Last Updated: 2026-10-26
A recurring purchase, a trend rule and a range rule solve different problems. This reference compares what each rule does, the observations needed to test it and the conditions that can invalidate it. The numbers below are invented teaching examples, not recommended settings or reported performance. For order types and fee-inclusive sizing, start with the beginner reference.
Scroll sideways to see every column. Keyboard: focus the table region and use the arrow keys.
| Approach | Rule to specify before testing | Main failure case | What to record |
|---|---|---|---|
| Scheduled accumulation | Cash amount, dates, asset and stopping conditions | Asset keeps falling or becomes unusable | Contributions, acquired units, fees and ending value |
| Trend rule | Exact indicator, completed-bar timing and execution rule | Repeated false crossings in a sideways period | Every crossing, including rejected and losing signals |
| Range rule | Bounds defined using information available then | Price leaves the range and inventory loses value | Fills, remaining inventory and total account value |
| Partial hedge | Spot quantity, contract type and hedge quantity | Basis, funding, margin or custody failure | Dollar P&L on both legs and collateral separately |
Scheduled accumulation: average price is not profit
Three purchases of $100 at prices $100, $80 and $125 acquire 1, 1.25 and 0.8 units before fees: 3.05 units for $300, average cost about $98.36. At a final price of $90 the position is worth $274.50, a $25.50 loss before costs. Buying more units at the lower price did not protect the final value.
A fixed calendar removes one recurring timing choice. It does not remove the decision to select an asset, continue funding it or stop. Compare contributions on identical dates when evaluating alternatives; a lump sum invested earlier has different time exposure. Do not apply an active trade's stop-based risk percentage to an indefinite accumulation plan without defining its total capital commitment and exit conditions.
Trend rules: specify when a signal exists
For a moving-average crossover, record the price source, interval, window lengths and whether the crossing requires a completed bar. A strategy using the final daily close cannot assume an earlier fill at that same price unless its execution model supports it. A long position means positive price exposure; it does not mean using leverage. A flat state and a short position are different choices with different costs and risks.
Test a sideways period as well as a rising period. Repeated crossings can produce losses and fees. Choosing the best of many window pairs after seeing the data is selection, not an independent result. Freeze the rule before evaluating a separate period, and record all configurations tried. See indicator calculations.
Range rules: include the inventory left behind
Suppose a paper rule buys 10 units at $100. Five sell at $110, producing $50 of realized gain. The other five are marked at $80. Cash from the sale is $550 and the remaining inventory is worth $400: $950 total against $1,000 spent, or -$50 before costs. Reporting only the $50 closed-trade gain hides the losing inventory.
Define the range before the simulated trade. If the price leaves it, specify whether the rule stops placing orders, closes inventory or continues holding. Those are different strategies. A second, wider range adds another exposure; it is not automatically a hedge. Limits can remain unfilled, and a stop can execute beyond its trigger or not fill, depending on its type.
Partial hedges: keep dollars and margin returns separate
At an invented $100,000 price, 0.5 BTC of spot is $50,000. A linear short of 0.1 BTC has $10,000 notional. If both prices fall to $85,000, the spot loses $7,500 and the short gains $1,500 before funding, basis changes and fees. Combined P&L is -$6,000. The short offsets 20% of the spot quantity; it does not eliminate the directional exposure.
If initial short margin were $10,000 / 3, its $1,500 gain would equal 45% of that margin. It remains only 15% of the short notional. Comparing that 45% with the spot's 15% loss mixes denominators. On a price rise the short loses, and its margin can be liquidated even while spot elsewhere gains. Actual contract specifications and maintenance rules matter. This is an exposure example, not instructions to open a futures position.
Compare exits and costs on the whole position
Closing half a position at a gain of 1R and half at 2R produces 1.5R gross, where R is the initial risk assigned to the whole position. It does not produce 2R. Moving a stop to the entry price does not guarantee zero loss after fees or gaps.
Use a common evaluation sheet: rule version; data cutoff; fill assumptions; fees, spread and funding; cash flows; open inventory; worst observed drawdown; missing data; and the benchmark. A positive sample does not establish future profitability, and no fixed number of trades guarantees statistical significance. An unmet data or execution requirement is a reason to leave the result unresolved, not to invent a return estimate.
There is no universal progression from a few months of spot trading to leverage. The useful next step is a reproducible paper test of one precisely defined rule, including its failure cases. The trade-review checklist helps distinguish a rule failure from a missing record or execution problem.
Educational mechanics, not investment advice or a return forecast. Crypto assets can lose their value. AITR may earn commissions from commercial links elsewhere; that does not establish a strategy's quality.