAA and 80% BBBB. If everyone would only buy or sell the ETF, but not AAAA or BBBB, then the relative prices of the companies would never change. The value of AAAA would be 80% of the ETF, the value of BBBB would be 20% of the ETF. Thus, BBBBs value would always be 1/4 of the value of AAAA, and relative price changes would not occur.

Another way in which ETFs make the markets dumber is the increase in correlation among the assets in the same ETF, even if they have nothing in common otherwise. Actually, I am not yet sure if this is a bad thing, since it could also mean that factor specific risks are priced more accurately.


As you can see, though ETFs have their advantages, they also have a bunch of drawbacks which do not affect the individual investor, but rather the market as a whole. Moreover the drawbacks require a more nuanced argumentation to state them and thus may not be widely known, especially among hobby investors.


27 January 2018


Quantitative Finance