ETF or Index Fund? The Wrapper Matters Less Than the Transacting

The same holdings, two different machines

An index fund and an exchange-traded fund tracking the same index own substantially the same things. The difference is not in what they hold; it is in how you get in and out.

With an index fund, you place an order with the fund. Units are created or cancelled for you, and the price is the fund's net asset value struck at the end of the day. You do not trade with another investor and there is no counterparty to find. Whatever the market does during the day, you get that day's calculated value.

With an ETF, you buy from another investor on the exchange, at whatever price the two of you agree, through a broker. The fund itself is not involved. There is a mechanism intended to keep the exchange price near the underlying value — large participants can create or redeem units in bulk and profit from any gap — but it is a mechanism operated by third parties for their own reasons, not a guarantee.

Everything that follows comes from that one structural difference.

Where the costs actually are

The expense ratio is the visible number and it is usually the smaller one. There are four other costs, and they do not apply equally to the two structures.

The spread. Buying an ETF means paying slightly more than a seller would receive at the same moment. That difference is a real cost, paid on the way in and again on the way out, and it does not appear in any ratio. An index fund has no spread — you transact with the fund at its calculated value.

The premium or discount. An ETF's exchange price can drift from the value of what it holds. In calm conditions the gap is generally small. It widens when the underlying market is disturbed, which is precisely when people want to trade, and an investor selling in a stressed session can receive less than the holdings were worth. An index fund cannot have this problem by construction.

Brokerage and the machinery around it. ETFs need a broking account and incur transaction charges, and each purchase pays them. This matters most for someone investing a modest amount every month, where fixed costs are a large proportion of each contribution. Index funds are bought directly with no broker involved.

Tracking difference. Both structures lag their index and the lag is not the same for both. It compounds silently and is separate from the expense ratio — the distinction is set out in tracking error against tracking difference.

Notice that three of those four apply only to the ETF, and that all three are transaction costs rather than holding costs. The wrapper with the lower annual charge is not automatically the cheaper one; it depends entirely on how often you transact and how much you transact at a time.

The honest gap, and why it is placed here rather than at the end

At this point an article of this kind normally produces a table of typical spreads and typical premiums. We are not going to, and the reason is the whole argument of the page.

Settling which wrapper is cheaper for a particular investor needs three things: bid-ask spreads for Indian equity ETFs, their premium or discount to net asset value — particularly during stressed sessions, when the question actually bites — and tracking difference for a matched pair of ETF and index fund following the same index over the same period.

This site holds none of them. Our figures come from broad index and interest-rate history, which says nothing about the cost of transacting in any particular wrapper.

So this page states no figure for any spread, premium or transaction cost, and none should ever be added from a plausible-sounding recollection. A made-up "typical spread" would sit in the exact place where the article's central claim is decided, and it would decide it by invention. Given that this page's whole argument is that the transacting costs are what matter, quoting an unsourced number for them would be worse here than almost anywhere else on the site.

What follows is therefore a way of working out the answer for yourself, using figures you can observe, rather than an answer we have not earned.

How to decide for your own case

Start with how you invest. If you are putting a fixed amount in every month, the fixed costs of transacting are paid every month and the index fund route avoids most of them. If you are placing a single large sum and holding it for many years, a spread paid once on a large amount is a small proportion, and the lower annual charge has many years to work.

Then observe the spread yourself rather than looking for a published figure. Open the order book for the ETF you are considering, at a normal time of day, and look at the gap between the best buy and best sell prices as a proportion of the price. Then look again at a quiet moment — early in the session, or on a day the market is falling. What you see is your answer, for that ETF, and it is more reliable than any average across ETFs.

Check the exchange price against the fund's stated value. Fund houses publish an indicative value through the day. A persistent gap tells you the arbitrage mechanism is not working well for that particular ETF, which usually means it is not traded much.

And weigh how you behave. An ETF can be traded all day at a live price. For some people that is convenience; for others it is an invitation, and the cost of a structure that makes trading easy does not appear in any ratio at all. If you know which sort you are, that is a legitimate input.

What to take away

The choice is not really between two products. It is between transacting with a fund at a calculated price and transacting with a stranger at a negotiated one, and which is better depends on how often you do it, how much at a time, and how calm the market is when you need to.

For most people investing regular monthly amounts, the index fund route removes several costs entirely and the annual difference is small enough to be outweighed by them. For a large one-off holding, the calculation can go the other way — and you can do that calculation with numbers you observe yourself in a few minutes, which is a better basis than a figure this site cannot source.