How Do ETFs Affect Stock Volatility? [SSRN]

Abstract: This paper offers evidence supporting that ETFs shield underlying from demand shocks. Because ETFs are designed with secondary markets, ETF trading can substitute trading in underlying, and thus absorb demand shocks that would otherwise move underlying prices. I find U.S. common stocks that would have been more heavily traded without equity ETFs display lower volatility from 2011 to 2021, consistent with ETF secondary market trading providing liquidity buffer to underlying. Further, I show this liquidity provision channel complements the demand propagation channel established in the literature by differentiating ETF trading from ETF ownership. The effects from both channels are significant even after restricting volatility changes to these exogenously driven by the NASDAQ 100 index reconstitutions.

This figure illustrates the mechanisms how ETFs could operate similar to mutual funds or index futures. ETF authorized participants can create (redeem) fund units and purchase (dispose) underlying assets - called the arbitrage mechanism but akin to the creation and redemption mechanism in mutual funds. Via this channel ETFs propagate demand shocks from investors to underlying. ETF marker makers can also match investor buy (sell) orders to sell (buy) orders, identical to market makers in index futures. In this channel ETFs shield underlying from investor demand shocks.

Optimized Index Replication and the Cross-section of Trading Volume: Evidence from Germany and the U.S.

joint with Erik Theissen

Abstract: Investors wishing to trade a basket of stocks (e.g. the DAX or the S&P500) can either exactly replicate the basket, or construct a replicating portfolio consisting of only a subset of the basket securities. The latter strategy has higher tracking error but lower transaction costs. We construct such replicating portfolios using algorithms exploiting return correlations, and using heuristics based on market capitalization or bid-ask spreads. In a German (DAX) and a US (S&P500) sample we find that stocks included in the optimized replicating portfolios have higher trading volume than other index constituents. These results suggest that such replicating portfolios are used in practice, and they add to our understanding of the cross-sectional determinants of trading volume.

Page updated

Google Sites

Report abuse