TAA.For The Masses
Strategy Library
Static benchmarksby Mebane Faber

Static Ivy

The endowment-style five-way split, no timing.

Holding now

Signal from Dec 2025 close
  • VTI20%
  • VEA20%
  • IEF20%
  • VNQ20%
  • DBC20%
  • VTIUS Total Market20.0%
  • VEAIntl Developed20.0%
  • IEF7-10Y Treasuries20.0%
  • VNQUS REITs20.0%
  • DBCCommodities20.0%

Trailing returns

Through Sep 25, 2026 · 3Y+ annualised

1M

-0.5%

YTD

+15.6%

1Y

+18.0%

3Y

+15.0%

5Y

+7.9%

10Y

+8.4%

Growth of $100

Backtest from Mar 1, 2006 to Sep 25, 2026 (20.5 years)

Static Ivy+265%
60/40+397%
S&P 500+765%

Growth of $100, rebased to the start of the selected range. Month-end prices, dividends reinvested, no taxes or trading costs.

Risk & return, full backtest

CAGR

+6.7%

Volatility

13.7%

Max drawdown

-46.5%

Sharpe

0.42

Sortino

0.59

Calmar

0.14

Best year

+23.4%

Worst year

-25.6%

What if you added a trend filter? → GTAA 5 / Timed Ivy

The same weights, but each sleeve steps into T-bills when it closes a month below its 10-month average.

CAGR
+6.7%
+5.0%
Max DD
-46.5%
-15.9%
Sharpe
0.42
0.46

Drawdowns

Percent below the prior peak, vs the S&P 500

Calendar-year returns

Faded bars are partial years · 64% of months positive

Strategy Lab

Stress-test the implementation: trade on the signal day's close or later, split the portfolio into tranches that rebalance on different days (reduces timing luck), or swap a fund for a same-asset-class alternative.

The idea

The five asset classes of The Ivy Portfolio held at 20% each. Comparing it with the timed version isolates exactly what the trend filter adds or costs.

The rules

  1. 120% each: US stocks, developed international, 10-year Treasuries, REITs, commodities.
  2. 2Rebalance annually.

Funds used

  • DBC Commodities
  • IEF 7-10Y Treasuries
  • VEA Intl Developed
  • VNQ US REITs
  • VTI US Total Market

Early history uses proxy funds: EFA, SHY.

Recent allocations

Decided at each month-end, held the following month

  • Dec 2025
  • Dec 2024
  • Dec 2023
  • Dec 2022
  • Dec 2021
  • Dec 2020
  • Dec 2019
  • Dec 2018
  • Dec 2017
  • Dec 2016
  • Dec 2015
  • Dec 2014

More static benchmarks