ETF vs index fund: key differences, costs and what the price data shows

An ETF (exchange-traded fund) and an index fund are not opposites. An index fund is any fund that follows an index, and it comes in two wrappers: a mutual fund, which you buy from the fund company once a day at its net asset value, and an ETF, which trades on an exchange all day like a share. "ETF vs index fund" usually means index ETF against index mutual fund.

Two funds tracking the same index deliver close results; the experience of owning them differs. Below is the investor answer first, then the data view: what an index level, an ETF and a mutual fund look like to an ETF API and to an indices API.

ETF vs index fund at a glance

What is an ETF, and how do ETFs work

An ETF is a fund that holds a portfolio (stocks, bonds, commodities or futures) and splits ownership into shares that trade on an exchange like any other stock. Most ETFs follow an index; some are actively managed. The clever part is how the share price stays tied to the value of what the fund holds.

An ETF has two markets. In the primary market, large dealers called authorized participants hand the fund a basket of the index stocks and receive new ETF shares in large blocks, or hand shares back and take stocks out. In the secondary market, everyone else trades those shares with each other all day. If the ETF drifts above the value of its basket, dealers create shares and sell them; if it drifts below, they buy shares and redeem them.

A mutual fund works differently. Every order collected during the day executes at the same NAV after the close, and the fund buys or sells the underlying stocks for cash to meet them. Selling stocks for cash can realize capital gains inside the fund, and in the US those are passed on to every holder. In-kind redemptions let an ETF hand out stocks instead of cash, which is the main reason ETFs tend to distribute fewer capital gains.

Costs: expense ratio, spread and the rest

Both wrappers charge an expense ratio, taken from fund assets a little every day, and for large broad-market index funds it is small either way. The costs that differ sit around the edges, and the biggest one for an ETF is also the easiest to measure from market data: the bid-ask spread.

• Bid-ask spread (ETF). Paid each time you trade. Under a basis point on the largest funds, much wider on thin ones, and wider for everyone at the open and in stressed markets.
• Commissions (ETF). Many brokers now charge none for ETF trades. Check yours.
• Sales loads and minimums (mutual fund). Some share classes charge a sales load, and many funds require a minimum first investment.
• Short-term redemption fees (mutual fund). Some funds charge you for selling within a few months of buying.
• Tax drag (US). Capital-gain distributions from a mutual fund are taxable in a regular account even in a year you sold nothing.

What the price data shows: index, ETF and fund

Put the three objects side by side as data and the difference becomes concrete. The index level is a calculation you can read but not trade. The ETF is a security with its own quotes and trades. The index mutual fund has one price a day, published by the fund, and no intraday data at all.

Look at the spread row. The index quote has a bid and an ask because the API publishes index-style levels in a two-sided, CFD-like form, but no order book stands behind it and nobody fills an order against the index itself. The ETF spread is the actual cost of trading its shares. The mutual fund has neither: you would see one NAV per day, after the close.

The two lines are almost one. Across all 78 five-minute bars of that session, measured as percent change since the first bar, the fund never drifted more than 0.041 percentage points from the index. What differs is the level: the index sat at about 10.04 times the ETF price all day. That ratio is arbitrary, since each fund picks its own starting share price, and it moves slowly with fees and with dividends the fund holds until it pays them out.

import os
import sys

import requests

BASE_URL = "https://api.tickerlayer.com"
HEADERS = {"x-api-key": os.environ["TICKERLAYER_API_KEY"]}
DAY = "2026-09-25"


def five_minute_closes(asset, symbol):
    url = f"{BASE_URL}/{asset}/agg/{symbol}/5/minute/{DAY}/{DAY}"
    resp = requests.get(url, headers=HEADERS, params={"sort": "asc", "limit": 5000}, timeout=15)
    if resp.status_code != 200:
        sys.exit(f"{symbol}: HTTP {resp.status_code} {resp.text[:160]}")
    return {bar["t"]: bar["c"] for bar in resp.json()["results"]}


index = five_minute_closes("indices", "US500")
etf = five_minute_closes("etfs", "US500ETF")
times = sorted(index.keys() & etf.keys())
if not times:
    sys.exit("No overlapping bars for that day.")
t0 = times[0]
gaps = [abs((etf[t] / etf[t0]) - (index[t] / index[t0])) * 100 for t in times]
ratios = [index[t] / etf[t] for t in times]
print(f"{len(times)} five-minute bars in both series")
print(f"largest gap in % change since the first bar: {max(gaps):.3f} percentage points")
print(f"index / ETF ratio: {min(ratios):.4f} to {max(ratios):.4f}")

78 five-minute bars in both series
largest gap in % change since the first bar: 0.041 percentage points
index / ETF ratio: 10.0377 to 10.0417

The same check works for any fund and the level it follows: swap the symbols, keep the logic. The ETF bars reference lists the intervals, and US500ETF has a delayed chart of the fund.

When each wrapper tends to fit

One caveat for commodity funds: an ETF that holds futures rather than the physical asset can drift away from the spot price for reasons that have nothing to do with fees. Contango and backwardation explains that roll cost. And if you want to know what an index fund is tracking in the first place, what a stock index is covers how the weights are set.