New Developments in Long-Term Asset Management

Second Annual Conference
May 19-20, 2017

ETF Arbitrage under Liquidity Mismatch

As one of the fastest-growing asset classes, exchange-traded funds (ETFs) have been increasingly invested in illiquid assets. A natural liquidity mismatch emerges when liquid ETFs hold relatively illiquid assets. Kevin Pan of Harvard University and Yao Zeng of the University of Washington theorize and present empirical evidence, in the particular context of corporate bond ETFs, that this liquidity mismatch can reduce market efficiency and increase fragility. They leverage two key institutional frictions: the liquidity mismatch, and the dual role of authorized participants (APs) as both bond dealers and the only arbitragers able to create and redeem ETF shares.  [Download the paper]

Other Conference Papers

Institutional Investors and Information Acquisition, Matthijs Breugem and Adrian Buss

The Relevance of Broker Networks for Information Diffusion in the Stock Market, Marco Di Maggio, Francesco A. Franzoni, Amir Kermani, and Carlo Sommavilla

Asset Insulators, Gabriel Chodorow-Reich, Andra C. Ghent, and Valentin Haddad

Do Institutional Incentives Distort Asset Prices? Anton Lines

Chasing Private Information, Marcin T. Kacperczyk and Emiliano Pagnotta

Replicating Private Equity, Erik Stafford

Efficiently Inefficient Markets for Assets and Asset Management, Nicolae Gârleanu and Lasse Heje Pedersen

< 2016 Conference Papers>

Pan and Zeng propose a theoretical framework showing that APs' ETF creation and redemption motives are influenced by two opposing effects: an arbitrage effect and an inventory management effect. When the absolute magnitude of inventory imbalances of APs (also as bond dealers) is small, APs use ETF creations and redemptions to close bond-ETF relative mispricings. APs are willing to close relative mispricings when the marginal benefit of expected arbitrage return outweighs other costs. However, ETF arbitrage remains far from frictionless and is limited by the liquidity mismatch. In contrast, when the absolute magnitude of APs' bond imbalances is large, the inventory management effect, which captures the motive to trade towards an optimal bond inventory level, rises. While APs still create and redeem ETF shares, ETF arbitrage may go in the opposite direction to what would be implied by the initial relative mispricing. Intuitively, APs may strategically use ETF creations and redemptions not to correct relative mispricings, but to unwind bond imbalances, reduce existing inventory risks and facilitate future market-making in their role as bond dealers. In this sense, the ETF arbitrage mechanism becomes "distorted" — creations and redemptions are disconnected from fundamentals — and gives rise to the possibility of larger relative mispricings.

Pan and Zeng develop empirical evidence by leveraging unique datasets that capture APs' ETF transactions and corporate bond trading. They obtain lists of APs for each corporate bond ETF of two ETF issuers, and obtain a proprietary version of all secondary market corporate bond transactions with dealer identifiers, allowing them to directly link secondary market bond transactions of APs. Using these identifiers, they impute bond order flow imbalances on APs' balance sheets and identify their impact on APs' arbitrage activity. Finally, they obtain time-series data on the daily create and redeem baskets for each ETF in their sample to identify the subset of bonds used in the ETF arbitrage. They use these datasets in their empirical design to trace the effect of initial bond pricing and inventory through to realized AP arbitrage.

The researchers' findings are consistent with the model's predictions. First, increases in market volatility and bond market illiquidity reduce ETF creations and redemptions, suggesting the existence of limits to arbitrage under liquidity mismatch. An increase of 1 percent in the ETF premium generates an increase in AP arbitrage of 50 basis points. However, as market volatility rises, holding fixed the ETF premium, a one standard deviation increase in market volatility generates a 10 percent decline in AP arbitrage. Similar results hold when bond illiquidity increases.


Second, bond inventory generates unique risks to ETF arbitrage through the inventory management effect. When APs experience larger bond inventory imbalances, APs' creation and redemption activities become less sensitive to perceived arbitrage opportunities, and more sensitive to the size of their bond inventory imbalances.

Third, realized arbitrage by APs affects relative mispricings across markets as well as liquidity in the underlying corporate bond market. AP creation and redemption activities reduce the persistence of relative mispricings and increase liquidity in the corresponding arbitrage baskets. However, these effects may become less pronounced or even reversed when APs' arbitrage capacity becomes more constrained due to either greater liquidity mismatch or AP bond inventory imbalances.

Overall, these findings suggest a new type of risks that is worth further attention as ETFs continue to grow and invest in illiquid assets.

NBER Videos

National Bureau of Economic Research, 1050 Massachusetts Ave., Cambridge, MA 02138; 617-868-3900; email:

Contact Us