GuluStock and AutoServa
The GuluStock Backtests Page
The Backtests page in GuluStock is the run journal of every walk-forward backtest of a strategy on Bursa Malaysia data, with its verdict, out-of-sample statistics after trading costs, cross-validation grade, equity curve and individual trades. It is used by clients who want to see the evidence behind each strategy grade.
Open a Backtest Run
- Select Backtests in the left sidebar.
- The Run Journal lists every walk-forward run, newest first (up to 100). If you arrived filtered to one strategy family, press Show Every Run to see all families again.
- Click a row to open its details: equity curve, CPCV splits and trades.
- Press Close in the footer, or the X at the top right, to return.
Verdicts
| Verdict | Meaning |
|---|---|
| SHIP | Passed the backtest gates; still needs the full ladder before it can trade live |
| REVIEW | Parked for more evidence; the normal resting state of active research |
| KILL | Failed the gates; it cannot be admitted |
Key Statistics
The Out-of-Sample columns hold the only statistics that count for a ship or kill decision, because they come from data the strategy was not fitted on.
| Statistic | What It Means |
|---|---|
| OOS Net | Out-of-sample net P&L in RM after the run's broker costs. The out-of-sample slice is the last 40% of the lookback window |
| OOS PF | Profit factor: gross wins divided by gross losses. 1.4 or more is the ship threshold |
| OOS WR | Win rate. A low win rate is fine if winners are much larger than losers |
| OOS DD | Maximum drawdown out of sample |
| Trd | Out-of-sample trades; at least 30 are needed for SHIP, and under 10 forces KILL |
| IS→OOS PF | In-sample against out-of-sample profit factor; a large drop suggests overfitting |
| CPCV | The cross-validation grade, one rung of the ladder, not admission on its own |
| Stop / Trail / Time | How trades ended; the run detail breaks this down further under Exit Reasons |
| Avg Hold | Average days in a trade |
Inside a Run
- Strategy Rules & Justification and Why This Version Exists explain what the strategy does and why it was tested.
- Hypothesis, Lesson Learned and Next Action record what the run taught.
- The trade list shows Stock, Entry and Exit times (MYT), Days, Qty, Entry Px, Exit Px, Buy Amount, Sell Amount, Net P&L (after costs), Ret %, Out (exit reason) and Smp (in-sample or out-of-sample) for every simulated trade.
Frequently Asked Questions
Does a SHIP Verdict Mean the Strategy Trades?
No. It must also clear the full validation ladder. The Strategies page shows which strategies are actually admitted.
Are Backtest Results Included With Costs?
Yes, for current runs: brokerage, clearing, stamp duty and slippage are subtracted. Older runs are flagged where this was not the case.
Does a Good Backtest Mean Future Profit?
No. A backtest is a simulation of the past and does not guarantee future results.