Rule reference
Holding several things at set weights
Portfolios with no buy or sell signal: fixed weights, buying and holding, and risk parity. Each one with what it does, its settings, and an example to try.
On this page
Risk parity
Holds the calmer assets of a list in equal parts, capped per asset, and scales the whole portfolio down when its ups and downs run above a target. Ask for hierarchical risk parity to weight them by how they move together instead.
- When
- At each month-end, weight every symbol by 1 / its volatility over the last 60 daily returns
- Then
- Hold every symbol at those weights, at most 40% each (anything the cap leaves over stays in cash). Exposure is scaled down whenever portfolio volatility runs above 10% a year.
- How often
- Rebalanced monthly
| Setting | Value |
|---|---|
| Scales down above | 10% a year of ups and downs |
| At most in one holding | 40% |
| How often | Rebalanced every month |
Risk parity portfolio across SPY, TLT and GLD over the last 10 years
Try it: open the Strategy Agent with these words typed in, unsentHierarchical risk parity across SPY, EFA, TLT, GLD and VNQ. Backtest over 10 years.
Reads as: scales down above 10% a year of ups and downs; at most in one holding 40%; weighting hierarchical risk parity; how often Rebalanced every month.
Try it: open the Strategy Agent with these words typed in, unsentTemplates in the community library that use it:
- Risk-parity mode: stocks, bonds, gold, REITs: The agent's risk-parity mode on SPY, TLT, GLD and VNQ: hold the calmer assets, capped at 40% each, with a 10% volatility target.
- Hierarchical risk parity across 5 ETFs: HRP weights across the calmer of SPY, EFA, TLT, GLD and VNQ, capped at 40% each, with a 10% volatility target.
Recorded tests of risk parity rules, beside simply holding, in the evidence explorer.
Explained step by step in the Education Centre.
Fixed weights
Holds set weights, such as most in a stock fund and the rest in a bond fund, and puts them back to those weights on a schedule. Weights that add up to less than the whole leave the rest in cash.
| Setting | Value |
|---|---|
| Weights | SPY 60%, TLT 40% |
| How often | Rebalanced every quarter |
60/40 portfolio of SPY and TLT rebalanced quarterly over the last 10 years
Try it: open the Strategy Agent with these words typed in, unsentEqual-weight portfolio of Apple, Microsoft and Nvidia, rebalanced quarterly, over the last 5 years
Reads as: weights AAPL 33.3%, MSFT 33.3%, NVDA 33.3%; how often Rebalanced every quarter.
Try it: open the Strategy Agent with these words typed in, unsentBuy and hold
Buys on the first day and keeps it to the last. Every result is set beside this, so a plain question like "what if I had bought it" is answered the same way.
Several names bought once in equal parts drift with their prices, because nothing puts them back.
| Setting | Value |
|---|---|
| Weights | SPY 100% |
| How often | Bought on the first day and never rebalanced |
Buy and hold SPY over the last 10 years
Try it: open the Strategy Agent with these words typed in, unsentWhat would have happened if I bought Nvidia 5 years ago
Reads as: weights NVDA 100%; how often Bought on the first day and never rebalanced.
Try it: open the Strategy Agent with these words typed in, unsentRelated
- Rule referenceRule referenceEvery kind of rule the engine runs, with what it does, the settings the agent uses when you name none, and an example to try.
- Rule referenceTiming rules: in or out of one shareRules that decide, each day, whether to hold a share or fund or to sit in cash. Each one with what it does, its settings, and an example to try.
- Rule referenceRanking rules: hold the leaders of a listRules that score a list of shares or funds and hold the ones at the top, swapping as the order changes. Each one with what it does, its settings, and an example to try.
- Strategy patternsTrend followingBe in while prices have been rising and out while they have been falling, judged by moving averages.
These pages describe how Quantis Trade works and what its tests show about the past. They are not forecasts and not advice.