A13: Exchange Traded Funds and Tail Risk

Passive investment products, such as Exchange Trade Funds (ETFs), have been steadily and significantly growing in size for several years, and the rising market share of passive products has attracted a large amount of attention, both from industry practitioners and policy makers as well as academics. One view in this debate is that ETFs are innocuous, since they serve as a low-cost way of investing and that ETF liquidity might increase the liquidity and price discovery of the underlying assets. Contrary to this view, regulators and policy makers have frequently expressed concerns that a high share of passive investments (relative to total market capitalization) means that there are not enough active managers left in the market for price discovery to work and that the market can effectively become illiquid when passive vehicles, such as ETFs, all try to trade in the same direction. ETFs are also relevant in the context of market functioning and market stability because these funds can be traded in secondary markets by both institutional and retail investors. Our project aims at contributing to this key debate on passively intermediated funds by testing for a link between ETF flows and measures of market quality and market risk. A key obstacle is to identify a causal link between ETF flows and market statistics. For example, isolating the effect of ETF ownership on relevant asset pricing moments is not straightforward, because stocks do not randomly enter or exit ETFs. A recent literature uses changes in the composition of stock indices that are tracked by passive ETFs to overcome this identification problem. These constituent changes lead passive index trackers towards (away from) the stock that is newly included (excluded) in a benchmark stock index. We want to build on these shifts in index composition and the associated changes in ETF ownership as an instrument to test for the causal effect of ETF ownership on stock market risk. Specifically, since a major concern in the policy literature on ETFs is that passive flows may increase the probability of large price swings, as per the factors discussed above, we will put particular focus on the tail risk of stock markets in response to ETF flows. Tail risk itself can be inferred from options prices or from the cross-section of stock returns. We also want to bridge the gap to another recent strand of the literature that tries to estimate the impact of flows on market outcomes such as, e.g., prices, liquidity, and volatility of the market, more generally. Gabaix and Koijen (2020) develop a new approach to analyze fluctuations at the level of the aggregate stock market in response to shifts in quantities (flows) based on so-called `Granular Instrumental Variables'. We will build on their work to estimate the causal effect of changes in equity holdings of different types of intermediaries (e.g., passive versus active) on key market statistics such as prices, liquidity and volatility.

Project Leader: Prof. Dr. Maik Schmeling and Prof. Dr. Christian Schlag