How simulations are calculated

Many all-in-one funds are only a few years old. To compare them with older funds, a simulation estimates how their mix of investments would have done before they launched. This page explains how it works and how close it has come to real results.

The short version

Every all-in-one fund publishes its mix, for example 36% Australian shares, 26.5% global shares and so on. This page calls each kind of investment in a mix an ingredient. For the years before a fund launched, the simulation works out what its mix would have returned each day from the real returns of ASX index ETFs that hold the same kinds of investments. Before 2013 it uses stand-ins: US-listed index funds converted to Australian dollars, and Australian interest rates for bonds and cash. It then takes off the fund's fee. For a geared fund (one that borrows to invest more), it also allows for the loan and its interest.

From launch day, a fund's figures use only its own real prices. The exceptions are DZZF, DGGF and IGRO. They traded so rarely at first that their prices didn't follow the value of what they held, so DZZF and DGGF are simulated until 2021 and IGRO until 2024.

The three periods in a fund's history

1. Before every ingredient has data. Nothing is shown. The simulation never guesses at a missing ingredient or goes back before January 1997.
2. From when every ingredient has data, up to launch. This part is simulated. Charts draw it as a dotted line, and tables mark any figure that uses it with an asterisk (*).
3. From launch to today. Real prices, with distributions (the income a fund pays out) reinvested.

For example, VDHG launched in November 2017 and its simulation starts in January 1997. GHHF launched in April 2024 and also starts in January 1997, so all but the last two and a half years of its figures since 2000 are simulated. VanEck's all-in-one funds start in April 2019, when data for their youngest ingredient (global listed property, hedged) begins. The Since column of the All-in-one ETFs table shows each fund's start dates.

Where each fund's mix comes from

Each mix comes from the documents of the company that runs the fund (the issuer): its product disclosure statement (PDS), web page or holdings file, using the most detailed split the issuer publishes. When an issuer gives only a broad group, such as "international shares", the whole group goes in one ingredient, in this case global shares.

The ingredients

Each ingredient uses ASX index ETFs from the dates in the second column. Before that, it uses the stand-in in the third column, if it has one. A hedged fund cancels out most of the effect of exchange rate changes on its returns. The Portfolio page uses the same ingredients to extend single ETFs, such as VAS or GOLD, back before they launched.

A stand-in only fills the dates before the ETF that takes over from it has data. It must also pass a test against that ETF, over at least 3 years when both have data. Their monthly returns must move closely together, with a correlation of 0.9 or more (where 1 is a perfect match). Their annual returns must be within 1.5% a year of each other. Some older ETFs failed and aren't used. For example, VGS failed as a stand-in for VGAD, which holds the same shares but is hedged. Exchange rate changes made their returns differ too much.

BEMG and EXUS don't yet have 3 years of real prices, so their stand-ins can't be tested this way. BEMG's stand-in is IEM, which tracks the same index. EXUS's is IVE, which holds a slightly different set of countries. Their pages show how closely IEM and IVE have followed them so far.

Before 2013

Yahoo Finance prices for ASX ETFs include distributions only from January 2013. Before then they show price changes alone, so they aren't used. For earlier years:

• US-listed index funds, with distributions reinvested, are converted to Australian dollars at each day's exchange rate. The US market closes before the ASX opens the next morning, so each US day's return counts on the next business day.
• Hedged ingredients use the US fund's return in US dollars, plus the difference between Australian and US 3 month interest rates. That difference is roughly what a currency hedge adds or costs.
• Global shares use three Vanguard index funds covering the US, Europe and the Pacific. The split between them follows each fund's price, as the size of each market did. It is 73%, 18% and 9% in June 2026, and the US part was about 41% in 1997 and 65% in 2020.
• Australian bonds are worked out each day from the 10 year Australian Government bond yield (the interest rate on those bonds). The return is the interest earned, plus the rise or fall in bond prices when the yield moves.
• Cash earns the 3 month bank bill interest rate.
• Gold before November 2004 uses gold futures: contracts to buy gold on a set date, traded in New York. The contract due next trades close to the price of gold itself. It is converted to Australian dollars and counts on the next business day, as the US funds do.

Each of these stand-ins passed the test above against the ETF that took over from it:

Cash returns barely change from month to month, so correlation says little about them. Cash is tested on annual return alone. The Australian shares, global shares, Australian bonds and cash ingredients were also compared with published index returns for each decade from 1996 to 2026.

Factor funds

Dimensional's and Avantis's funds are factor funds. Compared with the market as a whole, they hold more of smaller companies, value companies (those with low share prices compared with the value of their net assets) and more profitable companies. Index ETFs don't do this, so these funds aren't simulated from a mix. Where they exist, each uses stand-ins that pick companies the same way:

• The fund's own unlisted units. Dimensional ran DACE, DGCE and DFGH as unlisted funds for years before they listed on the ASX, and their daily unit prices are available from 2014.
• The same manager's US-listed funds, converted to Australian dollars. Avantis's go back to 2019, so before that Avantis's funds use Dimensional's, which pick companies in a similar way.

Where there are 3 years to compare, each stand-in passed the test above against the one that takes over from it. DGVA and DGSM have traded for only 2 years, but Dimensional publishes the yearly returns of their unlisted funds, which gives a longer check: over the 10 years to August 2026 their simulations came within 0.1% and 0.4% a year of them, and since those funds began (1999 and 2000) within 0.5% and 0.1% a year. Each fund's page gives its own checks.

How each day's return is worked out

For an ungeared fund, the simulation multiplies each ingredient's return that day by its percentage of the mix, adds them up, and takes off one day of the fund's fee. Say a fund is 40% Australian shares and 60% global shares. If VAS rises 1% and VGS falls 0.5% on the same day, the mix returns 0.4 × 1% + 0.6 × -0.5% = 0.1% before the fee. If an ingredient doesn't trade on a day, its move counts on its next trading day.

The simulation resets to the fund's mix every day. Real funds rebalance (trade back to their mix) less often, but checks found this makes little difference for ungeared funds. The ingredient ETFs and US funds already have their own fees taken out of their returns. The fund's fee comes off as well, so the simulation tends to come out slightly low.

Geared funds

The geared funds (GHHF, GGBL and GNDQ) keep their borrowing between 30% and 40% of their total assets. This is their loan-to-value ratio (LVR). The loan is a share of the fund's total assets, not of investors' own money, so $1 of investors' money buys more than $1.30 to $1.40 of assets:

• 30% LVR: $0.30 of every $1 of assets is borrowed, so $1 of investors' money buys 1 ÷ 0.70 = $1.43 of assets.
• 35% LVR: 1 ÷ 0.65 = $1.54.
• 40% LVR: 1 ÷ 0.60 = $1.67.
• Twice investors' money ($2 of assets) would be 50% LVR.

These funds don't reset their borrowing every day, as many leveraged funds in the US do. The loan doesn't move with prices, so the LVR rises when the market falls and drops when it rises. At 35% LVR, $1.54 of assets carries a $0.54 loan. After a 10% fall the assets are worth $1.38 and the loan is still $0.54, an LVR of 39%. Betashares checks the LVR every day and says it rebalances only when the LVR moves outside 30% to 40%. Above 40% the fund sells assets and repays part of the loan. Below 30% it borrows more and buys more.

The simulation works the same way. It starts at 35% LVR. Each day it:

• applies the mix's return to all the assets,
• adds interest to the loan, charged at the cash return (AAA, or the bank bill rate before 2013) plus 1% a year,
• adds the fund's fee on all the assets to the loan too,
• and if the day ends with the LVR below 30% or above 40%, borrows more or repays part of the loan to bring it back to 35%.

The funds' product disclosure statement doesn't say what level they go back to. The simulation uses the middle of the range, as Betashares does in its own simulations of these funds, and the LVRs Betashares publishes fit it. GHHF's went from 34.6% at launch in April 2024 to 30.5% in June 2025, and was 35.4% by August 2025. GNDQ's rose from 30.8% to 37.7% in the fall of early 2025 without going back to 35%. Replayed on the ETFs GHHF holds, the simulated LVR stayed within 1.4 points of every GHHF value found, and topped up the loan in late July 2025. Adding interest and the fee to the loan matched those values more closely than taking them from the assets.

From 1997 to its launch, the simulated GHHF borrowed more 12 times and repaid part of its loan 9 times. Because the loan carries over from one day to the next, the return over any period depends on where the LVR stood when it began. Every chart, table and portfolio that shows a geared fund's simulated years uses the same simulation, which starts in January 1997 (March 1999 for GNDQ).

Joining the simulation to real prices

The simulated line is scaled to end exactly at the fund's first real price, so the chart is one continuous line with no jump on launch day.

How close has it been?

This table compares the simulation with real returns over the years each fund has traded, as at September 2026. Annual return is the average growth per year, with growth compounding. Drawdown is the largest drop from a high point to a later low. For the geared funds, this check starts the loan at 35% LVR on launch day, close to where the real funds began (GHHF 34.6%, GNDQ 34.8% soon after launch).

For the five funds with 7 to 9 years of real prices, the simulation came within 0.4% a year of the real annual return, and within 0.6 percentage points of the real drawdown. For geared GHHF it came in about 1.4% a year below. Most of that gap is the ingredients, not the borrowing. Over those years the Betashares ETFs GHHF holds returned about 0.8% a year more than the Vanguard ETFs the simulation uses, and gearing magnifies the difference. Run on GHHF's own ETFs, the simulation came within 0.2% a year. GNDQ holds NDQ, the same ETF its simulation uses, so its row checks the borrowing alone: it came within 0.2% a year. Over so short a time, where the loan starts matters too. Carried on from 1999, the simulation reaches GNDQ's launch at 31.3% LVR and comes in about 3% a year below. For VDAL it came in about 2% a year below, but over only 1.5 years, which is too short to judge. All of these funds launched after 2013, so this table only checks the ASX ETFs. The table above checks the stand-ins used before them.

Limits of the simulation

• It copies only the mix. VanEck's all-in-one funds, for example, choose and weight their investments in their own way, which the simulation doesn't follow.
• It uses each fund's mix today, even if the fund held a different mix in the past.
• Ethical and ESG (environmental, social and governance) funds leave out companies that fail their screens, and some favour companies with lower carbon emissions. VDIF picks most of its shares for high dividends. The ingredients are ordinary index ETFs, so these funds' simulated figures are a rougher estimate, marked with two asterisks (**).
• EXUS leaves out the US and Australia, and no ETF has done that for long. Its stand-in, IVE, includes Australian companies (about 7% of IVE) and leaves out Canada (about 14% of EXUS). EXUS's simulated figures are marked with two asterisks (**).
• Before September 2005, DGCE, DFGH and AVNG use the same index funds as VGS and VGAD, so their fall from 2000 to 2003 follows the index. The fund's fee also comes off the unlisted unit prices of DACE, DGCE and DFGH again, so those years come out about 0.3% a year low.
• Avantis's funds have traded for less than a year, too short to test their stand-ins against. Their simulated figures are marked with two asterisks (**) until October 2028.
• Before 2013, Australian shares come in about 1% a year below the Australian share market. EWA holds only large companies, charges its own fee, and as an overseas fund loses some of its dividends to tax.
• Before 2016, hedged global bonds are US bonds only, and there is no ETF to check that stretch against. This matters most for funds with a lot of global bonds, such as VDCO (42%) and VDBA (35%).
• From 1996 to 2006, Australian bonds come in about 0.5% a year above the bond market. The calculation assumes longer-term bonds than the market held then.
• Before the ETFs, global shares have no Canadian fund (Canada's part goes to the US fund), and the Pacific fund includes some Australian companies.
• Mixes, fees and gearing levels are entered by hand from issuer documents. They were last checked in September 2026.
• It is an estimate for comparing funds, not a record of what any fund returned. Past returns, real or simulated, don't tell you what will happen next.

Where the data comes from

• Fund prices are daily closing prices from Yahoo Finance, adjusted so that distributions are reinvested.
• The 10 year Australian Government bond yield and the 3 month bank bill rate are from the Reserve Bank of Australia (tables F1.1 and F2), used under CC BY 4.0.
• The US dollar to Australian dollar exchange rate and the 3 month US Treasury bill rate are from the Board of Governors of the Federal Reserve System, retrieved from FRED (Federal Reserve Bank of St. Louis).
• Gold prices before November 2004 are daily closing prices of COMEX gold futures from Stooq.

This page explains a method. It is general information only, not financial advice.