The exchange-traded product (ETP) market is becoming increasingly crowded with more products and providers surfacing all the time. The industry’s heady growth and increasing complexity has left many investors perplexed. Here I will provide investors with the top three criteria I identified when comparing ETPs; namely, knowing the index, understanding the ETP's structure and assessing the total cost of ownership. This is by no means an exhaustive listed as there are many points to consider when evaluating ETPs, but it is a good starting point.
First, investors need to evaluate which index can best deliver their investment objective. As every index is different, investors need to understand the mechanics of the index being tracked to avoid undesired surprises.
The risk of excessive concentration in individual stocks or in specific sectors is fairly easy to spot. Some index providers implement caps on the weightings at the individual constituent level and broader sector level to avoid such concentration. The oil giant Petrobras (PBR), for instance, represents almost 20% of the value of the MSCI Brazil Index. BP's (BP.) disaster in 2010 is a prime example for the potential dangers of such concentration risk.
It is also important to understand the index’s weighting methodology. Most indices are constructed using the market-capitalisation-weighting method. This can cause some problems during bubbles, e.g. the tech-bubble in the late 1990s, as it can lead to excessive concentration in overvalued stocks. Implementing this methodology for fixed-income indices can introduce similar risks as the most heavily indebted issuers will be most heavily weighted in market capitalisation weighted indices. Concentrated exposure to the largest corporate and sovereign debtors carries the obvious risk that these issuers might eventually leverage themselves to the point where they are unable to service their debt.
Besides understanding the index's construction, it is also very important to understand the different structures used by providers to deliver the performance of the reference index.
For exchange-traded funds (ETFs), physical replication can be achieved either through full replication or optimised sampling. The optimised sampling approach is more common for indices with a large number of components or with less liquid constituents, in which case the provider would only buy a basket of selected component stocks reflecting the same risk-return characteristics as the underlying index. This model works well during normal market conditions. However, in an adverse market environment historical statistical attributes are thrown out of the window, and sampling can result in higher than expected tracking error. In addition, some physically replicated ETFs make use of securities lending to generate an extra income. This will expose investors to counterparty risk. However, the level of risk tied to securities lending will depend on the level of securities lent by the fund.
Providers of synthetic replication ETFs deliver the performance of their funds’ benchmarks via a swap agreement with a single or multiple counterparties. The provider agrees to pay the return of a pre-defined basket of securities to the swap provider in exchange for the index return. Synthetic replication generally reduces costs and tracking error, but increases counterparty risk. For difficult-to-access markets, swap structures have demonstrated superior tracking relative to physical replication funds. Investors interested in a more in-depth analysis about the different structures can read an article we published last year: The Difference Between an ETF and an ETP.
Last but not least, investors need to assess the total cost of ETP ownership which can extend well beyond the headline expense ratio.
The total expense ratio (TER), representing the portion that will be deducted from the funds’ net asset value (NAV) to pay for the manager’s effort, is the most explicit cost of owning an ETP. However, investors should not neglect trading costs which will come on top of the TER. These costs include brokerage commissions and bid-offer spreads. Furthermore, premiums and discounts to NAV represent another important consideration.
Tracking difference is an implicit cost of owning an ETP and represents the difference between the ETP’s return that of its benchmark. The biggest part of the difference can be attributed to the TER. It is also important to distinguish between the tracking difference and the tracking error, which measures the short-term volatility of an ETP's return relative to its benchmark.
As mentioned, one source of the tracking error can stem from the replication method as a result of the index turnover. Realigning the portfolio of a physical replicated fund due to mergers, bankruptcy or acquisitions can change the index’s composition and force rebalancing trades of the ETP.
Moreover, the timing of dividends can have an impact on the tracking difference. For physical ETFs, cash received as dividends from the underlying stocks is held in the fund’s income account until it is distributed to fund holders. This dividend treatment can potentially create a drag on returns in upward trending markets as dividends are not immediately reinvested into the fund. In practice this cuts both ways. It could also result in outperformance if the benchmark falls in this interim period.
A more detailed analysis on the total cost of ownership can be found in our article The Total Cost of ETF Ownership.
We hope this brief guide serves as a useful starting point for investors comparing different ETPs.
