Aussie Lessons About Nontransparent ETFs

John Hyland, CFA, is a retired ETF executive and longtime investment industry professional who has contributed articles to ETF.com in the past.

In the very near future, “active, non-fully transparent ETFs” will begin to be offered to U.S. ETF investors. They are designed to allow a fund manager to offer active management with an ETF without having to release their daily portfolio holdings.

Active portfolio managers have been reluctant to offer fully transparent active ETFs out of fears that their trading could be front-run by other market participants.

This helps explain why many mutual fund companies have steered away from getting into the ETF space despite the fact that the ETF wrapper is generally both cheaper and more tax efficient for their investors than the traditional mutual fund wrapper.

Active Floodgates

Given that there is more than $10 trillion in active U.S. mutual funds, the ability for mutual fund companies to now offer the same active portfolio, but in a potentially “better” wrapper, could quickly lead to tens or hundreds of billions in fund assets to migrate to the ETF side within a few years. This could turn out to be a huge story.

As an investor or financial advisor with an interest in these new funds, it is easy to appreciate the value of their being more tax efficient, and perhaps cheaper, than a mutual fund.

However, buying any ETF requires an investor to face the cost of trading represented by the bid/ask spread. If you buy a passive ETF in the U.S., the bid/ask spread typically runs from 1 or 2 basis points wide for large, liquid ETFs to 15 basis points or more for the average ETF. For the U.S. ETFs that are active, but still fully transparent, their spreads also run from 1 or 2 basis points to an average of 20 basis points or more. But at what spread will these new style ETFs trade?

What Will Bid/Ask Spreads Be?

The mutual fund companies getting ready to launch nontransparent ETFs are not likely to publicly venture a guess about bid/ask spreads. In addition, ETF market makers—who likely have already war-gamed the process and know exactly what the spreads are likely to be—are a tight-lipped group.

‘Down Under’ Test Case, Perhaps

Fortunately, the U.S. is not the first established ETF market to go the active, nonfully transparent ETF route. The Australian ETF market introduced active, nonfully transparent funds in 2012.

While there are some differences between the U.S. and Australian ETF markets, the Australian experience can help still teach us something about what to expect.

First, a few caveats. The U.S. has more than 2,300 ETFs, several trillion dollars in assets and trades extremely cheaply.

The Australian ETF market has a little more than 200 ETFs and less than $100 billion in assets. In addition, almost half of their ETFs focus on Australian equities or fixed income. Finally, the average size of an Australian ETF skews fairly small.

Spreads Wider

Not surprisingly, most ETFs in Australia tend to trade with much wider bid/ask spreads than their U.S. peers.

On the other hand, the ETF market makers in Australia are almost the exact same ones who make markets in the U.S. So, by making some adjustments, we may still learn some useful things here.

As the table below shows, active, nontransparent ETFs in Australia are found in most ETF categories, although primarily they are found in three main categories (as defined by the Australian Stock Exchange, “ASX”).

Equity & Fixed Income Focus

For this next look, we are going to focus just on the equity and fixed income ETFs and ignore cash ETFs, currency ETFs and commodity ETFs.

A key takeaway here is not to focus on the absolute spreads of the Australian active, nonfully transparent ETFs. Remember that Australian ETFs in general have much wider spreads than U.S. ETFs, so focus on the difference between the passive ETF and the active ETF.

In at least one area, fixed income, there is actually only a modest difference between the passive and the active spreads. This is likely because even if an ETF market maker does not know exactly what a bond ETF’s portfolio is today, they certainly knew what was in it the last time the ETF released its quarterly statement 60-90 days ago.

And since bond portfolios tend to change only slowly, it is still pretty easy for a market maker to know what to use to hedge their market-making risk.

Drilling Down On Equities

Now let’s take another look at just the equity ETFs. Here we are comparing global equity ETFs versus domestic Australian equity ETFs (we removed all of the “Mixed Asset” ETFs, as they hold a mixture of Australian and global investments).

In addition, we are going to split out the Australian large cap segment. There are no active ETFs reported in that segment, which will allow us to better compare domestic active ETFs to their passive peers.

The results are interesting. On a more apple-to-apples basis, the difference between most domestic passive and active ETFs is only about 11 basis points.

To put this into dollars and cents, if you had an ETF with a $25 net asset value (NAV), the passive domestic large cap ETF would be trading about 2 cents wide, the other passive domestic ETFs would be trading about 6 cents wide, and the domestic active ETFs would be trading about 8-9 cents wide.

For the global equity products, the gap is wider: 16 basis points. The lower passive global equity number is helped by the presence of several large S&P 500 listings. As a result, the gap—once again, assuming a $25 NAV—is between a spread of 4-5 cents for passive versus 8-9 cents for active.

‘Sticky Spreads’

Analyzing the data from 2012 to the present also produced two other interesting observations that may be seen in the U.S.

First, spreads on active, nonfully transparent ETFs have historically proved “sticky” on the ASX; that is, whatever the spread was the first month or two after listing years ago—12 basis points or 30 basis points—is likely still about the same spread today. There are some exceptions, but generally, time itself does not led to spreads changing.

Second, in general increases in assets under management (AUM) in an active ETF seem to have, at best, only a minor impact on spreads over time. An ETF that started with $10 million in AUM that now has, say $100 million, likely still has the same range of spreads. Once again there are exceptions

Key Takeaways

Getting out my crystal ball, what do I think this all means?

Considering that the U.S. markets tend to trade more cheaply than the Australian markets, I would expect U.S. active, non-fully transparent ETFs that own U.S. fixed income to trade only modestly more expensively than similar passive, U.S. fixed income ETFs.

For U.S. equity, I would not be surprised if nonfully transparent ETFs that track active portfolios with fairly low turnover—say, annualized turnover of 40% or less (which equates to only 10% a quarter)—might only trade a few pennies wider than their fully transparent peers.

Keep in mind that, based on the Australian experience, neither time in market nor an increase in one of these ETF’s AUM is automatically guaranteed to reduce trading spreads. It may be that the spread you see after the first few weeks is the spread you should assume will be there when you look again a year later.

A new chapter of the ETF industry evolution is starting. It will make 2020 an interesting year.

John Hyland can be reached at [email protected]