> For the complete documentation index, see [llms.txt](https://help.hydravision.ai/llms.txt). Markdown versions of documentation pages are available by appending `.md` to page URLs; this page is available as [Markdown](https://help.hydravision.ai/products/backtest-analysis/strategy-tester-journal-print.md). # Strategy Tester Journal Print How to read the performance summary Hydra Vision prints to the MT5 Journal at the end of a backtest. ## Reading the Journal Summary When a backtest finishes, Hydra Vision prints a full performance summary to the **Strategy Tester → Journal** tab. It is the fastest way to see how a run behaved without opening any other tool. {% hint style="info" %} The summary is printed during shutdown. You will see it after the test completes, at the very bottom of the Journal. Scroll to the end and read upwards. {% endhint %} ### When it appears (and when it doesn't) | Situation | Summary printed? | | ----------------------------------- | ------------------------------------------------------------------ | | Normal backtest finishes | Yes | | You remove the EA from a live chart | Yes | | Terminal closed / chart closed | Yes | | You changed an input (EA restarts) | No — the EA is coming straight back | | The EA failed to start | No — you get a one-line "deinit after failed init" message instead | | Optimisation pass | No | *** ### Header ``` ═══════════════════════════════════════════════════════════════ HYDRA VISION PCL PERFORMANCE SUMMARY License: Fusion Strategy: Dynamic PCL Symbol: EURUSD | Period: 2025.01.01 - 2025.12.31 (364 days) ═══════════════════════════════════════════════════════════════ ``` Confirms **what was actually tested**. Check this first — it is easy to run a different symbol or date range than you intended. `days` is calendar days, not trading days. Every "per day" figure further down uses this number. *** ### Profit Performance | Line | What it means | | ----------------- | ---------------------------------------- | | **Total Profit** | Net result after costs. The bottom line. | | **Gross Profit** | Everything the winners made, added up. | | **Gross Loss** | Everything the losers cost, added up. | | **Profit Factor** | Gross Profit ÷ Gross Loss. | **How to read Profit Factor:** above `1.0` means the winners outweighed the losers. `1.5` means you made £1.50 for every £1.00 lost. Below `1.0` the run lost money. If the run had no losing trades at all it prints `No Losses` rather than a number. {% hint style="warning" %} A high Profit Factor on a grid strategy is normal and is **not** by itself a sign of quality. Grids win often and lose rarely but heavily. Always read it next to the drawdown section below. {% endhint %} *** ### Risk Metrics This is the most important block in the summary, and the one most often misread. There are **six drawdown figures and they measure different things.** #### The two words that matter: Maximal vs Relative * **Maximal** — the biggest drop in *money terms*. * **Relative** — the biggest drop in *percentage terms*. These are usually different moments. A £500 loss on a £10,000 account is a bigger *percentage* hit than a £700 loss on a £20,000 account. Each row shows its own value, its own percentage and its own timestamp, so the numbers on one line always belong together. #### The two words that also matter: Equity vs Balance * **Equity** — includes open, floating positions. This is your real exposure. * **Balance** — closed trades only. For a grid strategy **equity drawdown is the number that matters.** Balance drawdown can look calm while an open basket is deeply underwater. #### The rows
RowRead it as
Equity Maximal DrawdownWorst money drop including open positions. The headline risk number.
Equity Relative DrawdownWorst percentage drop including open positions.
Balance Maximal DrawdownWorst money drop on closed trades only.
Balance Relative DrawdownWorst percentage drop on closed trades only.
Equity Absolute DrawdownHow far equity ever fell below your starting deposit.
Balance Absolute DrawdownSame, measured on closed balance.
Last Capital OpOnly printed if money was deposited or withdrawn during the run: when the last deposit or withdrawal happened, and the total deposited and withdrawn. If you see it, the drawdown rows above were measured on an account whose capital changed mid-run.
Grid Floating DD (worst)The worst unrealised loss carried by a single grid, on whichever side was worse.
Recovery FactorTotal Profit ÷ Equity Maximal Drawdown.
{% hint style="danger" %} **Grid Floating DD is not the same measurement as the rows above it.** The account rows measure your whole account. Grid Floating DD measures one basket of positions. It is labelled separately for exactly this reason — do not compare the two numbers as if they were the same metric. {% endhint %} **How to read Recovery Factor:** how many times over the strategy earned back its own worst drawdown. `3.0` is comfortable. Below `1.0` means the run never made back what it risked at its worst point. #### Compared with MT5's own backtest report The account rows above measure the whole account — in a backtest, that is just this EA. **Total Profit matches MT5's report to the cent.** The drawdown rows are measured slightly differently from MT5's, so expect small gaps: * **Equity drawdown** — Hydra usually reads a little **deeper**, typically by less than one percentage point. Hydra checks equity on every price tick it processes, including the instant after it opens a position, and catches low points MT5's report can miss. Hydra's is the more conservative figure. * **Balance drawdown** — should match MT5's within a few cents. (The JSON report's per-EA `hydra_dd` figures can differ by a little more — see the JSON report guide for why.) {% hint style="info" %} If profit matches but MT5 shows the **deeper** equity drawdown, or the gap is well over a percentage point, that is worth reporting. The JSON report guide has a worked example with real numbers. {% endhint %} *** ### Efficiency | Line | What it means | | ------------------- | ------------------------------------------------ | | **Profit/Day** | Average daily earnings across the whole period. | | **Profit/Grid** | Average earnings per completed grid. | | **Profit/Position** | Average earnings per individual position opened. | Use these to compare runs of **different lengths**. Total Profit rewards a longer test; Profit/Day does not. *** ### Grid Statistics | Line | What it means | | ---------------------- | ------------------------------------------------------------------------------------------------- | | **Grids Completed** | Total grid cycles, split BUY / SELL. | | **Total Positions** | Every position opened across the run. | | **Avg Positions/Grid** | Typical grid size. Close to `1.0` means most grids never needed to build out. | | **Avg Grid Duration** | How long a typical grid stayed open. | | **Max Grid Depth** | The widest a grid ever spread, in market distance (pips/points as appropriate to the instrument). | | **Avg Grid Depth** | The typical spread of a grid. | {% hint style="info" %} **Depth here is distance, not a position count.** "Max Grid Depth 180 pips" means price travelled 180 pips against the grid before it closed. A large gap between Max and Avg tells you the run was usually comfortable but was stretched badly at least once — go and find that event. The JSON report's `grids.max_depth` is different: there, depth is the **number of positions** a grid held. Same word, different measurement. {% endhint %} *** ### PCL Transactions *(Printed for every licence.)* | Line | What it means | | -------------------- | -------------------------------------------------------------- | | **Total PCL Closes** | How many times the Profit Close Logic engine closed something. | | **Total PCL Profit** | Money made by those closes. | | **Avg Profit/PCL** | Average per close. | | **PCL/Grid** | Average PCL closes per grid. | | **PCL/Day** | Average PCL closes per day. | #### By Engine Each engine that actually fired gets a line: ``` By Engine: Dynamic 42 ( 68.9%) GBP 1284.50 Avg: GBP 30.58 depth 4.2 dd 3.1% vol 6.30 hunters 5 By Candidate: Anchor + Worst 18 ( 29.5%) GBP 612.20 Avg: GBP 34.01 depth 3.8 dd 2.7% vol 2.40 hunters 2 Whole Grid 19 ( 31.1%) GBP 503.10 Avg: GBP 26.48 depth 5.6 dd 4.4% vol 7.90 hunters 0 ``` Read the first line as: **fire count**, **share of all PCL closes**, **money contributed**, **average per fire**. The line under it gives the average number of positions in the grid and the average drawdown % at the moment that engine fired, the total volume it closed, and how many Position Hunter setups it created. An engine with a large share but a small average is doing a lot of small work; a small share with a large average is doing occasional heavy lifting. **By Candidate** breaks the Dynamic engine down further, showing which combination of positions it chose to close. Engines and candidates that never fired are omitted entirely — an absent line means "never happened". *** ### Engine and feature blocks Most of the following blocks share the same shape: ``` MULTI-TRAILING ───────────────────────────────────────────────── Closes: 12 Profit: GBP 340.20 Loss: GBP -85.10 Net Contribution: GBP 255.10 (8.4% of total) ``` | Field | What it means | | -------------------- | ---------------------------------------------------------------------------- | | **Closes** | How many times this feature closed positions. | | **Profit** | Total from its winning closes. | | **Loss** | Total from its losing closes. | | **Net Contribution** | Profit minus Loss, and what share of the run's total result that represents. | **Net Contribution is the field to compare across blocks.** It answers "how much of my result came from this feature?" The percentages across these blocks add up to roughly 100% once the PCL engines' share is included — the Journal shows the PCL share in money only (Total PCL Profit), not as a percentage. Two blocks print a different set of lines before their Net Contribution: * **Position Hunter** prints its own counters (see Position Hunter extras below), then **Hunter Profit** and **Hunter Loss** in place of Profit and Loss. * **Injected Trailing SL** prints **Injections** (how many were placed), **Invalidations** (how many turned back into a grid instead), **Succeses** (how many finished on their trailing stop — the spelling is as printed) and a single **Profit** line, which is the net result. It has no Closes or Loss line. The blocks you may see:
BlockWhat it covers
Anchor Trailing SLTrades that never grew into a grid — a single position trailed out.
Position HunterTrailing stop management on a remaining position after a grid close.
Injected Trailing SLGrid continuation after an injection.
Multi-TrailingMultiple positions trailed together.
BounceBounce strategy closes.
FrontierFrontier (hedge) strategy closes.
Risk – Max LossHard stop from your Max Drawdown before auto close (negative) setting.
Risk – Step GapHard stop from your Maximum missing steps gap setting.
#### Position Hunter extras | Line | What it means | | ---------------------- | ---------------------------------------------------------------- | | **Setups Created** | How many times a hunter was armed. | | **SL Placed** | How many got a stop loss attached. | | **Trail Activated** | How many actually began trailing. | | **Completed (via SL)** | How many finished by being stopped out — with a conversion rate. | | **Converted to Grid** | How many turned back into a grid instead. | Conversion rate is `Completed ÷ Setups`. A low rate is not automatically bad — it means most setups turned into grids instead of trailing out. #### Reading the Risk blocks {% hint style="warning" %} The two Risk blocks print a note reminding you: **a hard stop only ever realises a loss earlier than the alternative.** By design it has no winning trades. Judge it on the drawdown it avoided, not on its win rate. A Risk block showing pure losses is the feature working correctly. {% endhint %} *** ### Safety Top-Up Printed whenever the feature is enabled — **including when it never fired**, because "armed but never needed" is a meaningful result.
LineWhat it means
Buffer ConfiguredThe simulated capital you made available.
Trigger ThresholdThe floating loss level that draws on it.
Deposits / Total DepositedHow often it was drawn, and how much.
Withdrawals / Total ReturnedHow often and how much went back.
Peak DeployedThe most that was in use at any one time.
Still DeployedAmount still in use at the end of the test.
Consumed (unreturned)Buffer money that was never paid back. This is a real cost.
Buffer RemainingWhat was left.
{% hint style="danger" %} If deposits occurred, the summary tells you so explicitly: **all profit and drawdown percentages above include injected capital.** They were measured against a capital base that grew mid-run, so they are **not comparable** to a run with top-up disabled. Compare like with like. {% endhint %} *** ### New Position On Close | Line | What it means | | ----------------------- | ---------------------------------------------------------------- | | **Attempts** | Times the EA tried to open a replacement position after a close. | | **Opened** | How many succeeded, with a success rate. | | **Blocked (Too Close)** | Rejected — price too near an existing position. | | **Blocked (Dual Grid)** | Rejected — dual-grid rules. | | **Blocked (Count Low)** | Rejected — not enough positions to justify it. | High block counts are not errors. They show the filters doing their job. ### Backfill | Line | What it means | | ------------------ | ------------------------------------------ | | **Evaluations** | How many times backfill was considered. | | **Conditions Met** | How many times the setup qualified. | | **Signal Blocked** | Qualified, but the entry signal vetoed it. | | **Fired** | Actually placed, with a fire rate. | The gap between **Conditions Met** and **Fired** tells you how much your entry signal is filtering out. *** ### Trailing sections After the main summary, further diagnostic reports print automatically:
ReportWhat it shows
SmartEntry 3.0 PerformanceSignal-by-signal entry quality, both sides. Blocked counts are per checked tick, not per setup: SmartEntry checks on every tick while it waits for an entry, and one tick can count under several vetoes at once.
Signal Co-OccurrenceWhich signals tend to fire together. Useful for preset tuning.
Dormancy ReportHow much of the test the EA spent doing nothing.
Grid Compression AnalysisHow tightly positions bunched inside grids.
Drawdown Breach AnalysisHow often drawdown crossed your configured tiers.
Open PositionsAnything still open when the test ended.
MT5 History ComparisonHydra's own accounting checked against MT5's.
{% hint style="info" %} **MT5 History Comparison** is a self-check. If Hydra's totals and MT5's totals agree, the numbers above are trustworthy. A reported difference is worth investigating before you act on the run. {% endhint %} *** ### A suggested reading order 1. **Header** — did I test what I meant to test? 2. **Equity Maximal Drawdown** — could I have survived this? 3. **Total Profit** and **Recovery Factor** — was I paid enough for that risk? 4. **Avg / Max Grid Depth** — how stretched did it get? 5. **Net Contribution** across the engine blocks — what actually made the money? 6. **MT5 History Comparison** — do the numbers reconcile? {% hint style="warning" %} One backtest is not evidence of profitability. Vary the period, the symbol and the modelling quality before drawing conclusions, and remember that past results do not predict future results. {% endhint %}