Beware of taking labels at face value

Sebastian Ceria and Melissa Brown warn that exchange-traded funds with similar labels can generate widely different returns because of the way their portfolios are constructed 

At a glance

• Although many exchange-traded funds (ETFs) have similar names they are unlikely to deliver similar returns.
• There were significant differences in the performance of US high-dividend funds in the wake of the country’s presidential election last year.
• The big drop in the oil price also affected funds with similar names differently.

Smart beta has had a foothold in the foreign exchange market for years, although the approach is less pervasive than those of other asset types and it rarely gets called by that name. There are certainly arguments for the presence of risk factors, most notably the currency carry, which investors have long sought to exploit.

What’s in a name? For smart beta exchange-traded funds (ETFs), let the buyer beware, as funds of the same name may not generate returns as sweet.

ETF fund families rely on common themes to create single-factor smart beta ETFs. The objective is to capture systematic returns using a rules-based approach that is transparent and replicable. Yet, while many ETFs have similar labels, such as ‘US Large-Cap Value’ or ‘High Dividend Yield’ they are unlikely to deliver similar returns.

Why? Portfolio construction is often the culprit. Portfolio-construction rules and methods vary widely from one ETF to the next, and those variations can significantly affect performance. Variables include:

• The universe of stocks from which the provider chooses.
• The scheme for weighting assets – market capitalisation, equally weighted, or some other method.
• The number of names allowed in the portfolio.
• Whether other factors (such as sector exposures) are constrained.

The application of these rules determines the portfolio’s contents which, in turn, drive the portfolio’s returns. Portfolio construction matters, and investors should be aware of the key return drivers when comparing smart beta portfolios, similarly named or not.

In this article, we illustrate the effects of two sample market moves on a number of similar-sounding smart beta ETFs. The impact of portfolio construction on a portfolio’s exposures is clear: the performance of such portfolios can and did vary substantially. While we focus here on US high dividend yield portfolios, we have reached similar conclusions with other ETF types as well.

By examining data from big, impactful events – such as the US presidential election and the sharp decline in oil prices in 2014 – we show how several dividend-yield smart beta funds reacted differently to those moves because of unintended consequences of portfolio construction.

In these particular cases, market moves were driven by factors – such as sector returns, beta and size exposure – and since most smart beta funds do not control exposures to these factors, big moves in these factors significantly affected the performance of the funds.

The US election and its aftermath: one of the effects of the US election was a sudden and dramatic sorting of the expected sector winners and losers. In the three months to 8 February 2017, US financial stocks soared about 18%, whereas utilities and consumer staples stocks gained just 2%. Technology stocks, up 8.5%, were in line with the benchmark. Clearly, sector allocation had a big impact on returns during that three-month period, and that impact would have swamped the benefits of investing in a particular factor.

To illustrate the impact of portfolio construction, we chose five funds that have ‘dividend’ or ‘dividend yield’ in their names but differed in a number of ways:

• SDY: S&P Dividend ETF
• VYM: Vanguard High Dividend Yield ETF
• FVD: First Trust Value Line Dividend Index Fund
• HDV: iShares Core High Dividend

The impact of portfolio construction stands out clearly, mainly in HDV’s underperformance relative to the others. (They all lagged the market.) HDV was extremely underweight financial stocks, and that was the main source of underperformance. Other factors also contributed to the underperformance, including its large-cap bias. Without detailed analysis, it is impossible for an investor to differentiate this fund from other similar-sounding funds. And would an extreme underweight in financial stocks even be considered acceptable?

To be sure, all of the portfolios shared certain characteristics, such as overweight positions in utilities and consumer staples, underweights in information technology, and low beta – all of which led them to lag the market over this period. But the magnitudes of those exposures, and their impact on performance, also differed. In addition, the excess return from their (positive) exposures to dividend yield was muted, and could not offset the large effects from other factors.

The oil price drop had important implications for energy stocks. We chose the same set of portfolios to provide another example of the impact of portfolio construction. Oil prices halved from their peak in June 2014 to December of that year, so exposure to the energy sector would have likely driven big differences in returns. Two of our sample portfolios, VYM and HDV, had energy weights that well exceeded the 8.7% exposure in the broader market, whereas the other two were quite underweight.

During the second half of 2014, the broad US market rose about 4.5%. SDY and FVD managed to far outperform during that period, whereas HDV lagged substantially. VYM was slightly ahead of the market. Sector weight decisions clearly helped SDY and FDV and hurt the other two – again, despite the similar labels.  

SDY’s big underweight in energy offset the drag from overweights in others. FVD was also helped by its energy underweight, but returns were boosted more by the ETF’s overweight in utilities. In contrast, energy overweights for both HDV and VYM hurt returns, and those shortfalls were not offset by better selection of other sectors. All four portfolios had low betas, and that low beta was an important source of return as the market was driven higher by lower beta stocks.

In addition, having high dividend yields was a drag on returns for all of our test ETFs over this period. As in our US election example, this exercise demonstrates how an ETF’s exposures, which may or may not be intentional, are likely to drive big enough differences in contribution to return that two similar-sounding funds may end up with vastly different performance. 

Sebastian Ceria is the CEO and founder of Axioma. Melissa Brown is the firm’s managing director, applied research 

Have your say

You must sign in to make a comment


Your first step in manager selection...

IPE Quest is a manager search facility that connects institutional investors and asset managers.

  • QN-2311

    Asset class: Cash.
    Asset region: OECD.
    Size: EUR 20-200 m.
    Closing date: 2017-05-26.

  • QN-2315

    Asset class: All/Large Cap Equities.
    Asset region: Global Emerging Markets.
    Size: $150m.
    Closing date: 2017-05-29.

  • QN-2316

    Asset class: All/Large Cap Equities.
    Asset region: Global Emerging Markets.
    Size: EUR 200m.
    Closing date: 2017-05-31.

  • QN-2317

    Asset class: All/Large Cap Equities.
    Asset region: Global Emerging Markets.
    Size: EUR 200m.
    Closing date: 2017-05-31.

  • QN-2321

    Asset class: All/Large Cap Equities.
    Asset region: Switzerland.
    Size: CHF 150m.
    Closing date: 2017-06-05.

Begin Your Search Here