Common patterns

Most tactical strategies, however clever they sound, are built from a handful of moves. This page shows each move as a real tree you can open, says how to build it, and says what tends to go wrong with it. Every tree here comes from a starter pattern or from the classic strategies, so you can open the original, read where it comes from, and copy it to your account.

A fixed mix#

Hold set proportions and rebalance back to them. The 60/40 portfolio is the best-known:

Build it with one Weight set to Specified, and a ticker for each holding. Choose a monthly or yearly schedule.

Watch for nothing clever, which is the point. A fixed mix makes no prediction; rebalancing simply sells some of what rose and buys some of what fell. It's the baseline every tactical idea on this page should beat after costs, so it makes a good benchmark. Other mixes worth a look: the Permanent Portfolio and the All Weather portfolio.

A trend filter#

Hold a risky asset while its price is above a long average, and step aside when it falls below. Faber's version checks once a month:

Build it with an If / Else. On the left, the current price of SPY; on the right, SPY's moving average, switched to months so it reads month-end closes. Hold the risky asset in the first branch and a safe one, such as BIL, in the second. Variations compare a short average with a long one (the golden cross), or check daily (200-day moving average).

Watch for whipsaw. In a choppy market the price crosses its average again and again, and the filter sells after a dip and buys back higher. Checking monthly instead of daily, or requiring the signal to confirm for a few days, trades less but reacts later. A trend filter also misses the first days of every rebound, since it waits for the price to recover above its average. It earns its keep in long declines like 2000–2002 and 2008.

Hold the strongest#

Rank a list of assets by recent return and hold the leaders. Papa Bear holds the three strongest of thirteen asset classes, scored by their 3-, 6-, and 12-month returns added together:

Build it with a Rank block: choose the reading and lookback, Add score term for each extra lookback, set Keep to the number of winners, and add the candidates.

Watch for two things. A Rank always holds something, so in a broad crash the "strongest" asset can still fall. And the choice of lookback matters more than it looks: blending several lookbacks, as this one does, makes the result less dependent on any single number.

Strongest, but only if it beats cash#

Dual momentum adds a filter to the Rank: hold the stronger of US and foreign stocks, but only while stocks have beaten Treasury bills over the past year. Otherwise hold bonds. This is Antonacci's Global Equities Momentum:

Build it as a trend filter whose first branch is a Rank. The condition compares the 12-month return of SPY with the 12-month return of BIL, rather than with zero, so stocks must beat cash, not just rise.

Watch for slow signals. Checked once a month on a 12-month return, GEM was still fully in stocks when the market fell in February and March 2020, and lost about a third. A slow filter sidesteps long bear markets, not sudden crashes. The full article shows how it has done since its book came out.

A canary#

Watch a few sensitive assets and turn defensive the moment any of them weakens, even if what you hold looks fine. Keller and Keuning's Vigilant Asset Allocation holds the strongest of four assets only while all four have positive momentum; otherwise it moves to the best of three bond funds:

Build it with an If / Else whose conditions are set to all (or at least some number of them), one condition per canary, and a Rank in each branch. The canary doesn't have to be something you hold: Hybrid Asset Allocation watches TIP, an inflation-protected bond fund, as an early warning.

Watch for false alarms. The more canaries, the more often one of them wobbles, so a canary strategy can spend long stretches on the defensive side while stocks keep rising. Check the allocation chart to see how much of the time it actually held its growth assets.

Balance the risk#

Instead of splitting money equally, give calmer assets more of it so each contributes a similar amount of risk. The Static allocation starter pattern does this with stocks, tech, and gold:

Build it with a Weight set to Inverse vol, or to Minimum variance to account for how the holdings move together, as Adaptive Asset Allocation does after ranking ten asset classes by momentum.

Watch for the bond problem. Risk balancing hands most of the money to whatever has been calmest, usually bonds, and that hurts when calm assets suddenly aren't, as bonds weren't in 2022. Minimum variance can concentrate in one or two holdings; look at the allocation chart before you trust it.

Buy the dip#

Add to an asset after a short, sharp fall, betting on a bounce. The Buy the dip starter pattern holds a risk-balanced core and switches entirely into QQQ when its 10-day RSI drops below 30:

Build it with an If / Else on a short-term reading such as RSI. To hold the position until the bounce has actually happened, rather than only while the dip lasts, use an Enter / Exit block instead: enter when RSI drops below 30, exit when it climbs above 70. The builder guide shows that tree.

Watch for falls that keep falling. Dip buying works in markets that recover quickly and hurts in a long decline, which is why it's often paired with a trend filter, as in the Trend following starter pattern.

Ask the economy#

Use economic data as a second opinion. Jesse Livermore's growth-trend timing obeys a broken trend only when a recession looks likely. This version sells stocks only when the S&P 500 is below its 10-month average and the unemployment rate has risen above its own 12-month average:

Build it with an If / Else set to all: one condition on price, one Economy condition. Tactfolio uses each figure only from the day it was published, as it stood then, so the backtest can't peek at revisions.

Watch for small samples and sudden crashes. There have been only a dozen or so US recessions since the data begins, so a rule that dodged most of them has been tested on very few events. And unemployment rises only after a recession starts, so a crash as fast as 2020's is over before the signal turns. Read the growth-trend timing article for the sources and caveats.

Combine strategies#

Once you have strategies you trust, hold several side by side instead of betting on one. Add a Strategy block for each inside a Weight: your own strategies, or published ones from Discover.

• A monthly strategy inside a daily one keeps trading monthly; a faster one inside a slower one isn't allowed.
• Set the Weight's Split resets to yearly to let each sleeve's share drift with its results and reset once a year.
• Other authors' strategies are pinned to one published version, so your combined results can't change unless you choose an update.

Watch for strategies that are secretly the same. Two trend filters on different stock funds will usually sell on the same days; combining them adds little. Combine ideas that fail at different times.