Commonality in the Mispricing of ETFs

M. S. Broman released his research in 2013 wherein he used three sets of tests that verified the existence of commonality in the mispricing of U.S. exchange-traded funds. 

He began his first test by investigating the existence of premiums, and confirmed that it is the ETF that is mispriced rather than its net asset value (NAV). Accordingly, the premiums predict ETF returns and not the returns of NAVs. 

He also found in his second test evidence of persistence in mispricing following a period of negative average mispricing. He found a high correlation between the changes in ETF mispricing and the average change in mispricing of other ETFs in the cross-section. 

Under his third test the author found that mispricing was systematically affected by local risk. He found these excess exposures to be amplified in periods when the average international fund is highly mispriced, an indication of times when arbitrage is limited. More specifically, a one standard deviation increase in the absolute level of mispricing of international ETFs is associated with an increase in U.S. factor loadings by 6.14% and a decrease in EAFE factor loadings by 5.25% in the following week. 

Once again as in previous literature this author found trading locations to be important to mispricing. The clientele for international ETFs are different than the clientele for U.S. ETFs. In the creation and redemption trading used for ETFs the mispricing is usually arbitraged away in short order. 

However, in the international arena the mispricing lingered persistently. Investors in international ETFs benefit by heeding these findings on mispricing in the two markets.

sources: http://ssrn.com/abstract=2227004.