The Hidden Risk of ETFs

A Bit of Background


Mike Akeroyd

3 years ago | 6 min read

Passive investing was recently identified as a bubble by Michael Burry of ‘The Big Short’ fame. He called out this style of investing as a bubble like the tech bubble of the late 90’s and mortgage securities bubble of the mid-2000’s.

The premise is interesting because when it comes to investing, it is well accepted that one cannot consistently beat the market through active investing.

The common advice is to invest passively by attempting to track an index. John Bogle and Vanguard made a career and business respectively out of that advice. Passive investing has become so common that the total invested value of exchange traded funds (ETFs) has grown to over $6.1T in assets in 2019.

ETFs are an investment fund that holds stocks, bonds, commodities, or other asset classes and usually tracks to an index.

Growth of Worldwide ETFs over Time; Data from Statista

A Bit of Background

Before I breakdown some potential risks associated with passive investing, I would like to cover a few basics. There are a variety of ways to make passive investments.

They include purchasing shares in an index-tracking mutual fund, purchasing shares in an ETF, or buying the underlying assets in an index. For example, if your goal is to match the return of US Large Cap stocks, then an index to track could be the S&P500. You could either buy a mutual fund that tracks this index like the FXAIX or buy shares in an ETF such as the SPY.

You could also buy most or all of the 500 underlying stocks in the appropriate proportions to the index.

There are subtle differences between these three approaches. Right off the bat, purchasing the underlying stocks is going to be time consuming. It will need continual management to ensure that the portfolio is in alignment with the index. Transaction costs will also be significant. By comparison, there won’t be any transaction costs associated with portfolio re-balancing if you buy a mutual fund or ETF.

There will still be entry and exit transaction costs associated with the ETF. In some cases, the mutual fund will also have transaction costs depending if there are any applicable loads.

One can also buy the underlying stocks in the ETF and the ETF shares intra-day. For mutual funds, investors can place their orders intra-day but they are not filled until after market close. If placed after market close, orders are filled after market close the next day.

The value of an ETF is derived from the value of the fund’s underlying holdings. Theoretically, there should be no difference between the value of the ETF and its net asset value.

The net asset value is the total value of the ETF’s underlying assets. For example, an ETF has 100K shares outstanding and the total value of its underlying assets total $15M. This means the price per share of the ETF should be $150 ($15M / 100K). However, theory can differ from reality.

Supply and demand pressures may not always be proportional between the ETF and underlying holdings. For example, if the broad market is on an upswing, and investors make disproportionately larger bets on the ETF than on the underlying holdings, then a disparity in value can occur. To reduce these disparities, there are arbitrage mechanisms that exist.

If the value of the underlying portfolio is worth less than the value of the ETF, then authorized participants can buy the individual securities from the secondary market. The authorized participants, often broker-dealers, can then sell short the ETF. This locks in the arbitrage gain but the short sell obligation exists.

After the market close, the authorized participant can deposit the individual securities with the ETF’s issuer. These individual securities are also known as a “creation basket”. The issuer will then convert these into a “creation unit” or new ETF shares. The authorized participant receives these ETF shares from the issuer. These new ETF shares are used to close out the intra-day short position.

But what happens if this arbitrage process breaks down?

On to the Risk…

This is where it starts to get interesting. In every index, there are the household names like Apple (AAPL), General Motors (GM), and Wells Fargo (WFC). Usually, these companies make up a large part of the index. There are also smaller companies included in the index but they represent a smaller allocation.

This also happens within ETFs and passive mutual funds. For example, the Vanguard High Dividend Yield ETF (VYM) allocates roughly 10% of the portfolio to three companies. These three companies each have a $300B+ market cap: JPMorgan Chase (JPM), Johnson & Johnson (JNJ), and Procter & Gamble (PG).

On the other end of the spectrum, this ETF also invests a fraction of its portfolio in a company with a $400M market cap: Green Plains Inc (GPRE). Smaller companies usually have less trading liquidity making ETF arbitrage challenging.

Excluding inverse, active quant, or leveraged ETFs, I found approximately 800 ETFs that specialize in US Equities. To run a quick analysis, I only looked at the largest 200 representing 95% of the total assets of ETFs specializing in US equities. Within these ETFs, there are 3,800 stocks and on average, each stock is a component within at least 20 of those 200 ETFs.

The chart below shows all these stocks plotted by daily turn-over and how much of the stock’s market cap is contained in ETFs. Daily turn-over is a measure of liquidity.

Specifically, how much of the stock’s value is traded on a daily basis. Usually a higher turn-over percentage is an indicator of higher liquidity.

The metric on the y-axis is a measure of the stock’s exposure within ETFs. The higher the exposure, the more the stock could be impacted by the dynamics of ETF arbitrage. If a stock is largely held within ETFs, then the price of the stock can be driven not only by its intrinsic value but also the supply/demand of ETFs.

Stock Concentration within US Equity ETFs

Three stocks are highlighted in the chart above. Sprint (S) has a market cap of $27.5B, is held by 28 ETFs representing 1% of its market cap, and has a daily turn-over of 0.49%. Amphenol Corp (APH) has a comparable market cap of $25.3B and is held by 55 ETFs.

Half a percent of this stock’s market cap turns over each day and 7% of its market cap is held by ETFs. In other words, it would take 14 days, or roughly 3 trading weeks, to buy/sell the value held by ETFs. The third stock is Western Digital (WDC). It has a larger portion of its market cap held by ETFs but has a higher daily turn-over rate at 2.1%.

If there is a market sell-off, ETFs will likely have far greater trading activity than individual securities. This is for two reasons. First, the $4T in investments held in ETFs.

Second, if a trader is looking to bet as the market sells-off, it is more efficient to short sell an ETF compared to short selling a basket of securities. In this situation, ETFs will have lower values compared to the underlying assets.

Authorized participants will undoubtedly step in. They will buy the ETF, convert the ETF to the underlying assets, and then sell those assets. By purchasing the ETF, this puts upward pressure on the price slowing its decline. Selling the underlying assets puts downward pressure on those prices but in doing so, aligns the ETF with its net asset value.

The systemic risk happens when the authorized participants stop stepping in. If it becomes difficult or impossible to sell the underlying assets, then there is no longer an arbitrage incentive. Liquidity and ETF exposure are key metrics for stocks and other asset classes. If an ETF is comprised of assets commonly held by other ETFs and if those assets have low liquidity, then that ETF could be at increased risk during a sell-off.

I’m not advocating eliminating ETFs from your portfolio. There are many reasons to maintain passive exposure to the markets. Before making any changes, consult your financial advisor. Given the increased discussion on the risks of passive investing, I am merely providing extra insights into navigating this risk.


Created by

Mike Akeroyd

For over the last 10 years, I’ve led the development of ML-based products at Amazon, Disney, and financial services companies. Prior to starting my business career, I served for 5 years as an active duty Army Officer. MBA from UNC, BS from West Point.







Related Articles