The Couch Potato strategy thrives on simplicity, but advanced index investors (geeks) should understand what goes on under the hood of ETFs. One of the most important concepts is how ETF shares are created and redeemed. I’ll warn you that this gets a bit technical. But it turns out that this process is the single most important difference between ETFs and index mutual funds.
Let’s begin by looking at how a mutual fund creates new shares (or units). When you make a $2,000 contribution, your money goes directly to the mutual fund’s manager, who uses it to buy more securities. If the fund’s net asset value (NAV) per share is $20, the manager then creates 100 new shares ($2,000 ÷ $20) just for you. This is what’s meant by the term open-end fund: the number of units changes every time money moves in or out.
Closed-end funds, by contrast, do not create or redeem new units. They are launched with a finite number of shares, and if you want to invest in the fund you have to buy your shares from another investor who is willing to sell.
An open-ended discussion
ETFs trade on an exchange like closed-end funds, but they are open-ended: new shares are created to meet investor demand. Unlike mutual funds, however, ETFs do not create new units every time money flows in.
Instead, whenever necessary, ETF providers issue a large block of shares called a Prescribed Number of Units (PNU), typically in multiples of 50,000. The providers then work in partnership with third-party “designated brokers,” or DBs, who distribute these units to the public. A new ETF might be launched with 200,000 units, for example, and several DBs would divide these up and sell them on the secondary market. Going forward, when invetsors want to sell their units, the DBs are obliged to buy them back at a price very close to their net asset value.
(Geeky footnote: ETF providers south of the border use different terminology. If you’re reading American books or websites, PNUs are called creation units, and designated brokers are called authorized participants. Feel free to share this fun fact with your friends and coworkers.)
When you purchase $2,000 worth of an ETF trading at $20, the DB will simply fill your order with 100 shares from its inventory. However, if an institutional investor wants to buy $2 million worth of shares, the DB will not have enough inventory. So it will credit the investor’s account for 100,000 shares and simultaneously purchase $2 million worth of the fund’s underlying holdings. Then the broker will deliver that basket of stocks or bonds to the ETF provider, who will create 100,000 new shares and send them to the DB as payment.
That wasn’t so hard, was it?
This creation-redemption process may be complicated, but it has a couple of important benefits.
First, if the ETF is selling at a premium, a designated broker can buy the fund’s underlying securities and simultaneously sell shares of the ETF, thus making a risk-free profit. (If it is selling at a discount, the DB can do the opposite.) This presents an arbitrage opportunity, and when multiple DBs compete with each other to profit from it, the result is that the ETF’s price stays close to the NAV. That, of course, is exactly what investors want.
Second, the process prevents buy-and-hold ETF investors from being penalized by active traders. An inherent problem with mutual funds is they must always keep cash on hand to pay investors who redeem their shares, and this uninvested money is a drag on the fund’s returns. Worse, when investors sell in droves (such as during a market decline), the mutual fund has to liquidate holdings to pay them, which can mean forced sales and tax consequences for the fund’s other investors. ETFs never have to do this: if investors flee, the designated broker can just return blocks of shares to the ETF provider and receive the underlying securities in exchange. Because this is an “in-kind redemption” and not a sale, there is no taxable event.
For more information about the nuts and bolts of ETF construction, I recommend All About Exchange-Traded Funds, a new book by Scott Paul Frush (McGraw-Hill, 2011).