Tactical asset allocation (TAA) changes a portfolio's target weights using a rule tied to market data. A rule might hold an equity fund while its trend is positive and switch to a defensive asset when that condition fails. The purpose is to make the decision repeatable. It cannot promise an early exit from every crash.
Strategic and tactical allocation
A strategic portfolio can rebalance regularly while keeping the same long-term target. A tactical portfolio changes the target itself. Neither approach removes investment risk.
| Example | Equity target | Bond target | What changes it? |
|---|---|---|---|
| Classic 60/40 | 60% | 40% | Rebalance back to fixed weights |
| GEM equity signal | 100% | 0% | Equity momentum passes its rule |
| GEM defensive signal | 0% | 100% | Equity momentum fails its rule |
A concrete example: GEM
BestFolio's standard GEM implementation compares the 12-month total return of US equities with T-bills. If US equities beat that hurdle, it holds the stronger of US and international equities. Otherwise it holds aggregate bonds. The rule is evaluated monthly, and the portfolio holds one asset at a time.
That last point matters. Defensive bonds can lose money. The signal can react after a loss has begun, and it can switch out just before a rebound. Other tactical methods use faster momentum, trend filters, volatility weights or several separate sleeves. Their risks differ even when they share the TAA label.
How tactical rules respond to a crisis
No rule predicts a crisis. What a tactical rule can do is react, on a schedule, to measurable changes in prices. Most strategies combine some of 4 mechanisms.
Momentum rules cut or drop an asset when its return over a lookback period, typically 1 to 12 months, turns negative or falls behind cash. Trend filters move a sleeve to cash or bonds when its price drops below a long moving average, such as 10 months. Canary rules watch a small set of bellwether assets as an early warning: Defensive Asset Allocation used emerging market equities and aggregate bonds, and Hybrid Asset Allocation reduces it to a single canary, US TIPS. Volatility scaling shrinks positions as volatility rises, so the portfolio de-risks during turbulent periods.
None of these needs a forecast. Because they respond to data as it arrives, they are often partly or fully defensive by the time a decline is well under way, and they can just as easily switch out right before a rebound. The defensive asset matters as much as the signal. In 2022 stocks and bonds fell together. GEM, ADM and CDM, whose defenses are dominated by bonds, lost between 10 and 24%, while 3 canary strategies (BAA-G4, BAA-G12 and HAA), which can retreat into T-bills when longer bonds are falling too, ended the year positive. The 2022 dual momentum study has the full scoreboard.
What to check in a backtest
- Compare strategies and the passive benchmark on the same start and end dates.
- Read the drawdown alongside the return. A monthly chart can hide deeper losses between month ends.
- Separate observed fund prices from pre-inception proxies and synthetic histories.
- Check how trading costs, distributions, cash returns and execution timing were modeled.
- Look for tests beyond the period used to design the rule. Selecting the best historical variant can overstate what a new investor should expect.
The GEM guide includes a worked comparison of GEM, ADM and CDM on shared dates. The Classic 60/40 page provides a simple baseline.
Make a plan you can follow during a drawdown
A rule only defines the portfolio you actually hold if you keep following it. Before investing, write down the chosen variant, the source of its signals, when you review them, the allowed assets and the costs you can accept. Decide how much loss your overall finances can absorb. A strategy's worst historical loss is not a ceiling on future losses.
After a difficult month, distinguish an implementation problem from an uncomfortable but valid signal. Check whether the data arrived, the rule ran and the trades matched the intended weights. A missed trade or a changed financial need can justify action. Recent underperformance alone does not establish that the rule stopped working.
Use a scheduled review to consider changing the strategy. Record the reason and compare alternatives over consistent windows, including the losing periods. Jumping to the latest winner after every setback makes it hard to tell whether the original rule or the switching decision caused the result.
Building your first tactical portfolio
Set expectations first. Tactical allocation is not about beating the S&P 500 every year: in a strong bull market most tactical strategies lag a fully invested equity portfolio. What they aim to do is make the deepest falls smaller, so the portfolio is one you can actually hold. You also commit to a monthly routine: each month you rebalance, meaning you trade back to whatever weights the rules now set. If you cannot check a signal once a month and place a few trades, a fixed allocation is the better choice.
Choose 2 to 4 strategies
For a first portfolio, look for rules that are fully published, simple to execute (1 to 5 ETFs, one trading day a month) and robust, meaning the results do not hinge on one narrow parameter. 3 well-known starting points use different kinds of momentum:
- GEM: Gary Antonacci's dual momentum, one ETF at a time, 12-month lookback.
- HAA: Wouter Keller and Jan Willem Keuning's Hybrid Asset Allocation, which holds the 4 strongest of a broad offensive universe while its TIPS canary is positive and retreats to Treasuries or cash when it is not.
- ADM: Accelerating Dual Momentum, which scores equities on 1, 3 and 6-month returns and reacts faster than GEM, at the cost of more turnover.
One strategy alone will have stretches of underperformance that test your resolve. 2 to 4 strategies with different signals smooth that out because their bad stretches rarely line up. Beyond that, each extra strategy adds work faster than it adds diversification.
Decide the weights
Equal weight is the simplest choice and the hardest to overfit: with 3 strategies, a third each. Risk weighting gives more capital to the calmer strategies, in proportion to the inverse of their volatility. Conviction weighting leans toward the rules you understand best. Start with equal weights and refine only after living with the portfolio through different markets. If you want the weights to adapt on their own, walk-forward portfolios re-weight strategy sleeves monthly using only past data.
Check your broker and the ETFs
You need an account that can trade every ETF your rules use, at a cost you accept. The 3 starter strategies above use these funds:
| Strategy | Offensive assets | Defensive assets |
|---|---|---|
| GEM | SPY, VEU | AGG (BIL as the hurdle) |
| HAA | SPY, QQQ, IWM, VEA, VWO, VNQ, PDBC, IEF, TLT | IEF, BIL (TIP as the canary) |
| ADM | SPY, SCZ | TLT, TIP |
Investors in Europe often cannot buy US-listed ETFs because of the PRIIPs rules, and need UCITS equivalents for each ticker. Check this before you commit, not on the first trading day.
An example starter blend
A simple blend is 33% GEM, 33% HAA and 34% ADM. With $100,000 that is roughly $33,000 per strategy. Each one tells you which ETFs to hold, so you end up with 3 to 8 positions at a time and trade a few of them at the start of each month. The 3 rules use different momentum methods and partly overlapping universes, which diversifies the approach and not just the assets. This is an illustration of the mechanics, not a recommendation.
What not to do
Do not skip months because markets seem fine, and do not override a signal you dislike: if the rule says bonds, the plan is bonds. Do not watch the portfolio daily either, since these rules are built for monthly decisions. And do not fire a strategy after a bad quarter. Give it 12 to 24 months and judge it at the review you planned, as described above.
Starting with a monthly workflow
- Choose a rule and inspect its holdings and limitations.
- Confirm the instruments, order sizes and costs available in your account.
- Wait for the confirmed signal and its stated execution date.
- Record the trades and check the resulting allocation.
- Review the process at the interval you set in advance.
Keep a simple log of each month's signals, trades and portfolio value; it makes the scheduled review far easier. A signal calculated from a market close cannot be assumed to have been known before that close. The execution-delay study explains why that distinction belongs in the plan.
You can compute signals yourself from price data in a spreadsheet, and many investors did for years, but it is slow and error-prone. BestFolio computes the signals for every strategy in its catalog and shows what each one holds every month; 6 strategies, including GEM, are free. Whatever you use, have one reliable signal source and stick to the process. The moment you start making exceptions, it stops being a rule and becomes guessing.
Educational research, not personalized investment advice. Backtests are hypothetical; past performance does not guarantee future results.