Following is courtesy of insight from Chris Hempstead, the head of ETF Trading for agency broker WallachBeth Capital LLC.
With respect to analyzing and selecting ETFs, one of the most common and frustrating mistakes that I overhear is "..unless the fund has at least some minimum AUM ($50mm in many cases)," or has average daily trading volume less than X (some arbitrary number -- say 250k shares) it should be avoided…"
Some other unfounded arguments against ETFs go so far as to suggest that "...ETFs need to have a certain history or track record before they should be considered…" Adding insult to injury is the claim that "investors are at risk of losing all their money if an ETF shuts down."
In light of recent articles being picked up by media from New York to Seattle, I would like to dismiss a few of these common unwarranted reasons to avoid an ETF based solely on AUM, ADV or track record.
First, let's address AUM:
"ETFs with less than $50mm should be avoided"
In order for an ETF to come to market (list on an exchange), the fund needs to have shares created. This process is often referred to as seeding. The "seeder" is the initial investor who delivers into the custodial bank the assets required to back the initial tradable shares of the ETF in the secondary market. ETFs issue shares in what are known as creation units. The vast majority of ETFs have creation unit sizes of 25k, 50k or 100k shares.
When a "new" fund comes to market, they are usually seeded with at least 2 units of the fund. There are very few examples of ETFs that come to market with more than $5mm in AUM or an excess of 200k shares outstanding. (One recent exception comes to mind: the Pimco Total Return ETF (NYSEARCA:BOND) launched with ~$100mm AUM and 1mm shares outstanding)
Understanding that ETFs have to start somewhere, it would be difficult to explain how more than 55% of ETFs (excluding ETNs, Leveraged ETFs and Inverse ETFs) have garnered AUM in excess of $50mm.
In other words, someone had to take a close look and invest into the funds. The "I will if you will" mentality is probably not how the most successful fund managers find ways to outperform.
Ten of the top 30 performing ETFs year to date have AUM below $50mm.
(EGPT, TAO, HGEM, IFAS, SOYB, PKB, RTL, QQQV, ROOF and RWW)
Congratulations to the pioneers who "went it alone," as they say.
Now, let's address Average Daily Volume:
Why not start with another top 30 stat? Twenty of the top 30 performing ETFs year to date have ADV below 250k shares.
This is the one that really gets my attention, mainly because ETF ADV is not a true measure of the liquidity of the underlying assets and by default, the actual ETF.
In its most basic form, simply looking at ADV and dismissing a fund based solely on that metric might steer you away from funds that are quoted in block size at the minimal increment (one penny).
Here is an actual ETF example:
The First Trust REIT ETF (NYSEARCA:FRI), has an ADV of roughly 100k shares per day. The quote in this name however, is also roughly a penny wide, with 75k shares on each side of the market. This means you can trade roughly 75% of the ADV in a single trade with little or no impact to the quote.
This kind of liquidity (as a % of ADV) is rare in single stock names. Unfortunately, I suspect that the aversion to trading ETFs with low ADV stems from the realities that exist in single stock trading which, again, is just not true for ETFs.
Recently, there has been regular chatter about a metric called "implied liquidity." This is an important step in educating investors as to why they can effectively trade more than just a small percentage of an ETF's already low ADV. That being said, this metric has many variations and assumptions, which need to be considered when ascertaining whether or not the desired investment size can be efficiently executed in a particular ETF.
One other reality in trading low-ADV ETFs that every investor should keep in mind is that because an ETF has less volume and in many cases is less widely held, you are likely to be trading against a professional market maker.
Traders make markets in ETFs based on the fund's real time NAV (aka INAV), not last sale. Last sale is one of the most irrelevant metrics in determining the "right" price to enter an ETF order.
Knowing how and where liquidity might exist based on a fund's NAV is the most important tool in efficiently sourcing block liquidity in low-ADV names at minimal premiums/discounts to the ETF's NAV.
Regarding Track Record:
The argument for avoiding ETFs with little or no track record carries with it a tad bit more merit than the aforementioned. Obviously, one cannot predict future performance, however one can "back test."
Many of the "new" ETFs track or attempt to replicate an index that has been in existence for some time. I would argue that these ETFs should not be dismissed simply because they are "new" or without a "3-year" track record.
An example here might include the SPDR S&P Health Care Services ETF (NYSEARCA:XHS). Take note here that the ETF has ADV of less than 2k shares per day, but implied liquidity in excess of 750k shares per day. This fund launched in September of 2011, however the index has been around far longer.
Since the ETF tracks the index, wouldn't it make sense to give the ETF the benefit of the doubt here and look at the historical performance of the index when back testing?
Now, I am not oblivious to the fact that not all ETFs track Indices that have been around a long time. There are many actively-managed ETFs available today with even more in the pipeline. These active managers with proprietary methodologies may very well have to earn their track record the traditional way, by the clock.
Lastly, I would like to explain what happens to an ETF when it de-lists or closes down:
While in most cases, the investors will, in fact, lose the ability to trade the product on a listed exchange (they can often times trade off exchange), the fund itself will liquidate the assets in the fund (which, by the way, are held at a custodial bank). Once these assets are liquidated, the ETF holders will receive what you should know as FINAL NAV. This will be calculated based on the proceeds of the fund liquidation.
Fund managers do not look to execute poorly, so please understand that careful attention is given to the liquidation process to insure that the ETF holders get maximum value on closing.
The major risk is that the proceeds from liquidation could take a week (or more) to receive once a fund is delisted and the liquidation process commences.
It is very important that investors and fund managers understand how and why liquidity exists in ETFs, and that sourcing ETF liquidity is materially different in more ways than sourcing single stock liquidity.
AUM and ADV do NOT accurately assess an ETFs' implied liquidity. ETFs are backed by assets, not promises. While some ETNs and leveraged/inverse ETPs do contain swaps, the majority of ETFs are backed by equities, bonds and physical assets.