Fetching historical prices and checking every monthly purchase.
DATA CHECK REQUIRED
The backtest is unavailable.
OCTOBER 2016 — SEPTEMBER 2026
Lag7, over ten years.
The same monthly investment. Three different destinations.
$
per month · 120 purchases
PRICE ONLY
Cash dividends are excluded from all three strategies. These are not dividend-reinvested or total-return results.
THE LONG VIEW
Portfolio value
LIKE-FOR-LIKE
Return came with a cost.
Full-period metrics
Cumulative return = investment gain ÷ total contributions × 100. It is not annualized or time-weighted; dividends are excluded. XIRR measures the annualized return on your dated contributions. TWR removes the effect of deposits. Drawdown is measured on the cash-flow-adjusted TWR index, not account balance.
ACCUMULATION, NOT REBALANCING
What you would own.
Capital in. Value out.
Each holding’s growth, excluding dividends.
Units and quoted prices are split-adjusted equivalents, not original historical share counts. Value = adjusted units × adjusted close. Existing portfolio weights are never reset.
LAG7 PURCHASE FREQUENCY
Which stocks were the laggard?
Full 120-month window
FOLLOW THE DECISION
Every month, an auditable choice.
Signal ranking
Highest average rank wins
THE PURCHASE RECORD
One square. One monthly buy.
Choose a month to inspect
TRUST THE PROCESS, NOT THE HEADLINE
What this test does. And doesn’t.
This is a retrospective simulation of a fixed basket of today’s seven companies. It does not prove that these names could have been identified in advance or that the result will repeat.
Execution rules
120 monthly deposits of $5,000 per portfolio ($600k total).
First trading-day open; final valuation at September 30, 2026 close.
Fractional split-adjusted units. No sales or rebalancing.
All cash dividends are excluded, including SPY dividends.
No taxes, commissions, spreads, slippage or cash interest.
The contrarian signal
Final month-end close versus month-ends 3, 6 and 12 months earlier.
Rank 1 = best; rank 7 = worst. Average the three ranks equally.
Select the highest average rank.
Ties: worse 12-month return, then 6-month, then 3-month return.
Use only information available before the next session’s open.
Backtest timing vs. live timing
This backtest uses final month-end signals. The live Lag7 rule locks its selection one trading session earlier, after the penultimate trading day’s close, so these results are not an exact test of live execution.
Validation ledger
Checks passed
Read the returns correctly
XIRR is the annualized money-weighted return using each actual contribution date. It is not ending wealth divided by total deposits and annualized.
Time-weighted return separately links overnight and intraday returns on contribution days. Deposits do not count as investment gains.
Drawdown uses the daily TWR index and includes the initial index peak. It does not capture intraday lows.
Material limitations
Dividend omission: this is price performance only. A dividend-reinvested comparison can produce different outcomes and rankings.
Hindsight universe: today’s Mag7 are held fixed across the entire period. There is no survivorship-free selection process.
Idealized fills: fractional execution at the published opening price is assumed. Historical data precision and revisions can affect near-ties.
REPRODUCIBLE BY DESIGN
Sources and audit files
Historical split-adjusted prices cover the fixed research period. Downloads include the computed results, monthly selections, purchase ledger, daily values, methodology and calculation code. Raw provider price files are not redistributed.