Slight confusion regarding arbitrage opportunities in ETFs as mentioned by investopedia

Question

On the "how ETF arbitrage works" page, investopedia says

The arbitrage opportunity happens when demand for the ETF increases or decreases the market price, or when liquidity concerns cause investors to redeem or demand the creation of additional ETF shares. At these times, price fluctuations between the ETF and its underlying assets cause mispricings. The NAV of the underlying portfolio is updated every 15 seconds during the trading day, so if an ETF is trading at a discount to NAV, a company can purchase shares of the ETF then turn around and sell it at NAV and vice versa if it is trading at a premium.

For example, when ETF A is in high demand, its price rises above its NAV. At this point the AP will notice the ETF is overpriced or trading at a premium. It will then sell the ETF shares it received during creation and make a spread between the cost of the assets it bought for the ETF issuer and the selling price from the ETF shares. It may also go into the market and buy the underlying shares that compose the ETF directly at lower prices, sell ETF shares on the open market at the higher price, and capture the spread.

From what I understood, for the above explanation to be correct, the ETF's (NAV per share) and the price of the ETF share that's traded, are two completely separate entities. The former is determined by the price of the ETF's underlying assets and the latter is defined by the supply and demand of the traded ETF share itself.

If I'm correct, then this implies that the ETF share price and the NAV per share are forced to be equal due to arbitrage traders only? I mean, if for example I was an insane/idiot trader, willing to quote a bid price higher than the current ask price and I complete that trade, this would mean that the ETF share price definitely goes higher than the NAV per share value. Now, how does the ETF share price again go back to matching the (NAV per share) value? Is it just that the market 'believes' that the ETF share price is supposed to be equal to its NAV per share, and hence no one would bid for the ETF share at a price higher than (NAV per share)?

Also, in this case, how would someone, who doesn't own any of the ETF shares, be able to capitalize on this mispricing? First of all, am I even correct about how the mispricing occurs?

Answer 1

Yes. Your understanding is correct. The only thing that keeps the ETF price in line is the aggregate market participants willingness to keep it in line. Under normal circumstances, that willingness is available usually in plenty of volume for most any size of ETF trader.

Consider there to be a continuum of ability to be able to purchase or sell at NAV. On the one hand is the open-end fund where you transact at the NAV on a guaranteed basis (but you give up intra-day trading ability). At the other end is a closed end fund where you have no ability to guarantee execution at NAV (even though it is likely to be a well known quantity). In between is the ETF. There is no explicit guarantee that you can execute at NAV. Instead a process was created to incentivize explicit or implicit market makers to provide an ability to enter or exit an ETF position. That process allows Authorized Participants to exchange the underlying basket of assets for the ETF and vice versa. Since this is a pretty straightforward (generally) problem to solve, this usually acts as a powerful incentive and the market sees plenty of ability to transact in many ETFs. Consider that the NAV is likely calculated based on a mid-market price for all of the basket. While in large, liquid ETFs, eg SPY, there are many non-market making participants looking to trade and there will likely be an ability to participate very very very near to NAV. On the other hand, a market maker will be looking to sell (buy) the ETF at a slightly (deliberately vague) higher (lower) price than where they can obtain the basket.

This is generally the case. On the other hand, consider a couple of other cases that might allow supply/demand to drive ETF prices away from NAV. First might be that the NAV calculation is not as accurate as might appear. For instance, high yield bonds where pricing might be subject to wide markets or obtaining liquidity in large size is impossible. Second, consider that APs tend to take what are considered to be low risk trades that require large amounts of leverage to make profitable. That means that they are dependent on their ability to obtain credit from their clearing firms or from banks. Should that credit get tighter, that could either force APs to widen their markets, stop operating, or even liquidate at adverse prices (there may be some sort of an inability to deliver baskets due to some settlement or credit issues). MUB (the muni bond ETF) traded at a discount for an extended period in the past.

In order to take advantage of these opportunities, the most effective way is to become an AP, but that is likely out of reach for most unless you wish to join or become a broker-dealer and do it full time. The second way is to find a correlated asset to trade against. For instance, if one found that HYG was trading rich but JNK was fair, one could do a pair trade. Given normal conditions (HYG not hard to borrow and reasonable costs to trade) that should be within an aggressive retail trader's scope.