Weekly Rebalancing: How Often Should a Bot Trade?

There is no single best rebalancing frequency. Trading more often lets a strategy react faster but costs more and reacts to more random noise; trading less often is cheaper and calmer but slower to respond. Weekly rebalancing is a common middle ground, and the right choice for any strategy depends mostly on how quickly its signals actually change.

This post explains what rebalancing means for a trading bot, works through the cost arithmetic, and shows why our five strategies use five different schedules.

Two meanings of "rebalancing"

People use the word in two ways, and it helps to separate them.

Restoring weights. A portfolio has target weights, such as 50% in one fund and 50% in another. As prices move, the weights drift. Rebalancing sells some of what has grown and buys some of what has shrunk to get back to the targets.

Re-running the rules. A strategy recalculates its signals and decides what it should hold now. That might mean the same holdings at new weights, or a completely different set of holdings.

For a trading bot the second meaning is usually the important one. The rebalancing frequency is really the decision frequency: how often the strategy asks "what should I own?" and acts on the answer.

What changes when you trade more often

Reaction speed

A strategy that checks daily can respond to a sharp change within a day. A strategy that checks monthly might sit through three or four weeks of a decline before it acts. For strategies built around fast-changing signals, such as volatility spikes or short-term stretch, slow checking defeats the purpose.

Costs

Every trade has costs: commissions in some cases, the bid-ask spread, and the gap between the price a signal saw and the price the order gets. These are small per trade but add up with frequency. AlgolaneBots's paper account takes 0.05% off every fill to reflect this, and fills happen at the next real market open, so the gap between signal and fill is included too.

Here is an illustration of how frequency multiplies a small cost. The numbers are made-up but the arithmetic is simple:

The cost per fill never changed. Only the amount of trading did. A strategy that switches its whole portfolio often has to earn a lot just to cover the cost of moving.

Sensitivity to costs is not theoretical

Our Gold Breakout strategy, which checks gold every hour, shows how much this matters. In a historical backtest from June 2004 to November 2026, at 2% risk per trade, it returned +15.1% a year with a worst fall of 41%, compared with +11.0% a year and a worst fall of 45% for simply holding gold. Its deepest fall, from May 2018 to October 2019, took until March 2023 to recover, 41 months after the bottom. Those are simulated results that nobody earned, and future results will differ. Its caveat is the point of this section: doubling trading costs cuts its yearly result roughly in half. A strategy that trades often lives or dies on its costs.

Noise and whipsaw

Markets move randomly from day to day far more than they trend. A strategy that checks daily sees every random wiggle, and if its rules are sensitive, it will sometimes switch on noise and then switch back. This is called whipsaw, and each round trip pays costs while often selling low and buying high.

Checking less often filters out some noise automatically. A weekly check ignores everything that happened on Tuesday and reversed by Thursday. The trade-off is that it also ignores real changes for a few days.

How to match frequency to the signal

A useful way to think about frequency is to ask how long a signal stays meaningful. Some signals change slowly: a ranking of companies by their 12-month performance does not change much from one day to the next. Checking it daily adds cost without adding information. Other signals are fleeting: a volatility spike can come and go in a week, so checking it monthly would miss it entirely.

A rough guide:

• Slow signals (long-term trends, yearly momentum, market health): monthly checks are often enough.
• Medium signals (weight changes among a fixed list, multi-week trends): weekly checks are a common fit.
• Fast signals (volatility, short-term stretch, breakouts): daily or intraday checks are needed, and costs must be kept low.

This is a guide, not a law, and it is worth testing alternatives. Our post on how to read a backtest explains how to check whether results depend heavily on one schedule choice.

Our five strategies and their schedules

Each strategy's schedule follows from what its signals measure.

You can see each strategy described on the strategies overview.

A closer look at weekly: Steady Growth

Steady Growth is our weekly strategy. It holds seven fixed names: NVDA, GLD, COST, LLY, MCK, FIX and AXON. Each week it works out new weights for the seven, and it can put most of the money in one or two names when its rules favour them.

A weekly schedule suits this design. The list never changes, so the only question each week is how to weight it. Weights based on recent performance and risk move gradually, and a daily check would mostly add trading costs. A monthly check might leave too much in a name that has turned.

In a historical backtest from January 2011 to November 2026, Steady Growth returned +26.5% a year with a worst fall of 20%. Over the same period the S&P 500 (SPY) returned +14.0% a year with a worst fall of 34%, and the Nasdaq 100 (QQQ) returned +18.3% a year with a worst fall of 35%. Its deepest fall, in the Covid crash from February 19 to March 20, 2020, recovered by April 14, 2020, about a month after the bottom. Nobody earned these results, and future results will differ.

The caveat is significant: these seven names were picked because they already did well, which flatters the test. The weekly schedule cannot fix a list chosen with hindsight. It only determines how the list is weighted.

Calendar or threshold?

Bots usually rebalance on a calendar: every day, every Monday, the first trading day of the month. An alternative is threshold rebalancing, where trades only happen when weights drift beyond a set band, such as more than five percentage points from target.

Calendar schedules are easier to reason about and to test, which is one reason most rule-based strategies use them.

Other things that frequency affects

• Taxes. In taxable accounts in many countries, frequent trading can turn long-term gains into short-term ones, which may be taxed differently. Rules vary, so this is worth checking for your own situation.
• Attention. A strategy that trades daily generates a lot of activity to review. Some people find that stressful; others prefer it.
• Execution timing. Strategies that decide after the close and trade at the next open face overnight gaps. More frequent strategies face this more often. Our post on how bots place orders explains how we guard against stale signals.

Key takeaways

• Rebalancing frequency is really decision frequency: how often a bot re-runs its rules and trades.
• Trading more often reacts faster but multiplies costs and reacts to more noise; a 0.05% fill cost can become several percent a year at high turnover.
• The right frequency depends on how long a strategy's signals stay meaningful, from hourly breakouts to monthly market health.
• Weekly suits strategies that adjust weights within a fixed list, like Steady Growth, but cannot fix a list chosen with hindsight.
• Cost sensitivity is real: in the Gold Breakout backtest, doubling trading costs cuts the yearly result roughly in half.

This is general education, not personal advice.

See each schedule play out

Watching strategies with different schedules side by side shows the trade-offs better than any table. AlgolaneBots runs all five on a $10,000 paper account, with real-price fills and costs included, during a free 90-day trial with no card needed. Start your paper account and compare how often each one trades.