Index Fund vs ETF: What Actually Differs for a Singapore Investor

The category error behind the question

An index fund and an ETF are not two parallel boxes sitting side by side on a shelf. One describes a strategy; the other describes a product form. Amova's page for the ABF Singapore Bond Index Fund states the distinction directly: ETFs and unit trusts are both investment vehicles, but they differ mainly in two ways. The page names those two as the way they are bought and sold, and — in its words — that ETFs also tend to have lower expense ratios and offer greater transparency.

The same page is explicit that the form does not determine the strategy. While most ETFs passively track an index, there are some ETFs that are also actively managed. That sentence is why the comparison cannot be made as a simple either-or. Tracking an index is a decision about what a portfolio holds. Listing as an ETF is a decision about how units in that portfolio change hands. A fund can be passive and exchange-traded, or active and exchange-traded. The evidence supports saying that most ETFs track an index; it does not support saying that every ETF does.

The question worth asking is therefore not "index fund or ETF". It is which index, and which wrapper for implementing it.

An index strategy can take more than one product form

An ETF is one way of implementing an index strategy, not the only one. The clearest illustration on a Singapore fund page comes from the ABF Singapore Bond Index Fund. Its benchmark field reads "iBoxx ABF Singapore Index calculated on a total return basis". Its fund-structure field reads "Open-ended Listed Unit Trust traded on Singapore Exchange". The index mandate and the legal wrapper are recorded in two separate fields on the same page, because they describe two separate things: what the portfolio is built to do, and what the vehicle is.

The Amova Singapore STI ETF (SGD Acc) Class page makes the strategy side explicit. It describes passive ETFs as having typically lower management fees, as they passively track an index to just deliver market returns with no expectations of outperformance. MoneySense's guide to choosing between ETFs tracking the same index, last updated 21 September 2026, goes further: for passive ETFs that track the same index, their portfolios are essentially identical.

That last point carries a consequence. If two passive products tracking the same index hold essentially the same portfolio, the difference between them is not what is inside the fund. It is everything wrapped around it.

How the two forms actually trade

The first of the two differences named on the Amova page is execution. ETFs can be purchased or sold on a stock exchange the same way that a regular stock can, and their price will fluctuate through the trading session. Just like stocks, you can trade ETFs on a stock exchange at any point during market hours.

Unit trusts work on a different clock. They only trade once a day after the market closes, based on their end-of-day net asset value price. MoneySense's unit trust guide, last updated 15 July 2026, sets out the mechanics behind that price: the price of each unit is based on the fund's net asset value divided by the number of units outstanding, and the NAV is usually computed daily to reflect changes in the prices of the investments held by the fund.

For an investor, the practical difference is the price you transact at. On-exchange, you deal at a price that moves during the day. In a unit trust, you deal at a price struck after the close.

What the fees are, and where they come out

The MoneySense ETF guide defines the measure: the TER is the annual cost of the fund, expressed as a percentage of assets under management and deducted directly from the fund's assets. It covers the fund manager's fee and other fund-level expenses. Being deducted from the fund's assets is the part that matters — the cost is taken out before returns reach the investor, rather than billed separately.

The range is wide even among products doing the same job. For similar ETFs tracking the same index, the TER can range from as low as 0.03% to over 0.5% of the net asset value of the ETF per annum. MoneySense's illustration of what that spread does over time is stated on the same page: on a $50,000 investment over 20 years, the difference between a 0.1% TER and a 0.5% TER — all else equal — compounds to thousands of dollars of foregone returns.

The unit trust guide frames the same idea in component terms. Recurring fees are paid by the fund and include management fees, trustee fees and other fees. These costs are passed on to you. The Amova page adds the comparative claim: ETFs also tend to have lower expense ratios and offer greater transparency.

The entry ticket: one board lot

Under section 8.3 of the SGX Rulebook, except for the unit share market, the minimum order size is 1 board lot. Orders may be in multiples of a board lot.

The rulebook fixes the floor at one lot. It does not fix the number of units in a lot, which is a product-level field. Both Amova products publish "Lot Structure 1 unit per lot" on their pages as of October 2026. A single unit is the published entry size for these two, but the lot structure has to be read off the individual fund, not assumed from the exchange rule.

Distributing or accumulating

Dividend treatment is a share-class decision, not an index decision. MoneySense lists it as one of the differences between ETFs tracking the same index: different dividend treatment, where some ETFs distribute dividends to investors and others reinvest them automatically. This "distributing vs accumulating" choice affects how returns are delivered.

The Amova Singapore STI ETF (SGD Acc) Class page describes the mechanics of the accumulation class in its own words: unlike the existing distribution share class which pays out dividends semi-annually in cash, this accumulation share class automatically reinvests dividend payouts back into the fund and accumulates towards the investors' holdings. Its published dividend distribution frequency is NIL.

The ABF Singapore Bond Index Fund sits on the other side of the line. Its published dividend distribution frequency is semi-annually at the discretion of the Manager. Two index-tracking products, one wrapper each, opposite treatment of the income.

Same index, different outcomes: domicile, spreads, brokerage

MoneySense is explicit that the TER is the starting point for comparison but not the only cost that matters. Bid-ask spreads, brokerage commissions, and withholding taxes also affect your net return.

Withholding tax is the item with the largest single quoted rate in this evidence set. An ETF domiciled in the US and held by a Singapore investor is subject to a 30% US withholding tax on dividends. For a dividend-paying ETF, this difference in withholding tax treatment can meaningfully affect your net return, sometimes more than the difference in TER. Domicile is a fund-level fact, invisible in the benchmark name and absent from the TER.

Spreads and brokerage sit on the trading side. They arise from the mechanics described earlier — transacting on an exchange, at a price that moves during market hours — and they are costs the TER does not capture.

Two Singapore-listed index products, in the same wrapper

Both Amova products carry the identical fund-structure field: Open-ended Listed Unit Trust traded on Singapore Exchange. Both publish a lot structure of 1 unit per lot and a daily valuation frequency. The differences sit elsewhere.

The Amova Singapore STI ETF (SGD Acc) Class tracks the Straits Times Index, in total return. Its management fee is 0.09% per annum. The page records the change: the management fee has been reduced from 0.20% p.a. to 0.09% with effect from 1 October 2025. Its total expense ratio is 0.24% per annum, audited as of the financial period ended 30 June 2025. Its trustee fee is up to 0.045% per annum. Its dividend distribution frequency is NIL. Its NAV was SGD 5.3945 as of 2 October 2026, with total fund size of SGD 52.03 million as of 2 October 2026.

The ABF Singapore Bond Index Fund tracks the iBoxx ABF Singapore Index calculated on a total return basis. Its management fee is 0.15% per annum. Its total expense ratio is 0.24% per annum, audited. Its trustee fee is up to 0.045% per annum. Its dividend distribution frequency is semi-annually at the discretion of the Manager. Its NAV was SGD 1.1002 as of 2 October 2026, with total fund size of SGD 1,137.71 million as of 2 October 2026.

Three observations fall out of those two columns without any calculation. The management fees differ, 0.09% against 0.15%, while the TER is identical at 0.24% for both — the management fee is one input into the TER, not the whole of it. The trustee fee ceiling is the same. And the fund sizes differ substantially, at SGD 52.03 million against SGD 1,137.71 million, both as of 2 October 2026.

What to do with the numbers

Treat the fund page as a set of fields to read in a fixed order. Read the benchmark field first, because it tells you what the portfolio is built to deliver. Read the fund-structure field second, because it tells you what kind of vehicle is delivering it, and the two are independent. Read the TER third, together with its audit period — a TER labelled "audited as of the financial period ended 30 June 2025" is a historical figure, not a forward guarantee, and the management fee on the same page can already have moved, as it did with effect from 1 October 2025.

Then read the fields that no TER captures: the distribution frequency, which decides whether income arrives as cash or is reinvested; the lot structure, which sets the entry size against the SGX minimum of one board lot; and the domicile, which decides whether the 30% US withholding tax on dividends applies. Check the spread and the brokerage on the trading side, since neither appears in the expense ratio.

Finally, re-read all of it at the point of dealing. Every figure quoted here carries a date — 2 October 2026 for the NAVs and fund sizes, 30 June 2025 for the audited TER, 1 October 2025 for the fee change. These values move. The method is the durable part; the numbers are a snapshot.