Defense First (Carlson’s defensive tactical model)

May 2008 – Sep 2026	Strategy	SPY
Annual return (CAGR)	9.7%	11.8%
Worst drawdown	-20.5%	-50.7%
Sharpe ratio	0.97	0.67
Volatility	10.0%	19.7%
Annual return since publication (Aug 2025)	26.1%	22.1%

Dec 1971 – Sep 2026	Strategy	SPY
Annual return (CAGR)	14.8%	11.1%
Worst drawdown	-23.5%	-55.2%
Sharpe ratio	1.27	0.70
Volatility	11.4%	17.4%
Annual return since publication (Aug 2025)	26.1%	22.1%

Defense First is a tactical model that Thomas D. Carlson published on SSRN in 2025 as Defense First: A Multi-Asset Tactical Model for Adaptive Downside Protection. Each month it ranks four defensive assets by momentum and holds them at 40%, 30%, 20%, and 10%, but any of them weaker than Treasury bills gives its share to the S&P 500 instead.

The idea#

Most tactical strategies watch stocks and retreat to safe assets when stocks weaken. Carlson turned that around. He watches the defensive assets: long Treasuries, gold, commodities, and the US dollar. When they are trending up, they are both a warning sign and a useful place to be. When they are not beating cash, markets are usually calm, and their share goes to stocks.

The result is a portfolio that is mostly defensive in stressed markets and mostly in stocks in quiet ones, with stock exposure moving in 10% steps from none to all of it.

How it works#

At the close of the last trading day of each month:

Score long-term Treasuries (TLT), gold (GLD), commodities (DBC), the US dollar (UUP), and Treasury bills (BIL) by 13612U momentum: the average of their 1-, 3-, 6-, and 12-month total returns.
Rank the four defensive funds from strongest to weakest and give them 40%, 30%, 20%, and 10% of the portfolio.
Any defensive fund whose momentum is below BIL's gives its share to the S&P 500 (SPY).

There is no test of SPY's own momentum, so the portfolio can be fully in stocks during a decline if all four defensive assets are also weak.

What the backtest shows#

Over the ETF era, from May 2008, Defense First returned about 9.7% a year, a little below the S&P 500's 12% and above a 60/40 mix's 8.4%. Its worst drawdown, about 20%, was well under both, and its risk-adjusted return was the best of the three.

Its worst drawdown ran from July 2008 to July 2009, recovered by December 2009; it lost 7% from late May to the end of 2008 while the S&P 500 lost 34%. It gained 8% in 2022, a year when the dollar and commodities rose, while stocks fell 18%. It trailed well behind stocks in strong years such as 2013, 2017, 2019, and 2023. Its best year was 2025, a strong year for gold, when it rose 32%.

It has done very well in the year or so since publication in mid-2025, returning about 26% at an annual rate. That is a short period, and it included a strong gold market.

With simulated history the test starts at the end of 1971 and returns about 15% a year against 11% for the S&P 500, with a worst drawdown of about 24%, in early 1980. The 1970s lean on gold, which also stands in for the commodity fund before 1980: the strategy gained 41% in 1973, 38% in 1974, and 83% in 1979.

When it struggles#

Stock crashes with weak defensive assets. When bonds, gold, and the dollar all trail bills, it holds stocks and has no trend check on them.
Calm bull markets. Defensive assets can keep mild positive momentum for months, holding back gains.
Currency and commodity noise. The dollar and commodities can trend for reasons unrelated to stock market stress.

Using it on Tactfolio#

The live strategy above runs Carlson's rules with DBC as the commodity fund; Allocate Smartly uses PDBC, which tracks the same index. Copy it to hold Treasury bills in place of UUP, as Allocate Smartly suggests, or to add a trend check on SPY like the one in HAA Simple.

Year by year

Year	Strategy	SPY
2026*	13.7%	14.0%
2025	31.9%	17.7%
2024	14.6%	24.9%
2023	5.3%	26.2%
2022	7.7%	-18.2%
2021	21.0%	28.7%
2020	22.0%	18.3%
2019	10.6%	31.2%
2018	-4.1%	-4.6%
2017	5.5%	21.7%
2016	5.5%	12.0%
2015	-0.5%	1.2%
2014	9.0%	13.5%
2013	16.1%	32.3%
2012	-0.2%	16.0%
2011	12.8%	1.9%
2010	12.2%	15.1%
2009	7.6%	26.4%
2008*	-6.7%	-34.3%

Year	Strategy	SPY
2026*	13.7%	14.0%
2025	31.9%	17.7%
2024	14.6%	24.9%
2023	5.3%	26.2%
2022	7.7%	-18.2%
2021	21.0%	28.7%
2020	22.0%	18.3%
2019	10.6%	31.2%
2018	-4.1%	-4.6%
2017	5.5%	21.7%
2016	5.5%	12.0%
2015	-0.5%	1.2%
2014	9.0%	13.5%
2013	16.1%	32.3%
2012	-0.2%	16.0%
2011	12.8%	1.9%
2010	12.2%	15.1%
2009	7.6%	26.4%
2008	4.1%	-36.8%
2007	16.7%	5.1%
2006	7.7%	15.8%
2005	14.8%	4.8%
2004	12.1%	10.7%
2003	17.9%	28.2%
2002	10.7%	-21.6%
2001	-11.7%	-11.8%
2000	21.5%	-9.7%
1999	14.6%	20.4%
1998	25.9%	28.7%
1997	14.4%	33.5%
1996	19.4%	22.5%
1995	31.3%	38.0%
1994	-4.2%	0.4%
1993	10.1%	9.7%
1992	6.0%	7.6%
1991	20.7%	30.3%
1990	6.0%	-3.2%
1989	26.1%	31.5%
1988	12.6%	16.4%
1987	13.1%	5.1%
1986	18.2%	18.6%
1985	22.5%	31.6%
1984	19.4%	6.2%
1983	1.6%	22.4%
1982	24.1%	21.6%
1981	7.5%	-4.9%
1980	20.4%	32.4%
1979	82.6%	18.3%
1978	17.1%	6.5%
1977	14.1%	-7.2%
1976	15.6%	23.8%
1975	10.5%	37.1%
1974	37.5%	-26.5%
1973	40.8%	-14.7%
1972	36.5%	18.9%
1971*	0.0%	0.0%

* Partial year.

The rules as implemented

This is the exact tree Tactfolio runs, rebalanced monthly with signals and trades at the close. Open it to inspect or copy it.

Sources and caveats

Thomas D. Carlson, Defense First: A Multi-Asset Tactical Model for Adaptive Downside Protection (2025)

Momentum is 13612U, the average of 1-, 3-, 6-, and 12-month total returns, on TLT, GLD, DBC, UUP, and the BIL hurdle, as in the source; SPY takes each failing slot.
DBC holds the same commodity index as PDBC, which Allocate Smartly uses in its test.
Signals and trades use the close of the last trading day of each month, as in the source.

Common questions#

What is the Defense First strategy?#

It is Thomas Carlson's 2025 tactical model. It ranks TLT, GLD, DBC, and UUP by momentum, holds them at 40/30/20/10, and moves the share of any that trails Treasury bills into the S&P 500.

What ETFs does Defense First use?#

TLT, GLD, DBC, and UUP as the defensive assets, BIL as the momentum hurdle, and SPY as the fallback. Allocate Smartly uses PDBC in place of DBC.

How is Defense First different from other tactical strategies?#

Strategies such as GTAA 5 or Bold Asset Allocation read stock momentum to decide when to become defensive. Defense First reads defensive assets' momentum to decide how much to put into stocks.

Does Defense First still work?#

It has been strong since it was published in 2025, but that is a short record in a period when gold rose sharply. The live record on Tactfolio shows how it is doing now.