Simulated history

Many funds are young. VEA, BIL, and DBC all started trading in 2006 or 2007, so a strategy that holds them can only be backtested from then on real prices. That leaves out the inflation of the 1970s, the 1987 crash, and the 2000–2002 bear market. Simulated history extends each young fund backwards with modelled values from before its first real price, so the same rules can be tested over decades instead of years.

Simulated values are estimates, not prices anyone could have traded. Use them to see how a strategy behaves in conditions its real history doesn't cover, and lean on the real-price results when you decide.

Where to turn it on#

• Builder and saved strategies: check Simulated history in the Window menu. The choice is remembered on your device, and a research link carries it as simulated=1.
• Public strategy pages: switch between Real prices and Simulated history above the backtest.
• Classic strategy articles: switch between the ETF-price results and the simulated results above the chart.

Real prices are always the default. Every simulated result lists the tickers it simulated and how.

How a young fund is extended#

Tactfolio keeps the ticker's name and its real prices. Only the days before its first real price are modelled, and the modelled values are scaled to meet the first real price, so there is no jump and every real price and return stays as it was. Each fund uses the first of these that applies:

1. A published simulation of the fund. For about a hundred funds, Tactfolio uses a frozen copy of testfol.io's simulated tickers, which rebuild a fund from the index or asset class it tracks. SPY, for example, reaches back to 1885 this way.
2. An older fund or index with the same exposure. VOO follows SPY, EEM follows Vanguard's emerging-markets index fund, and SH follows the inverse of SPY with daily resets, the fund's fee, and borrowing costs. Each of these was checked against the real fund where both exist: returns must move closely together and grow at nearly the same rate.
3. A stand-in, for asset classes older than any fund. Some asset classes had no fund at all for part of the period, so a related asset stands in for them. Stand-ins follow the real fund less closely than the first two kinds, so the backtest names each one it used.

"Correct direction of momentum" means the stand-in's average 1-, 3-, 6-, and 12-month return had the same sign as the real fund's at a month end, measured over the real fund's whole history. Strategies such as Hybrid Asset Allocation decide on exactly that sign.

Cash and borrowing earn or pay the 13-week Treasury bill rate. Before the rate history Tactfolio stores begins, a frozen copy of the same daily series reaches back to 1960.

Limits#

• Results tend to look better in simulated decades. Hybrid Asset Allocation's simulated history since 1970 shows a higher return and Sharpe ratio than its ETF years since 2008. Part of that is the market, and part is that models and stand-ins are smoother than real funds. Compare the two views before trusting the longer one.
• Old dates are a daily timeline, not exchange records. Before real prices, sessions come from the published simulations' dates.
• Some simulated days have a close only. Published simulations give one value a day, with no open, high, or low. Indicators that need a daily range, and fills at the next open, start at the fund's first real range instead, and the backtest window shrinks to match.
• Fees and leverage are modelled. Leveraged and inverse funds pay borrowing costs and fees at rates checked against the real funds, but a model can't capture every way a real fund drifts.
• Only research uses it. AI assistants over MCP, the strategy index rankings, live records, and Trade always use real prices.

Every simulated ticker#

As served on 2026-09-26. Your backtest names the exact dates for the tickers it uses. "Not stored yet" means Tactfolio has not loaded that ticker's prices; it is fetched the first time a strategy uses it.