ETFs: casting a critical eye

Drew Corbett outlines the rules behind ETF evaluation, and explains that not all ETFs are the same.

Exchange-traded funds (EFTs) are celebrating the 10-year anniversary of their first listing on the Australian Securities exchange (ASX).

Now, more than ever, evidence suggests there is growing interest among the financial planning and advisory community in ETFs as Government reforms and public pressure increase demand for low-cost, transparent and simple investment vehicles.

With some 47 (and growing) ETFs and ETCs (exchange-traded commodities) available to choose from, how can you decide which products to invest in?

One of the most common ways to assess an ETF is to focus on five key criteria, which we call the Five Pillars of ETF evaluation: issuer, structure, liquidity, cost and scale (see figure 1).

Issuer experience counts

The starting point for evaluating any ETF is the product issuer and the strength and depth of the management team.

An experienced team with particular ETF expertise is key. Also look at the experience of the portfolio managers and operations personnel.

It is also important to evaluate corporate governance behind an ETF. For example, is the investment manager authorised by the Australian Securities and Investments Commission (ASIC) as a Responsible Entity? It is also worthwhile to check if each ETF is a separate managed investment scheme.

All this information should be available in the Product Disclosure Statement.

Structure influences results

Once you’ve determined that the issuer is credible and experienced, the next step involves looking at the structure of the ETF.

Currently there are two main structures in the Australian market: direct replication ETFs and swap-enhanced ETFs (sometimes called ‘synthetic’ ETFs).

Direct replication ETFs aim to track the performance of an index by owning all, or a sample of, the underlying index components.

In a direct replication model, when the index changes, the manager of the ETF needs to modify (or rebalance) its portfolio by buying or selling shares to align with the changes.

The costs and potential errors associated with those changes may result in investors experiencing ‘tracking error’ (ie, some variance between the performance of the ETF and the performance of the underlying index).

Swap-enhanced equity ETFs, like direct replication ETFs, provide direct exposure to an index by investing in the underlying index securities.

In addition, they also enter into an agreement with a counterparty (called a swap contract) to ensure the closest possible tracking of the index.

Under this structure, the swap counterparty takes on the risk and cost of matching the difference in the performance of the investment portfolio and the index.

The performance of each ETF (before fees and expenses) is therefore expected to closely track the performance of the index, and tracking error should be minimal. The first swap-enhanced ETF listed on the ASX in December 2010.

Table 1 summarises the differences and similarities between the two ETF structures.

The swap-enhanced ETFs currently available in Australia directly hold the Australian shares that are part of the underlying index being tracked.

Since inception, the value of these shares has never dipped below 99.5 per cent of the net asset value of the fund. The swap agreement (which has not been valued at more than 0.5 per cent of the net asset value of the individual ETF since inception) is used only to account for the difference in performance between these shares and the index.

Both the investment portfolio and the net counterparty exposure are readily available on a daily basis.

Liquidity matters

Liquidity does not only play a role in tight bid/ask spreads but it also affects another factor known as ‘ETF depth’. This can be defined as the volume of the ETF available on the ASX for purchase at any given time.

Looking at figure 2, we can see there is, on average, $2 million of units available for purchase on the ASX of a financial sector ETF (orange line), whereas another product had less than $750,000 available (green line).

A higher depth level of an ETF ensures investors easy access to the product when buying or selling.

Another factor to consider when evaluating liquidity is average daily volume (ADV). This can be analysed by tracking ADV over an extended period of one month or longer to determine the overall turnover of ETF.

But while ADV helps analysis, it is less of an issue for ETFs than in normal shares since ETFs benefit from the liquidity of the underlying stocks of the index they are designed to track.

Look beyond headline costs

Costs to investors of ETFs include management fees, tracking error and bid/ask spreads.

Management fees are familiar to most advisers, and should be an essential part of any cost analysis. Advisers should understand that management fees are not the only cost associated with ETFs.

Tracking error is the difference between the performance of the ETF and the performance of the underlying benchmark index or asset class. This can be a cost to investors if the ETFs consistently underperforms the index being tracked (over and above fees).

Finally, there are costs associated with the ETFs bid/ask spreads.

These costs are those incurred when purchasing and selling ETFs and should be a major consideration for advisers when recommending ETFs.

Spreads can vary significantly between ETFs and advisers should evaluate these costs and, importantly, ensure low consistent spreads prior to investing.

To calculate bid/ask spreads, advisers should look at the difference between the ‘bid’ and the ‘offer’ of the ETF during ASX opening hours.

If, for example, the bid on the ETF was $10.00 and the offer was $10.02, one calculates the bid/ask spread by taking the difference between the bid and the offer (in this case $0.02), and dividing it by the midpoint of the bid and the offer (in this case $10.01). In our hypothetical example, the bid/ask spread is 0.20 per cent (0.02/10.01).

Figure 3 depicts average bid/ask spreads for two ETFs for a few weeks in January 2011.

Over that period, the resources ETF depicted in the red line traded an average spread of 24 basis points, and is consistent across the sample days.

Compare this with another resources ETF over the same period (grey line) which had an average spread of 78 basis points. These are real costs incurred by the investor to buy and sell their ETF on the market, which can be larger than the difference in the management expense ratio (MER) associated with holding the products.

Financial advisers looking to use ETFs should be looking at historical bid/ask spreads in their analysis of ETFs and look for consistently tight and even spreads. This way, advisers can be sure the costs incurred to buy and sell ETFs will be predictable for clients.

A mathematical way of summing up total cost associated of investing in ETFs is: MER + bid/ask spread + tracking error = Total cost of ETF. The lower the number, the cheaper the ETF.

Scale: bigger is better

This brings us to the final point: scale. The evaluation of scale can be achieved as simply as looking at the funds under management of an ETF.

Although there is no hard and fast rule on scale, it’s fair to say the bigger the better when looking at assets under management.

When comparing scale, it is important to compare ETFs that are designed to provide similar outcomes (eg, track the ASX200, provide high dividends or track a sector).

While ETFs are growing in popularity due to their simple, transparent structure, it’s important to realise not all ETFs are the same. These five points provide advisers with a valuable framework when evaluating which ETFs are most robust and cost effective for their clients.

Drew Corbett is head of investment strategy and distribution at BetaShares.