Exchange-Traded Funds are essentially open-ended mutual funds that trade on stock exchanges. Each fund is a trust that holds specified assets, typically shares of stock; the shareholders of the ETF are the trust’s “beneficiaries.” Where an ETF differs from a closed-end mutual fund is that the “trust agreement” requires the managers of the ETF to intervene in the market to keep the market price of each ETF share within a narrow range around its net asset value (value of the trust’s assets, the underlying stocks, divided by all outstanding shares). Note that closed-end mutual fund shares, unlike ETF shares, may, and do, sell at significant discounts from, and premiums to, their net asset value.

Here’s how the pricing mechanism works for ETF shares. Let’s say an ETF has an NAV of $100 and the shares of the ETF are trading at $100. Assume that several investment fund managers decide nearly simultaneously to sell their holdings of the ETF. These investment funds then put all of their ETF shares on the market. At this point the number of sellers of the ETF shares overwhelms the number of buyers and the floor specialist, to attract more buyers, pushes the price of the ETF down to say, $95. At this point some other investment fund managers may see a bargain opportunity (shares with a liquidation value of $100 trading at $95) and buy the ETF. If this occurs the price of the ETF will move back up to the NAV.

What if no other investment funds jump in to buy at the discounted price? At that point the managers of the ETF will cause the ETF itself to buy the shares. These bids will then move the ETF price upwards towards the NAV. Where will the ETF get the money to buy back its own shares? Eventually it will sell the stocks that it holds. These sales may then push down the market price of the underlying shares.

So, when many, many ETF sell orders go to the exchange floor there are several pricing factors at work: some investment funds selling the ETF shares, other investment funds attempting to get a bargain, the ETF buying back the ETF shares, and the ETF selling its assets (the underlying shares). Eventually an equilibrium is reached where the market value of the ETF share is approximately the NAV. You noted correctly that this equilibrium may be at a lower NAV. It all depends on the liquidity of the shares of the ETF itself, the relative size of the ETF’s holdings, the liquidity of the underlying stocks, and the extent of the imbalance between buyers and sellers of the ETF.

When several investment funds decide to buy an ETF the process operates in reverse; the ETF issues new shares and buys the underlying stocks.

So large purchases and sales of an ETF, just like large purchases and redemptions of an open-ended mutual fund, can affect the market value of the underlying assets be they stocks, bonds, or commodities.

If you want to find out the details for any specific ETF, read its prospectus and governing documents. Also, I believe an ETF can borrow short-term to maintain market liquidity without having to immediately place on the market the assets that it holds.

Peter F.