What a bucket does
| Action | Without a bucket | With a bucket |
|---|---|---|
| Deploying | Each strategy separately | The set in one action |
| Starting | Each bot each morning | The set together |
| Backtesting | One run per strategy | One action, credits per strategy |
| Reviewing | Strategy by strategy | Still strategy by strategy |
How backtest credits work
Sizing the strategies inside one
| Approach | When it fits |
|---|---|
| Equal lots across strategies | A sensible default when you have no strong view |
| Weighted toward the longest track record | When one strategy has far more live history |
| Weighted by margin requirement | When the strategies have very different funding needs |
| One at full size, others at one lot | While you are still building confidence in the newer ones |
What a bucket does not change
- The three-simultaneous-strategy limit — a bucket of five cannot all run at once
- Paper bots still count towards that limit
- The daily broker login, which is per connection rather than per strategy
- Margin, which is required per position regardless of grouping
- That each strategy is evaluated and executed independently
The correlation trap
Why grouping hides it
Building one sensibly
Start from strategies that already work individually
Check what they have in common
Size for the combined worst case
Keep it small enough to backtest
Review each strategy individually
When a bucket helps most
- Several strategies on the same daily cycle. One action instead of several every morning.
- Strategies you always run together. Grouping removes the chance of starting some and forgetting others.
- Testing a set as a whole, where you want to see how they behave over the same period.
The short version
- A bucket groups strategies for deploying, starting and backtesting together
- A bucket backtest costs one credit per active strategy inside it
- It does not raise the three-simultaneous-strategy limit
- Similar strategies on one index are correlated — grouping can hide that
- Review each strategy individually; the aggregate hides which one is working
Frequently asked questions
A group of strategies you can deploy, start and backtest together rather than one at a time. It is an operational convenience rather than a change to how each strategy runs.
One per active strategy inside it. A bucket of four costs four of your ten daily credits, and a failed run is refunded.
No. The limit of three running simultaneously applies regardless of grouping, and paper bots count towards it.
Not automatically. Strategies on the same index with similar exposure are correlated, so a sharp move can hurt all of them at once. Ask what single event would hurt every strategy in the bucket.
Each strategy is tracked individually, which is deliberate — an aggregate figure hides which strategy is carrying the result and which is dragging it.
Keep it small enough to backtest within your daily credits, and group strategies you genuinely always run together rather than everything by default.
Yes, and the thing to check first is whether it has an open position. Removing a strategy does not close what it is holding, so square off before you change the composition.
Start with a free 3-day trial
Build a strategy, backtest it and run it on paper — no broker, no IP and no money needed to try it.
