DAA-G12 (Defensive Asset Allocation)
Two assets watch the market, emerging markets (VWO) and aggregate bonds (BND), and their momentum decides whether the whole portfolio is in six risk assets, half and half, or in cash. The chosen risk assets pass no filter of their own.
What DAA-G12 is
DAA stands for Defensive Asset Allocation. It was published by Wouter Keller and Jan Willem Keuning in August 2018 in Breadth Momentum and the Canary Universe: Defensive Asset Allocation (DAA) (SSRN 3212862); this page follows the December 2018 revision. "G12" again means a global universe of 12 risk assets, the same as VAA-G12.
What is new compared with VAA is that risk assets are not counted to decide how much cash is needed. A separate universe of just two assets is watched, the so-called canary universe (named after the canary that warned miners of gas): VWO, emerging markets, and BND, US aggregate bonds. While both have positive momentum, the portfolio is all risk; with one negative, half goes to cash; with both, all of it.
Quick facts
- Author
- Wouter J. Keller and Jan Willem Keuning
- Year
- 2018
- Origin
- Paper
- Access
- Public and free
- Type
- Tactical
- Assets
- 15
- Frequency
- Monthly
- Difficulty
- Medium
Who it is for
Anyone who wants to study the canary-universe idea and a portfolio diversified across six positions when the market is doing well, with complete and public rules.
Who it is not for
Anyone who does not want just two signal assets to decide half or all of the portfolio, or who is looking for few trades a year.
Who it is for and who it is not for
It is a good read for understanding why a pair of assets can serve as an alarm. Unlike VAA-G4, it holds a portfolio spread across six positions when all is well, which softens the effect of one bad asset. In exchange, a regime change can mean several purchases and sales in the same month. It does not suit anyone who needs few trades a year.
Current signal
DAA-G12 (Defensive Asset Allocation)
Calculated with the closes of September 30, 2026.
What changes from last month
Rules, step by step
At the close of the last business day of each month:
- 13612W momentum of the 12 risk assets, the 3 cash assets and the 2 canaries (VWO and BND): 12 times the 1-month return, plus 4 times the 3-month return, plus 2 times the 6-month return, plus the 12-month return. The paper averages it; for the sign and the ranking it makes no difference.
- b = number of canaries with 13612W less than or equal to zero (0, 1 or 2). With B = 2, the cash fraction is b/2.
- b = 0: the 6 risk assets with the best 13612W, at 1/6 each.
- b = 1: the 3 best risk assets, at 1/6 each (50% in total), and the remaining 50% in the best of SHY, IEF and LQD by their 13612W.
- b = 2: 100% in the best of SHY, IEF and LQD, even if its momentum is negative.
- The chosen risk assets pass no absolute filter of their own: an asset can enter the portfolio with negative momentum if it is among the best. Protection comes only from the canaries.
The engine uses these US ETFs. Risk assets: SPY, IWM, QQQ, VGK, EWJ, VWO, VNQ, GSG, GLD, TLT, HYG and LQD. Canaries: VWO and BND. Cash: SHY, IEF and LQD. VWO is both a canary and a risk asset, and LQD is both risk and cash.
Decisions made by this site
- Tie-breaks. The paper does not say what to do if two assets have the same 13612W. The engine applies no rule of its own; with real prices an exact tie is very unlikely.
- Variant not implemented. The paper also presents DAA1-G12 (B = 1, T = 2, cash with SHV, IEF and UST; UST is a 2x leveraged ETF on 7-10 year Treasury bonds), which reports 22.3% a year and a -12.9% maximum drawdown between 1970-12 and 2018-03 (text next to Fig. 21, p. 20). It is not the one we implement: it depends on a leveraged ETF.
- 0.1% costs per trade, as in the paper.
Backtest and out-of-sample period
With real ETFs from Jun 2008 to Sep 2026, DAA-G12 returns 7.6% a year, with volatility of 8.5%, a maximum monthly drawdown of -19.2% and a Sharpe of 0.75. The average fraction of the portfolio in cash is 33% and annual turnover is 604%.
Since the paper's publication (Aug 2018), it returns 5.9% a year with a maximum drawdown of -19.2%.
What the authors publish
The paper gives few figures in text (those for the full sample are in an image and we do not transcribe them). Those that can be read, with 0.1% costs and the authors' proxies:
| Series | Period | Annual return | Worst drawdown | Source |
|---|---|---|---|---|
| DAA-G12, out of sample | 1993-12 to 2018-03 | 12.9% | -9.0% | Text, p. 10 |
| DAA-G12, last 10 years | to 2018-03 | 9.9% | n/a | Text, p. 10 |
| Equal-weighted 12 assets, no protection | 1993-12 to 2018-03 | 7.7% | -39.2% | Text, p. 10 |
Over that same stretch, according to the authors, DAA-G12 holds on average 28.8% of the portfolio in cash, against 60.8% for VAA-G12. "Out of sample" has the authors' meaning here (parameters optimised until 1993, validated since), different from "since publication", which is what our engine measures. The figures are not from the same period or the same data.
5.9% a year · max drawdown -19.2% · S&P 500 over the same period: 14.7%
An "out-of-sample" result: the period after the rule was published, which could not have been used to tune the rules.
Growth of $100 (logarithmic scale) since , with 0.1% costs per trade. Data: US-listed ETFs, total return (dividends reinvested), before taxes.
Drawdowns from peak
| Strategy | CAGR | Volatility | Max drawdown | Sharpe |
|---|
Year by year
| Year | DAA-G12 (Defensive Asset Allocation) | S&P 500 | 60/40 |
|---|---|---|---|
| 2026 (to Sep 2026) | +6.1% | +12.7% | +5.8% |
| 2025 | +18.5% | +17.7% | +13.8% |
| 2024 | +1.5% | +24.9% | +14.7% |
| 2023 | +5.3% | +26.2% | +17.2% |
| 2022 | -11.0% | -18.2% | -17.0% |
| 2021 | +1.8% | +28.7% | +15.9% |
| 2020 | +24.8% | +18.3% | +15.0% |
| 2019 | +8.5% | +31.2% | +21.9% |
| 2018 | -0.1% | -4.6% | -2.4% |
| 2017 | +14.8% | +21.7% | +14.0% |
| 2016 | +5.9% | +12.0% | +7.6% |
| 2015 | -4.3% | +1.2% | +1.3% |
| 2014 | +5.7% | +13.5% | +11.7% |
| 2013 | +9.4% | +32.3% | +16.9% |
| 2012 | +8.3% | +16.0% | +11.1% |
| 2011 | +5.4% | +1.9% | +7.4% |
| 2010 | +14.7% | +15.1% | +12.8% |
| 2009 | +24.9% | +26.4% | +13.1% |
| 2008 | +5.5% | -36.8% | -14.9% |
The first year may be partial. Returns are net of trading costs.
Taxes on rotating
The results above are before taxes. This strategy re-assesses the portfolio every month: historically it turns over an average of 604% of the portfolio a year and spends 33% of months in defensive mode. Every change is a sale.
In most countries each sale of an ETF in a taxable account realises a gain or loss that may be taxed, so rotating monthly is best done inside a tax-advantaged account (such as an IRA or 401(k) in the US, or an ISA or SIPP in the UK) where one is available. Rules vary by country and by account: check yours, or ask a tax adviser. See our methodology for how the results are calculated.
Criticisms and limits
The VWO and BND canary and the parameter T=6 came out of optimisation processes, although the authors separated sample and validation. For 1926-1970 long series were used rather than ETF prices. It depends on just two signal assets - a false alarm from emerging markets or bonds moves half the portfolio. It trades a lot and the chosen risk assets have no filter of their own.
- The canary is the result of optimisation. VWO and BND were chosen by optimising over 1927-1970 among combinations of SPY, VEA, VWO and BND, and T = 6 was optimised over 1970-1993. The authors separate sample and validation to limit bias, but the parameters are still the fruit of optimisation (pp. 6-9).
- Data that is not ETFs. For 1926-1970 long series and Meb Faber's backtest tool were used (note 13); later years use the authors' proxies. A history of tradable prices only exists for recent years.
- Two signal assets. A false alarm from emerging markets or bonds moves half the portfolio, and a double alarm sends all of it to cash. Our own observation, not the paper's.
- No filter on the chosen assets. With b = 0 the six best are bought even if one has negative momentum. In an environment where everything falls except the canaries, there is no correction.
- Turnover. Our history gives annual turnover of 604%, with each regime change turning into several sales and purchases.
How to run it
- Wait for the close of the last business day of the month. The signal is calculated with that close.
- Check the "Current signal". It tells you how many canaries are negative and which portfolio applies.
- Map each ETF to the fund you actually hold.
- Work out the orders with the rebalancing calculator.
- Trade on the first business day of the following month. The backtest assumes execution at the close.
Similar strategies
Frequently asked questions
What is the canary universe?
A small group of assets, here VWO and BND, whose momentum serves as a warning for the whole portfolio. The authors chose it by optimising among combinations of SPY, VEA, VWO and BND with US data from 1927 to 1970.
What does the portfolio do if only one canary is negative?
It keeps half in the three best risk assets, at one sixth of the portfolio each, and sends the other half to the best of SHY, IEF and LQD.
How does it differ from VAA-G12?
It uses the same 12 risk assets and the same cash, but decides the level of cash by looking only at the two canaries and not at how many of the 12 assets are negative. According to the paper, that reduces time in cash.
Is this investment advice?
No. It is educational information about a published strategy. Neither backtests nor past performance guarantee future results, and you can lose money.
Sources and review date
- Keller and Keuning, "Breadth Momentum and the Canary Universe: Defensive Asset Allocation (DAA)", SSRN 3212862 (revision v1.1, December 2018) · accessed October 10, 2026
Page last reviewed on October 11, 2026; figures and signal are recalculated at every monthly close (latest: September 30, 2026). How we work: editorial policy and methodology.
Educational and informational content. It is not investment advice or a personal recommendation. The results shown are historical simulations (backtests) using US-listed ETFs; past performance does not guarantee future results and you may lose some or all of the capital invested. Real returns will depend on your costs, taxes, currency and execution timing.