At FactSet ETF Analytics, we measure tracking difference over a rolling series of 12-month windows. We wind up with 252 12-month return gap values, one for each trading day in the calendar year. From that set, we cull the median, and the two biggest outliers, one on the upside and one on the downside. The median gives a pretty good picture of the fund’s expected behavior, while the outliers show us the best/worst-case scenarios and define the range of investors’ experience.
*Bond funds often appear to have wide tracking ranges, because of the variability in bond pricing across the industry.
For a larger view, please click on the image above.
Optimization Vs. Replication
Adjusting for the new, lower expense ratios, we can expect the revamped SPDR Portfolio funds to offer the best median tracking difference compared to their direct competitor tracking the Russell 1000/2000/3000 suite and the S&P 500 growth and value indexes. Yet we can expect Vanguard to continue dominating in bond funds tracking the Bloomberg Barclays Aggregate Bond and 10+ year Corporate Bond indexes. (Note that Vanguard’s BND tracks a float-adjusted version of the Agg.)
But things fall apart for the revamped SPDR funds when it comes to the tracking difference range. These funds have historically delivered excess variability in tracking, especially compared to their competitors. This will still be the case even with the new expense ratios unless SSGA changes its portfolio construction methodology.
In markets where liquidity can be tight, SSGA often optimizes portfolios, sometimes quite heavily. Optimization means that the portfolio holds some, but not all, of the securities in its underlying index. It’s the opposite of full replication. Optimization can lead to tracking error when the portfolio securities don’t fully reflect market activity.