Net asset value is an accounting fact struck at a moment. An exchange-traded fund's market price is a live opinion. The gap between them is sometimes a mispricing and sometimes the more honest number.
12 min read · Level 2
A premium or discount is almost always presented as a defect in the fund, and much of the time it is the most honest number on the page. Net asset value is an accounting output: it takes the last recorded price of every holding, adds them up and divides. When those last recorded prices are minutes old and executable, that output is excellent. When they are hours old, or when they are dealer quotes on bonds that nobody has traded this week, it is a construction - and a fund quoted at a discount to it is not mispriced, it is a live crowd of buyers and sellers disagreeing with a stale spreadsheet. I would put it more strongly: in the third situation the discount is the only genuine price discovery happening anywhere in the structure, because the holdings themselves are not being priced by transactions at all. That has an uncomfortable implication for the standard advice, which is to be wary of funds trading away from net asset value. The wariness is warranted, but it should attach to the holdings rather than to the wrapper. A persistent discount on a fund of illiquid assets is telling you something true about what those assets would fetch, and the alternative structure - an open-ended fund transacting at the struck value - does not make that truth go away. It just hides it until the day redemptions are gated, at which point the holder learns it all at once.
Net asset value is the arithmetic from lesson one: assets, less liabilities, divided by units. For Fund A it was 50.00. It is a valuation struck at a moment, from the prices of the holdings at that moment, and it is an accounting output rather than a negotiated figure.
For an open-ended mutual fund, that number is the transaction price. You buy and sell at net asset value, and there is no separate market in the units at all. The cost of that arrangement is timing: your order is filled at the next value struck, after you committed, which lesson one worked through.
For an exchange-traded fund, the transaction price is whatever the market is quoting. Those are two different numbers and they need not agree.
Net asset value 50.00 Best bid 50.09 Best offer 50.15 Mid 50.12
Read that carefully, because the published premium and the premium you pay are not the same figure. Measured at the mid, the fund trades at (50.12 - 50.00) / 50.00 = 0.24%, or 24 basis points, above the value of its holdings. That is the number a data provider reports. A buyer lifting the offer pays 50.15, which is (50.15 - 50.00) / 50.00 = 0.30% above. The spread itself is 0.06, which is 12.0 basis points of the mid, and the buyer wears half of it in the premium and the seller the other half in the discount.
What stops the gap widening is a wholesale mechanism running behind the exchange. Specialist firms - authorised participants - can assemble a basket of the fund's underlying securities and deliver it to the fund in exchange for a large block of newly created units, or hand back units and receive the basket. The block is fixed and large: take a creation unit of 50,000 units here.
Buy the basket at its market value 50,000 x 50.00 = 2,500,000 Deliver it, receive 50,000 units worth 50,000 x 50.12 = 2,506,000 Gross 6,000 Less the firm's own dealing costs, 0.09 per unit -4,500 Net 1,500
So the arbitrage only pays while the gap exceeds the firm's cost of executing it - here about 0.09 per unit, or 18 basis points. A premium of 24 basis points leaves 6 basis points of profit, so the gap is still worth closing here - and it stops being worth closing once it narrows to around 18. That is why the price settles near the cost of the arbitrage rather than at the value of the holdings. A small, persistent premium or discount is the price of the plumbing, not a malfunction.
The mechanism runs in both directions. At 49.85 the fund trades 30 basis points below net asset value; the firm buys units cheaply, redeems them for the basket, and sells the basket.
Because a premium or a discount means three completely different things depending on a fact about the holdings, and the third of them is where money is lost.
Situation one: the holdings trade in the same hours as the fund. A fund of large domestic listed companies, quoted while those companies are quoted. Here net asset value is a current figure, the arbitrage is cheap, and a gap beyond the cost of executing it is genuine and usually brief. The 24 basis point premium above is this case.
Situation two: the holdings' market is closed. A fund listed in one region holding shares that stopped trading nine hours ago. Its net asset value is a faithful record of a session that has ended. Suppose the relevant regional futures have fallen 2.1% since that close and the fund is quoted at 48.95 against a struck value of 50.00. That prints as a discount of (50.00 - 48.95) / 50.00 = 2.10%.
It is not a discount. The fund has not fallen below the value of its holdings; the recorded value of the holdings has not caught up with the world. The market price is the current opinion and the net asset value is the stale one, and a buyer congratulating themselves on a bargain has misread which of the two numbers is out of date. This is the case most readers have backwards.
Situation three: the holdings are genuinely hard to trade. Thinly traded corporate bonds, small property companies, frontier equities. Now the net asset value is itself an estimate - struck from dealer quotes that may not be executable in size - the arbitrage is expensive and slow, and the gap can be wide and persistent. Suppose the fund is quoted at 48.60 against a struck 50.00: a discount of 1.40 / 50.00 = 2.80%.
Here neither number is reliable, and the discount may be the market's honest statement about what the holdings are worth when somebody has to sell them.
That is the sentence in this lesson worth keeping: the wrapper's liquidity is not the holdings' liquidity. An exchange-traded fund can be bought and sold in seconds while the assets inside it cannot be, and nothing about the structure repairs that. Under stress it is the gap that moves, because the gap is the only thing free to move.
An open-ended fund does not escape this by transacting at net asset value. The same mismatch shows up in a different form: redemption gates, swing pricing that charges exiting holders the cost of the sales their exit forces, or outright suspension. The problem is a property of the holdings. The wrapper only decides which shape it takes.
Educational, not personalised financial advice. Nothing in the Investor Academy is a recommendation to buy, sell or hold anything — what is right for you depends on your own circumstances and the rules where you live.
Put the three situations side by side, on the same struck value of 50.00.
Situation holdings market price gap what it means One same hours, liquid 50.12 +0.24% real, small, bounded by arbitrage cost of ~18bp Two market closed 9h ago 48.95 -2.10% the price is current; the valuation is not Three thinly traded 48.60 -2.80% neither number is firm; the gap is information
The practical instruction differs in each. In situation one, transact with a limit order near the mid and think no further about it. In situation two, do not read the gap at all - compare the fund against an index or futures contract that is actually trading, because the published premium is measuring the clock rather than the fund. In situation three, treat the discount as a question about the holdings and go and read what they are.
One further point about situation one that gets stated backwards more than any other in this subject. An exchange-traded fund's own trading volume is a poor measure of how much of it you can buy. Because the creation mechanism exists, a market maker can quote you size in a fund that has barely traded all day, by buying the underlying basket and having new units created. The fund's liquidity is, to a first approximation, the liquidity of its holdings.
That cuts both ways, and the second way is the one that costs money. A fund with heavy daily volume whose holdings are thin is not liquid; it is busy. The volume is holders trading with each other, and the moment they are all on the same side, the mechanism that is supposed to absorb it has to go and transact in the thin market underneath.
Reading every discount as an opportunity. In situation two there is no discount at all, only a stale denominator, and buying on that basis is buying at the current price while believing you got a reduction.
Judging liquidity by the ticker's volume. A low-volume fund on deep holdings can usually be traded in size; a high-volume fund on thin holdings cannot. The question is always what is inside, which is the same question lesson one ended on.
Using market orders, particularly in the first and last minutes of a session. Spreads are widest then, the underlying holdings are least reliably priced, and a market order accepts whatever is showing. A limit order costs nothing and occasionally saves a great deal.
Assuming an open-ended fund is safer because it transacts at net asset value. It removes the premium and discount and leaves the underlying mismatch untouched, where it reappears as a gate or a suspension - and a gate arrives at the moment you most want your money, which a visible discount at least lets you decline.
Treating the published premium as the price you will pay. It is measured at the mid. You transact at the edge, and on the numbers above that is the difference between 0.24% and 0.30%.
Most fund pages publish a premium or discount measured at the mid or at the previous close. Take the live bid and offer, compare each with the most recent struck value, and know which side of the spread you are about to pay. On a broad liquid fund this is a formality; on anything narrower it is the largest single cost of the transaction.
If the answer is more than an hour ago, the published gap is measuring the passage of time rather than any view about the fund. Compare against something that is currently trading instead - an index future, or a listed proxy - and treat the printed premium as uninformative until the underlying market reopens.
Set a limit at or inside the mid and let it work. The spread you accept is a permanent cost of the round trip, paid twice, and it is the one cost in this whole subject that is entirely within your control at the moment of the transaction. Nothing about a broad fund requires you to transact in the first five minutes of a session.
Work out what your intended order is as a share of the daily traded value of the fund's underlying holdings, not of the fund itself. For a broad index that number is negligible and you can stop thinking about it; for a narrow fund it can be large enough that your own order is the reason the price moves.
Before every purchase and sale of an exchange-traded fund, whenever a fund appears to be trading away from the value of its holdings, and whenever you are choosing between an exchange-traded and an open-ended version of the same portfolio.
Do not use a persistent discount as a reason to buy without first establishing what the holdings are and when they last traded, and do not rely on net asset value as a realisable figure for a fund whose underlying assets do not trade regularly - it is an estimate wearing the clothes of a fact.
In stress episodes a shape recurs across funds holding assets that trade rarely. The fund continues quoting and transacting every second of the session, because an exchange does not require its underlying market to be functioning. The holdings, meanwhile, are being quoted by dealers who have stopped committing capital, so the struck valuation drifts along on stale marks. The printed discount widens to a level that reads as absurd, and commentary describes the fund as broken. When the underlying market reopens properly, the recorded valuations usually move most of the way towards where the fund had been trading. The wrapper was not broken; it was the only part of the structure still doing price discovery.
When a fund trades far from its struck value, the first question is whether the holdings are being priced by transactions or by estimates. A wide discount on estimated marks is evidence about the holdings.
Create a free account to read the rest and apply it to your own investing.