Can the new ETFs launched in the past seven years weather a market crisis? Is there is a growing number of under-capitalized ETFs that could destabilize the fund industry. Especially if a quarter of those may be over-exposed to derivatives risk?
Of course, reducing ETF offerings is difficult without penalizing small and innovative firms. We’ll leave that subject for another day.
In the above chart, the blue bars represent number of new ETFs launched in each quarter. As you would expect, new ETF launches grew, in fits and starts, until 2008, then took a dive, then regained previous levels and flattened. Around 2014 things became interesting. Almost every quarter saw about 50 ETFs launched.
The thin orange line is the average assets for each quarter’s ETFs. As you would expect, some ETFs groups that were launched more than seven years ago have significant assets. But it’s alarming that since 2013, recent ETF groups have average assets below $15 million.
Worse, 12-month average assets of new ETFs has been declining since 2013. At best, this shows a saturating market. At worst is shows a credit-fueled speculation in new ETFs (assuming managers are using credit to finance them).
Here’s annual ETFs launch and closure data.
Since 2008 there have been around one thousand ETF fund closures, mostly ETFs in established firms with a profitable ETF fund business.
Until 2011 no more than 60 ETFs closed each year. Since 2012, ETFs closures have been rising yearly, from 69 to 154 closures in 2018. If the trend remains, by 2025 ETF closures might run over 200 per year.
ETF Minimum Expenses
All ETFs must pay custodian, registration, auditing, legal, printing and other fees. Understand that many of them don’t break out these fees; they’re often included in one unitary management fee. Fortunately, there are many ETFs that report operational expenses for their shareholders. We can use these funds to estimate their minimum operational costs.
I selected 189, open-end type, ETFs that broke out their fees in 2017 in form N-SAR (which the S.E.C. has replaced with an abbreviated form N-CEN*). My quick estimate is that most ETFs incur at least $85,000 a year on average in operational expenses, depending on characteristics such as whether they hold actual assets or synthetics, the type of asset (fixed income, global equities etc.) and whether the fund is part of a series trust or stand-alone.
Often, these expenses are reimbursed by the advisor. Again, what is being internally reimbursed, by whom and why, across the industry, is impossible to ascertain, especially for funds that don’t itemize their expenses.
ETF Going-Concern Risk
There are two main risks for new ETFs. Assets lost if counter-parties default for leveraged ETFs (which make up 25% of new ETFs). Or portfolio losses experienced by funds desperate to liquidate their portfolios because their managers can’t meet the fund’s operating expenses
Let’s factor in management fees.
I’ll posit that an ETF can’t realistically be run for less than $250,000. (Outside exchange fees, lawyers, administrators, auditors, etc–stuff no fund manager could do themselves — as I estimate above, come to $85,000 alone).
In today’s market of 2,300 ETFs, almost 1,000 collect less then $250,000 in fees. Of course, large managers benefit from economies of scale. But even for them, I can’t see any fund collecting less than $165,000 (after the core $85K) adding to the bottom line. Again, there’s no way for any manager to escape exchange fees, outside counsel, audit fees, etc. There are 700 ETFs in that profitability-challenged group.
Looking at the smaller firms running ETFs, there are around 100 ETFs that are part of firms with less than $500,000 in total ETF expense revenues.
Here are some very small ETF families, to give you an idea what I mean by “sub-scale”.
Even for firms under $5 billion in ETF revenues, there exist 50 to 100 ETF providers who would find it difficult to remain in the ETF fund business if there was a multi-year bear market. Although some of them have diverse revenue streams, it’s hard to see how their ETF family would help their private wealth management business, for example.
It seems the ETF industry reached its saturation point sometime after 2014. Since then ETF launch success has been following the general trend in securities asset-valuation growth, not new money entering the markets. That is, easy money is inflating the number of ETFs launched along with stock prices. That’s highly debatable of course.
There is an interesting trend in recent ETF launch types. I calculate 1% of new ETFs in assets, around $50 billion, are dedicated to betting some aspect of the market. In number of ETF funds, however, they make up 18% of the market. The percentage is expanding. In the past 4 quarters 25% of all ETF launches have been to help investors bet, or hedge against, some kind of market movement.
These are funds highly sensitive to stock, bond and interest-rate spikes.
On the other hands, the Market Bets space remains small. But keep in mind their leverage implies a much greater amount of assets at risk.
Is it possible that in a harsh market downturn many ETFs would be forced into messy liquidations? If so, how many?
Distribution of ETFs by Expense Health
613 Exchange Traded Products (ETFs/ETNs) that collect less than $85,000 in fees, potentially fall into this ‘sub scale’ category.
Assuming that any ETF with above $100 million has a good chance of being, or becoming, viable, I’ve filtered the dataset to 1,237 ETF under that amount.
Next, let’s calculate how much these ETFs would collect in fees if they had a 30% reduction in AUM.
How can an investor trust that their fund is being managed with due care if the numbers aren’t available? How can anyone determine how much money is coming from revenues, and how much from other sources? Are these funds only hoping to gain enough assets to be profitable or are there hidden fees that make some of them profitable already?
If the market had a 30% correction, it looks like 717 ETPs would be losing money. That’s another 100 funds. A 50% correction would add another 100 on top of that.
I’m not predicting anything. I’m only pointing out data suggesting that a serious market crash could wipe out a third of all ETFs because many of them seem to be already unprofitable.
Max @ maxdatabook.com
Please note, this is a cursory study. There may be inaccuracies, not small. If you are interested in a deeper analysis please contact me through e-mail and we can set up a time to talk.
*Yes, there is data in N-CEN that wasn’t in the N-SARs; however, the fact remains that data that was once available in a standardized questionnaire is no longer.