Paul Novell's Tactical Bond Strategy

Dec 2013 – Sep 2026	Strategy	SPY
Annual return (CAGR)	3.0%	13.9%
Worst drawdown	-10.9%	-33.7%
Sharpe ratio	0.53	0.85
Volatility	5.9%	17.1%
Annual return since publication (Jan 2015)	2.4%	13.8%

Paul Novell's Tactical Bond Strategy applies dual momentum to the bond market. Novell published it on his Investing for a Living blog in 2014, adapting Gary Antonacci's bond momentum model, and revised its fund list in 2016. Each month it holds the three bond funds with the best six-month returns, a third each, and moves any of the three with a negative six-month return to Treasury bills.

The idea#

The bond market is not one asset. Short and long Treasuries, inflation-protected bonds, corporate and high-yield bonds, and foreign and emerging-market bonds respond to different forces, such as interest rates, inflation, credit risk, and the dollar, and they take turns leading. Novell's aim was a bond portfolio for any environment: follow whichever segments are being rewarded, and step aside from segments that are falling, as long bonds do when rates rise.

Holding three funds instead of one was his preferred version. In his tests it had the best risk-adjusted results and the smallest drawdowns of the variants he tried.

How it works#

The universe is nine bond ETFs from Novell's revised 2016 list:

US Treasuries: short (SHY), intermediate (IEF), and long (TLT)
US inflation-protected Treasuries (TIP)
US intermediate corporate bonds (VCIT) and high-yield bonds (JNK)
International government bonds (IGOV) and international aggregate bonds (BNDX)
Emerging-market bonds (EMB)

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

Rank the nine by their six-month total return.
Hold the top three, a third each.
Any of the three whose six-month return is negative is replaced by Treasury bills (BIL).

Novell's own tests traded every four weeks at the next day's close, a limit of his software; he said he prefers trading at the month-end close, which is used here.

What the backtest shows#

The test starts at the end of 2013, after BNDX's first six months of trading. Over that period the strategy compounded about 3% a year with low volatility and a worst drawdown of about 11%. As a bond strategy it is not built to match stocks, but it also trailed a 60/40 portfolio clearly, on both return and risk-adjusted return, while losing about half as much at its worst.

Its best years came when bonds trended: it gained about 10% in each of 2014, 2016, and 2019, and 11% in 2020. Its worst stretch was the March 2020 panic, when it lost 11% in nine days as credit and even Treasury funds sold off together; it recovered by August. It held up in 2022, the worst year for bonds in decades, losing about 2% while sitting mostly in Treasury bills. But since then results have been thin, and since its publication at the end of 2014 it has compounded only about 2% a year.

Simulated history cannot take this test further back, because several of its funds, including BNDX and IGOV, have no reference history.

When it struggles#

Sudden bond sell-offs. A six-month signal checked monthly cannot step aside from a shock that lasts days.
Trendless rate markets. When yields move sideways, the leaders keep changing and each switch costs a little.
Low returns everywhere. When every bond segment earns little, rotating among them cannot add much.

Using it on Tactfolio#

The live strategy above runs Novell's revised universe with his positive-return filter. Copy it to add Allocate Smartly's extra hurdle, a six-month return above BIL's, or use it as the bond sleeve of a stock strategy.

Year by year

Year	Strategy	SPY
2026*	-2.0%	14.0%
2025	4.1%	17.7%
2024	-0.3%	24.9%
2023	2.3%	26.2%
2022	-2.4%	-18.2%
2021	-1.1%	28.7%
2020	11.3%	18.3%
2019	9.7%	31.2%
2018	-2.6%	-4.6%
2017	5.1%	21.7%
2016	9.6%	12.0%
2015	-3.8%	1.2%
2014	10.1%	13.5%
2013*	-0.1%	2.9%

* 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

Uses Novell's revised 2016 universe, SHY, IEF, TLT, TIP, VCIT, JNK, IGOV, BNDX, and EMB, with his top-3 version and his filter of a positive six-month return. Allocate Smartly trades LQD and HYG instead of VCIT and JNK, drops IGOV, also requires a return above BIL's, and holds cash instead of SHY.
The six-month return is 126 sessions. BNDX began trading in mid-2013, which limits the backtest; Novell used BWX before then.
Novell's own tests traded every four weeks at the next day's close, a limit of his software; he prefers month-end trading at the same close, which is used here.
Signals and trades use the close of the last trading day of each month, as in the source.

Common questions#

What is Paul Novell's Tactical Bond Strategy?#

It is a monthly bond rotation that holds the three strongest of nine bond ETFs by six-month return, a third each, and replaces any pick with a negative six-month return with Treasury bills.

Which bond ETFs does the Tactical Bond Strategy use?#

SHY, IEF, TLT, TIP, VCIT, JNK, IGOV, BNDX, and EMB, with BIL as cash. Novell's 2014 original used a slightly different list that included municipal bonds.

How is this different from Allocate Smartly's version?#

Allocate Smartly trades LQD and HYG instead of VCIT and JNK, leaves out IGOV, holds cash instead of SHY, and also requires each pick to beat Treasury bills over six months. This version follows Novell's own list and filter.

Does the Tactical Bond Strategy protect against rising rates?#

In 2022 it lost about 2% while long Treasuries and most bond funds fell much further, because its filter had moved almost the whole portfolio to Treasury bills early in the year. Its long-run return since publication has been low. The live record on Tactfolio shows how it is doing now.