Credit: Shutterstock photoIt is well established that smart beta ETFs have continued to grow in assets and listings annually. Previously we had estimated the global smart beta ETF market to be $350B, or over 10% of global ETF assets at the time.
This rapid adoption prompted the creation of a smart beta definitions framework, since traditional 3x3 style boxes don’t adequately serve investors’ needs. The First Bridge smart beta framework is prospectus based i.e. if an ETF’s prospectus outlines an intended ‘value’ strategy, the ETF will be classified as ‘value’.
A well-designed prospectus based system has both advantages and disadvantages. The advantage is that ETFs that intend to provide exposure to the same investment factor can be found easily. In the First Bridge smart beta system, investors interested in low volatility strategies would find both SPLV and USMV, even though their prospectuses may have marginally differing investment objective language.
However, even a well-designed prospectus system has limitations. To further evaluate smart beta ETFs, we need to look at weight differences in the ETF constituent holdings. One way to do this is to examine the ‘active share’ of each ETF relative to a market cap weighted ETF.
SPLV vs. USMV using ‘Active Share’
It has been argued that smart beta ETFs, despite being index linked, are effectively active strategies. In 2009, Cremers and Pestajito proposed ‘active share’ as a metric to measure the extent of active management in a fund. Active share measures how the constituent holdings of a fund differ from a benchmark index like the S&P 500.
Here we adapt the concept of active share to evaluating smart beta ETFs. Rather than comparing a fund to the theoretical S&P 500 index, we measure active share relative to SPY, since it is the most widely used ETF for S&P 500 exposure.
Though SPLV and USMV both intend to provide exposure to low or minimum volatility stocks, they have had very differing returns (see Table 1). Both have underperformed SPY, but USMV has been closer to SPY in terms of its performance.
Table 1: Active Share & Returns of SPLV & USMV vs. SPY
* Each share class counted as a separate holding; Data:www.firstbridgedata.com; as of July 30, 2018
Their very different active share scores can partially explain this. Even though SPLV is a subset of SPY, it holds only 100 stocks and these stocks have very different weights from the parent SPY with no sector caps being applied. USMV on the other hand, is designed to have less sector concentration and to track more closely with the broader equity market by holding over 210 stocks. Consequently SPLV is effectively a more ‘active’ investment than USMV.
SPLV vs. Other Smart Beta ETFs using ‘Active Share’
We can extend this to other ETFs that track different investment factors. The chart below plots active share score (x-axis) vs. number of holdings (y-axis) for some smart beta ETFs.
- The Invesco S&P 500 Equal Weight ETF (RSP) holds the same stocks as the market cap weighted SPY. It is equal weighted, and has an active score of 46% relative to SPY.
- The iShares Select Dividend ETF (DVY), like SPLV, holds only 103 stocks and is quite active (84%) compared to SPY.
- Similarly the Invesco S&P 500 Pure Value ETF (RPV) also holds only 110 stocks and has an active share above 80% relative to SPY. This is because the underlying index methodology both screens and weights stocks based on style scores.
Using active score (relative to the relevant cap weighted benchmark) in combination with a prospectus based smart beta classification system can be an effective way to get further insight into smart beta ETFs.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
The views and opinions expressed herein are the views and opinions of the author and do not necessarily reflect those of Nasdaq, Inc.
