A systematic investment strategy is a rule you can write down and follow without judgment: what to hold, how much, and when to change it. The pages in this section take the strategies that keep reappearing in the academic literature and in practitioner portfolios, and treat each one the same way. What the rule is, why it might work, when it has failed, what a hypothetical backtest on twenty years of daily data shows, and the exact sentence you can hand to ENSEMBLE to build and test it yourself.
Why start with strategies rather than tools?
Most backtesting content starts with software and works backward to an idea. That gets the order wrong, because a backtest is only as useful as the hypothesis behind it. A momentum rule and a fixed-weight allocation are answers to different questions. Momentum asks whether recent relative performance persists. A fixed allocation asks whether diversification alone is enough. If you do not know which question you are testing, the metrics on the tear sheet cannot tell you whether the answer is good.
So each page begins with the idea, states the rule in one paragraph, and only then shows numbers. The numbers come from a real ENSEMBLE model built from a plain-language prompt, and each tear sheet card links to the full public model where you can inspect the equity curve, the drawdowns, the holdings over time, and the generated Python that produced them.
What the strategies have in common
Every strategy here is long-only, uses liquid US-listed ETFs, rebalances on a calendar schedule, and can be described in a sentence or two. Those constraints are deliberate. They keep the strategies implementable by an individual with a brokerage account, and they keep the backtests honest, because there is little room for the optimizer to hide. A rule that needs twelve parameters to look good is usually describing the past rather than a regularity in markets.
The strategies also span the main families of systematic investing:
- Fixed allocation. The 60/40 portfolio and its permanent and all-weather cousins hold set weights and rebalance back to them. The only decision is the weights.
- Risk-based allocation. Risk parity sizes each position by its volatility rather than by a fixed percentage, so that no single asset dominates the portfolio's risk.
- Trend following. The 200-day moving average rule holds an asset when it is above its long-term average and steps aside when it is below.
- Relative momentum. Dual momentum rotates into whichever asset has performed best over the past year, with an absolute-momentum filter that moves to cash when nothing is rising.
Reading them together shows the trade-offs plainly. Fixed allocations are simple and cheap to run but fully exposed to whatever the market does. Trend and momentum rules cut the deepest drawdowns in the historical record but pay for it in whipsaws, turnover, and long stretches of underperforming a buy-and-hold benchmark.
How to read the backtests
Every tear sheet card on these pages is labeled a hypothetical backtest, and the label is doing real work. The figures were computed by ENSEMBLE from daily closing prices, net of a simulated transaction cost, with the rules applied mechanically. They describe what the rule would have done over one specific history, not what it will do next.
Three habits make the numbers more useful than they look at first glance:
- Read drawdown before return. A strategy's compound annual growth rate tells you how it did on average; its maximum drawdown tells you what it would have felt like to hold. Most people abandon strategies in drawdowns, not in flat years.
- Compare to a benchmark in the same window. A 9 percent CAGR is a different achievement in a decade when equities returned 14 percent than in one when they returned 5 percent. The full model page shows the benchmark alongside the strategy.
- Look at the number of trades. Turnover is where real-world costs, taxes, and slippage enter. A rule that trades once a quarter and one that trades daily can have similar backtested returns and very different live results.
Our metrics explained guide goes through CAGR, volatility, Sharpe ratio, and maximum drawdown in detail, and common backtesting mistakes covers the ways a backtest can lie to you.
Testing your own variation
The reason each page carries a prompt is that the most useful thing you can do with a strategy is change one thing about it. Swap EFA for VEA in the dual momentum universe. Move the 60/40 rebalance from monthly to annual. Use a 10-month average instead of 200 days. Each of those is a one-word edit to the prompt, and each produces a new model in under a minute.
When two variations produce different results, the interesting question is whether the difference is a feature of markets or an accident of the sample. The pages point out which parameters are structurally important, such as the absolute-momentum filter in dual momentum, and which are cosmetic, such as the exact lookback length within a broad range.
What this section is not
It is not a ranking. The strategies are ordered by how often people search for them, not by any measure of quality, and a strategy that has done well over the last twenty years has no claim on the next twenty. It is also not a recommendation to hold any of them. ENSEMBLE is research software that builds and backtests the rule you describe; deciding whether to act on a backtest is yours, and for most people that decision belongs in a conversation with a qualified adviser.
What it is, is a place to see the canonical strategies stated precisely, tested on the same data with the same assumptions, and open to modification. That is what most strategy write-ups leave out, and it is the part that turns reading into research.