How to use FXAbsolute

Chapter 01

What FXAbsolute is, and what it is not

FXAbsolute is a browser-based manual market replay and trade-journaling workspace. It lets you make decisions against historical candles while bars to the right remain hidden, then records the simulated trades and notes produced by that process.

The central idea: practise one decision at a time

A normal historical chart exposes the entire outcome. Even if you promise yourself not to look right, the next swing, breakout, and reversal are already visible. FXAbsolute stops the display at the replay cursor. You can advance one candle, ten candles, fifty candles, or play forward automatically. That creates the information boundary a real decision requires: at each moment, you may use what has printed and nothing later.

The app combines four layers. Historical OHLC candles provide the market sequence. The replay controls decide how much of that sequence is visible. Trading and drawing controls capture what you would have done. The journal and statistics turn those actions into a reviewable record.

What the simulator does not reproduce

FXAbsolute is not a broker, exchange, signal service, or promise of profitability. It does not place live orders. Its simulator does not model spread, commission, swap, slippage, liquidity, margin, partial fills, order-book priority, broker symbol specifications, or tax. It evaluates candle ranges rather than tick-by-tick paths. Those omissions matter, especially for scalping or news-event strategies, and Chapter 27 explains exactly how the engine resolves an order.

How to choose your route through this guide

• If you have never opened the app, complete the next chapter first.
• If a control is confusing, use the contents search and jump to that feature.
• If you already know the interface but distrust your results, start at the execution model and testing protocol.
• If your question is about Google sign-in, passwords, Pro, or a completed payment, use Chapters 3 and 35.

Chapter 02

Your first complete backtest in 10 minutes

This route deliberately uses the fewest controls. Its purpose is to complete the loop once: open an unseen point in history, make a plan, execute it, move the market, close the trade, and inspect the record.

1. Open the backtester. On the home page select Try it instantly. The app creates a GBPUSD session with a $10,000 starting balance and one of several curated historical start dates.
2. Confirm the clock. Read the timezone selector in the top bar. If the candle times should match London, choose UTC+0; if they should match Dubai, choose UTC+4. The choice changes labels, date jumps, and session-band hours.
3. Read before moving. Note the selected instrument, timeframe, last visible candle, nearest swing high, nearest swing low, and whether the current market looks directional or balanced.
4. Write a falsifiable plan. Open the Journal tab and record an entry condition, an invalidation level, a target, and a reason to skip. Save with the button or Ctrl+Enter.
5. Place one trade. In the Trade tab set a positive lot size and optional absolute-price SL and TP values, then choose Buy or Sell. A market order uses the current candle close.
6. Advance without changing the story. Use +1 while price is near the decision point. Use +10 only when nothing important can occur between the visible bars. Do not scrub forward to inspect the answer and then return.
7. Finish and review. Let SL or TP trigger on a later bar, or press Close to exit at the current candle close. Read the history row, P&L, pips, win rate, and journal entry together.

What counts as success in this first run

The trade winning is irrelevant. The successful outcome is a complete, auditable sequence in which the note existed before the result, risk was defined, future candles stayed hidden, and the recorded outcome matches the rule. That workflow is the unit you will repeat hundreds of times; learning it matters more than the first P&L number.

Chapter 03

Accounts, Google sign-in, email sign-in, and passwords

You can begin practising before signing in. FXAbsolute asks for a free account when you start your second session, not on a timer. Google and email/password are separate sign-in methods.

If you chose Continue with Google

You do not create an FXAbsolute password. Google authenticates you and returns the account identity to FXAbsolute. On every later visit, choose Continue with Google again and select the same Google account. The “Forgot password?” flow is not for a Google-created account because FXAbsolute never received or stored a password for it.

1. Press Sign in, or wait until the account prompt appears.
2. Choose Continue with Google. The mobile-number field shown before Google is optional; leave it blank if you do not want to provide one.
3. Complete Google's account picker and consent screen.
4. After the redirect, confirm that your avatar or initials appear in the top-right corner.

If you chose email and password

Enter a valid email and a password of at least eight characters. The optional mobile field may be left blank. If the site says an account already exists, switch from Create account to Sign in. If you cannot remember an email-account password, choose Forgot password, enter the email, and use the link sent to that inbox.

What an account saves—and what it does not

Your authentication and Pro entitlement are server-side. Your backtest sessions, trades, journal notes, and drawings are stored in IndexedDB inside the browser on the current device. Signing into the same account on a second computer does not copy the first computer's sessions. Clearing site data, using an environment that discards storage, or switching browser profiles can remove or hide that local history. Chapter 31 explains this boundary in detail.

Chapter 04

Know what every site-level control opens

The top bar remains available across the main application views. It is the fastest way to move between the home page, the tracker preview, an active backtest, timezone settings, notifications, and your profile.

Notifications

A signed-in account can show a notification control beside the avatar. A badge indicates unread support replies or new community activity. Opening the relevant profile tab marks what is on screen as read. The app checks support unread state periodically while the page is visible, so a reply may appear without a full reload.

Footer controls

The home-page footer includes this User Guide, Privacy Policy, Terms and Conditions, Contact Us, the direct email address, and a feedback control labelled “Send us whatever you feel about our product.” Signed-out visitors can send a message, but they need to include contact information inside the message if they want a reply. Signed-in visitors can continue a support thread in Profile → Messages.

Browser Back and session URLs

The app gives the home, profile, tracker, legal pages, and backtests real paths. A backtest URL can include a local session identifier such as ?s=12. That identifier restores a row from this browser's IndexedDB; it is not a public share link. Sending it to another person will not send the session data.

Chapter 05

Create a session that answers a specific question

A session is the container for replay position, trades, drawings, and journal notes. Good configuration prevents two common errors: testing the wrong market conditions and measuring returns against an arbitrary balance.

Instrument

Open the instrument selector, choose a category on the left, then choose a symbol on the right. Each row displays the current candle count used for the default starting resolution. Select the market named in your written test rule; do not swap to a more active market simply because the first chart looks slow.

Starting Balance (USD)

The form accepts a minimum of $100 and starts at $10,000. Balance is the baseline used by the session statistics and the position tool's percentage-risk estimate. It is simulated bookkeeping, not a deposit and not an instruction to use that amount live. Keep the same balance across comparable runs so a $ result has the same meaning.

Start Date

The date is optional. Leaving it blank starts after a warm-up portion near the beginning of the available series. Selecting a date moves to the nearest available candle after enough prior bars have been retained for indicators. The date limits update for the selected instrument.

Try it instantly

This button skips the form. It creates a $10,000 GBPUSD session at one of several fixed dates spread across 2023 and 2024. Use it to learn controls, not to build a formal random sample: the date pool is intentionally small and curated, so repeated instant sessions are not statistically independent.

Recent sessions

The home page shows the three newest sessions. “Start” means the saved cursor has not advanced; “Resume” shows how many base bars the cursor has moved. The full signed-in list lives under Profile → Your trading. A session automatically remembers its most recent market, replay timestamp, timeframe, and speed.

Chapter 06

Understand the 53 instruments and their data depth

FXAbsolute currently groups the selector by what an instrument is—Forex, Indices, Commodity, Stocks, or Crypto. Data depth and available timeframe are properties of the selected symbol, not universal promises.

Why the candle count changes by market

A count represents actual stored bars at the default loaded resolution. It is not calendar minutes. FX closes for weekends; equity exchanges close overnight and on non-trading days; illiquid shares may have intervals with no recorded trade. FXAbsolute does not invent flat candles to fill those gaps. Consequently, two markets covering similar dates can have very different counts.

Core files versus UAE tiers

The 15 core markets load a minute file and construct 3-minute through weekly candles in the browser. UAE instruments are different: their data source supplies different history depths for 1-minute, 5-minute, hourly, and daily intervals. The app chooses the coarsest suitable source tier for the timeframe. A 1-minute UAE view may cover roughly a month while a daily view of the same share can reach back many years.

Names, quote precision, and pip units

The selector shows full company names for UAE instruments. FX pairs normally display five decimals, JPY pairs three, gold and core indices two, UAE shares three, UAE indices two, and UAEETF two. Display precision is not the same as a broker contract specification; Chapter 28 documents the simulator's pip and P&L conventions.

Chapter 07

Start dates, warm-up history, and chart timezone

Dates in a replay are only meaningful when you know which clock they use and how much prior history the indicator calculations receive. FXAbsolute handles both explicitly.

Why the session does not start at bar zero

An EMA 200 needs at least 200 earlier bars before its first meaningful value. A session opened at the first stored candle would show no long average and no context. For minute-based core markets, the launcher keeps 10,080 base bars—one week of minutes—as warm-up. For tiered UAE markets, it keeps at least 220 bars where possible, usually about 15% of the loaded file while preserving room to replay.

Warm-up candles are historical context, not sample trades. If your requested start date lies too early, the app moves the actual cursor forward enough to preserve warm-up. Record the timestamp shown after the session opens rather than assuming it exactly equals the typed date.

How the timezone selector behaves

On a first visit, the app detects the browser's current UTC offset and snaps to the nearest supported option. An explicit selection is stored locally and wins on future visits. The list spans UTC-12 to UTC+14 and includes UTC+5:30 and UTC-3:30. It uses fixed offsets, not named-city daylight-saving rules, for the general chart clock.

The chosen offset changes displayed candle time, the meaning of midnight in the start-date and Jump controls, and the hours printed in the Sessions legend. It does not alter the underlying UTC candle sequence, price, or trade order.

A practical clock rule

Choose one clock before a formal sample and keep it for the entire sample. If the rule says “London 08:00,” UTC+0 is easiest outside British summer time, but the Sessions overlay is safer because it calculates London and New York daylight-saving transitions by date. If the rule says “Dubai market open,” select UTC+4 and use the DFM/ADX band.

Chapter 08

Map the backtesting workspace before using it

The session screen has four functional zones. Knowing which zone owns a decision prevents accidental orders on the wrong instrument, especially after opening split view.

The information bar

The instrument pill is a button, not a static label. Timeframe buttons immediately follow it. Split opens a second synchronized chart. Indicators opens the seven available selections and shows how many are active. The time at the right belongs to the current replay candle. Moving the crosshair changes the displayed price to the pointer's vertical level; moving away returns it to the current close. The final counter is “current bar / total bars” for the selected timeframe.

The side panel acts on the focused chart

In a single chart, the Trade tab naturally applies to that chart. In split view, it applies to the pane most recently clicked, and the pair name above the tabs tells you which one. The Journal remains attached to the session and currently timestamps notes from the primary candle. The AI tab sends the primary chart context. Check the pair label before every split-view order.

Collapsing panels

The arrow beside the drawing rail hides or shows tools. The arrow on the side panel hides or shows the ticket. Both preferences are stored locally. Opening split view automatically collapses the side panel to preserve usable chart width; reopen it when you are ready to trade the focused pane.

Chapter 09

Navigate the chart without moving the replay

Panning and zooming change your view of information that is already available. Replay controls reveal or remove candles. Keeping those actions conceptually separate protects the historical sequence.

Pan, zoom, and inspect

Press and drag the chart horizontally to inspect earlier visible candles. Use the mouse wheel or a trackpad/pinch gesture to change scale and spacing. The chart automatically resizes with the browser. The right price scale uses instrument-specific precision, while the bottom time scale follows the selected chart timezone.

Move the pointer over the chart to activate the crosshair. The price in the information bar follows the crosshair's vertical coordinate, which is useful for selecting an exact level before creating a line or order. When the pointer leaves, the readout returns to the current replay candle's close.

Why an old section of history may load as you drag left

For performance, the chart initially draws up to 40,000 candles behind the cursor rather than allocating every earlier minute in a multi-year file. When you pan near the left edge, another older block is loaded. This affects rendering speed only; the replay timestamp, trade data, and indicator sequence are unchanged.

What you cannot pan into

Candles after the replay cursor are not placed on the chart. The empty space to the right is intentional. Drawings can project into that space because their anchors are saved as price and time, but no future OHLC values become visible. A trendline extending right is your projection, not leaked market data.

Resetting a crowded view

If the current candle is off screen after a long pan, a forward or backward replay action rebuilds the visible slice around the cursor and scrolls to the replay edge. Changing timeframe also recentres on the same timestamp. If a drawing is selected, click empty chart space before trying to pan so the gesture moves the chart rather than the drawing.

Chapter 10

Choose a timeframe and understand how its candles are built

Core instruments can expose 1M, 3M, 5M, 15M, 30M, 1H, 2H, 4H, 1D, and 1W. A timeframe changes the unit of observation, not the underlying market sequence.

Clock-aligned aggregation

Higher candles are grouped by clock boundaries, not by blindly counting a fixed number of stored rows. A 1H candle therefore opens on the hour. Gaps remain gaps: the app does not weld Friday evening onto Sunday night or one exchange session onto the next. Weekly candles start on Monday. Open is the first stored price in the bucket, high and low are the extremes, and close is the final stored price.

Switching intervals preserves the moment

The app maps the current candle timestamp into the new interval. It does not multiply a bar number and hope the calendars align. That distinction matters around weekends, overnight closures, and UAE tier changes. The new chart lands on the nearest candle representing the same historical moment.

Availability on tiered UAE instruments

A UAE symbol's buttons are the union of intervals its files can build. SUKOON, for example, may expose only daily-and-higher context because its manifest has a daily tier, while ADXGI can be minute-only in the current data. Even when an interval exists, it becomes disabled if its source tier does not cover the timestamp on screen. Hovering the disabled button explains the earliest available date.

Changing timeframe mid-trade

You may change intervals to inspect context, and open trades remain in the session. However, the engine evaluates exits as replay moves through the loaded base sequence. To keep a formal sample consistent, define which interval authorizes entry and which interval is allowed for management before testing.

Chapter 11

Reveal history with replay, speed, scrub, and date jump

The control strip is the experiment clock. It can move carefully one bar at a time or traverse a long quiet period, but every forward action permanently reveals information to you as the tester.

Step buttons

The three left buttons move back 50, 10, or 1 candle. The centre button plays or pauses. The three right buttons move forward 1, 10, or 50. Any manual step stops automatic playback first. A short forward move appends new bars efficiently; a backward move rebuilds the visible slice because it must remove candles that were previously shown.

Playback speed

Choose 1x, 2x, 5x, 10x, 30x, or 60x. Speed changes stop playback, so press Play again after selecting a new speed. The number controls bars revealed per unit of playback time, not simulated market minutes. At 60x on a daily chart, you are still revealing daily candles—just quickly.

Progress scrubber

The range slider moves directly to another candle and displays percentage progress. It is useful for navigation, demonstrations, and reaching a predeclared sample boundary. It is dangerous inside a blinded test because even a momentary forward scrub exposes the outcome. Once you have seen a later chart, dragging back does not remove that knowledge.

Jump date

Choose a date within the loaded interval's actual range and press the return-arrow button or Enter. The app interprets midnight in the selected chart timezone and chooses the nearest available candle. On a market that is closed at midnight, the nearest printed bar may be at the previous close or next open.

Autosave and background tabs

The replay position is saved approximately once per second with a trailing write, and flushed when the tab becomes hidden or closes. Playback stops behind an account wall. If a browser suspends a background tab, returning does not fast-forward through the entire elapsed wall-clock gap; catch-up is capped to protect the sequence and interface.

Chapter 12

Switch instruments inside one session

Select the instrument pill in the information bar to replace the primary market while preserving the session, historical moment, and nearest compatible timeframe.

What stays and what changes

The account, simulated balance, session ID, trade history, and historical timestamp stay. The candle series, price precision, available timeframe buttons, and visible drawings change to the selected market. Drawings are tagged by instrument, so a EURUSD resistance zone does not appear on XAUUSD. When you return to EURUSD, its drawings return.

Time is the join key. If GBPUSD is showing 15 March 2024 at 10:00 and you choose XAUUSD, the loader finds XAUUSD near that time. It does not use the same numerical row index, which would represent a different date because markets have different gaps and counts.

Open positions on more than one market

A session can hold one visible open position per market through the normal ticket flow, so GBPUSD and XAUUSD may both be open. Floating P&L for each position is calculated from its own market's cached price at the replay timestamp. The trade panel labels positions from another market with their symbol.

Use switching for confirmation, not rule-shopping

A multi-market rule can be valid—for example, only trade EURUSD when a predeclared dollar index proxy agrees—but define it first. Switching between many markets until one “looks good” changes the hypothesis after seeing the evidence. Record every required confirmation symbol and timeframe in the journal before the sample begins.

Chapter 13

Use split view without trading the wrong pane

Split creates a second chart with its own instrument, timeframe, and drawings. Both panes follow one master replay clock, so one Play button advances the historical moment across both.

Open and configure the second pane

1. Press Split in the information bar. The side panel collapses automatically to make room.
2. Select a different symbol from the second pane's instrument control.
3. Choose the second pane's timeframe. It maps the master timestamp to its nearest candle.
4. Click the pane you intend to annotate or trade. A focus outline identifies the active pane.
5. Reopen the side panel and verify the pair label above Trade, Journal, and AI.

Focus determines where actions land

A drawing tool is global, but the drawing is saved to the focused pane's market. The ticket's Buy, Sell, and limit draft also use the focused market. Changing focus cancels a half-built limit draft because its price belongs to the prior symbol. Closing split resets focus to the primary pane.

One clock does not mean identical bars

Two markets can have different trading hours and gaps. The second pane finds its nearest bar to the master timestamp; it does not invent a bar where the market was closed. A Dubai share and BTCUSD can therefore show very different local structures at the same UTC moment.

When not to use split

On a narrow phone screen, two panes can make prices and handles too small for precise work. Split can also invite confirmation bias if the second chart has no role in the protocol. Use it when the second pane answers a named question; close it when it only adds visual activity.

Chapter 14

Use indicators as defined measurements, not decorations

The Indicators menu offers EMA 20, EMA 50, EMA 200, Bollinger, Sessions, RSI 14, and MACD. Each selection has a fixed calculation so two runs remain comparable.

No future leakage in the values

Indicator arrays may be computed and cached efficiently, but the value at bar i depends only on closes at or before bar i. The chart slices each series to the replay cursor. Turning an indicator on does not reveal values from future candles.

Warm-up and “needs bars” messages

EMA 20 and Bollinger need 20 bars; EMA 50 needs 50; EMA 200 needs 200; RSI begins after 14 changes; MACD needs enough history for its 26-period slow average and 9-period signal. If the current point has too little prior history, the menu says how many bars are needed instead of silently drawing nothing. A weekly file with fewer than 200 earlier weeks cannot show an EMA 200 at that point.

Selections persist

The active indicator IDs are saved in local browser storage. Opening another session can therefore retain the same set. This is convenient for a stable template but can surprise a tester who expects a clean chart. Audit the indicator count in the menu before every formal run.

Protocol rule

If an indicator participates in entry, exit, or filtering, state the parameter and interpretation before the sample. “RSI confirmation” is vague. “At the close of the M15 signal candle, RSI 14 must be above 50 and rising versus the prior close” can be applied by another tester.

Chapter 15

Read EMA 20, EMA 50, and EMA 200 correctly

The three EMA overlays apply the same formula to different lookback lengths. A shorter EMA responds faster to recent closes; a longer EMA changes more slowly and requires more history.

Exact implementation

Each EMA is seeded with the simple average of its first N closes. Thereafter, the new value equals the current close multiplied by 2 / (N + 1) plus the prior EMA multiplied by the remaining weight. The displayed colours are blue for EMA 20, yellow for EMA 50, and purple for EMA 200.

Three valid ways to define their use

• Location filter: accept long setups only when the signal candle closes above EMA 200.
• Alignment filter: accept long setups only when EMA 20 > EMA 50 > EMA 200 at candle close.
• Dynamic-event rule: require price to touch or cross a named EMA and then close back on the qualifying side.

These are examples of testable definitions, not recommendations. Choose one interpretation. Do not count “near the EMA” in one trade and require a literal touch in the next.

Common interpretation errors

An EMA crossing after a large candle does not mean the crossover existed before the close. Record decisions at the same evaluation point—normally candle close. Avoid changing the timeframe until an attractive alignment appears. Remember that EMA 200 on M5 measures 1,000 minutes of bars, while EMA 200 on 1D measures roughly 200 trading days; the label is the same but the horizon is not.

Chapter 16

Use the Bollinger overlay as a volatility measurement

One Bollinger selection draws three lines: a 20-period simple moving average and upper/lower bands two population standard deviations from that average.

What changes the band width

The middle line follows the average of the latest 20 closes. The standard deviation measures how dispersed those closes are. When recent closes cluster, the bands contract; when they spread out, the bands widen. The calculation uses closes, not candle high-low range or volume.

Turn a visual idea into a rule

“The bands look tight” cannot be independently reproduced. A practical manual definition might compare current band width with the prior 20 bars, or require a close outside a band followed by a close back inside. If FXAbsolute does not display the numeric width you need, use the visual only for annotation and record an objective proxy you can observe consistently.

What a band touch does not prove

It does not by itself mean price is statistically due to reverse. In a trend, closes can remain near one band for many bars. Likewise, contraction does not specify breakout direction. The overlay describes recent closing-price dispersion; your protocol must supply the trade logic.

Chapter 17

Use RSI and MACD in synchronized lower panes

RSI and MACD have scales that do not belong on the price axis, so each opens a separate pane beneath the main chart. Its horizontal range follows the primary chart automatically.

RSI 14

FXAbsolute uses Wilder's smoothing over 14 changes. The first value appears after the initial 14-bar gain/loss window. Dashed references mark 70 and 30. RSI reaches 100 when the smoothed loss is zero. The pane disables independent scrolling and scaling so it cannot drift away from the candles it describes.

Define whether RSI is a level, direction, or divergence condition. For example: “RSI must close above 50,” “RSI must cross from below 30 to above 30,” or “price makes a lower low while RSI makes a higher low between two premarked swings.” Those are different hypotheses and should not be mixed under one label.

MACD 12/26/9

The blue MACD line is EMA 12 minus EMA 26. The yellow signal is a 9-period EMA of defined MACD values. The histogram is MACD minus signal, green at or above zero and red below. The signal begins only after enough values exist, so a newly opened high timeframe can have a blank warm-up region.

Close a pane

Use the X in the RSI or MACD pane title. This toggles the same indicator selection off and updates the persistent indicator set. If both are active, two synchronized panes appear and reduce the main chart height; hide what the protocol does not need.

Chapter 18

Read Tokyo, London, New York, and UAE session bands

The Sessions selection shades trading windows behind price and prints a legend in the current chart timezone. It uses actual named-city time rules for daylight-saving transitions.

Forex and core-market bands

The overlay defines Tokyo as 09:00–18:00 Asia/Tokyo, London as 08:00–17:00 Europe/London, and New York as 08:00–17:00 America/New_York. London and New York change UTC offset on their own daylight-saving schedules, which start and end on different dates. The overlay resolves each date separately rather than using one fixed UTC window all year.

UAE band

For UAE instruments, the overlay replaces the three global bands with DFM / ADX at 10:00–15:00 Asia/Dubai. Dubai does not use seasonal clock changes. The current data is gapped outside printed market activity; the band supplies the expected regular session context without inventing candles.

Legend and chart timezone

The legend converts each session's actual UTC window into the selected fixed chart offset for the day near the right edge. If the chart is UTC+5:30, London hours appear in that clock. Changing the top-bar timezone redraws the labels; it does not change the named exchange's local schedule.

Rendering limit

When the visible chart covers more than roughly 46 days, dozens of narrow bands would obscure price, so the overlay does not draw every band. Zoom into an intraday span to use it. Session shading is a clock aid, not evidence of volume; the candle files contain no volume field.

Chapter 19

Draw, select, edit, duplicate, and clear chart objects

Drawings are persistent session records anchored to chart time and price. They belong to the instrument on which they were created and survive timeframe changes, reloads, and ordinary navigation in the same browser.

Choose the toolbar you actually need

The Buy and Sell position tools are pinned at the top of the rail. Beneath them, the toolbar shows your favourite drawing tools. The default set is Trend Line, Ray, Horizontal Line, Rectangle, Fibonacci, and Text. Select the manage-tools button to choose up to eight favourites from all 13 drawing types, then Save. The favourite list is stored in local browser storage.

Place one-, two-, and three-point drawings

• One point: Horizontal Line, Horizontal Ray, and Vertical Line commit on one click.
• Two points: Trend Line, Ray, Extended, Rectangle, Ellipse, Fibonacci, Arrow, and Measure use a first anchor and second anchor.
• Three points: Parallel Channel uses two points for the base line and a third for the offset.
• Text: click its location, type the note, and press Enter; Esc cancels the editor. An empty note is removed.

The active tool shows a preview as the pointer moves. After the object is committed, the tool deactivates. Right-click while a draft is active cancels the draft and returns the chart to normal navigation.

Select and modify an existing object

With no tool active, click a drawing. Non-position drawings can be dragged by their body; anchor handles move individual points. A compact style bar appears near the object. It offers eight colours, line widths from 1 to 4, solid/dashed/dotted styles, duplication, and deletion. Text adds an edit button and can also be edited with a double-click.

Duplicate creates a separate object shifted three current-timeframe bars to the right, making the copy visible instead of stacking it exactly over the original. Edits are written to IndexedDB, so they remain after reload.

Clear means this market, not the whole session

The Clear control appears when the current chart has drawings. It deletes the drawings loaded for that chart's instrument. Drawings on other instruments inside the same session remain. There is no undo after clear, so verify the pair label first.

Chapter 20

Use every line and level tool for the right job

Six line tools differ mainly in direction and lifetime. Choosing the smallest object that represents your idea keeps the chart honest and readable.

Trend Line versus Ray

Use a finite Trend Line when the relationship is descriptive between two anchors. Use a Ray when the strategy explicitly tests reactions after the second anchor. The first point is the origin, so reverse-order clicks can project in an unintended direction. The Extended tool is useful for research but can encourage seeing alignments before the rule says the line existed.

Horizontal Line versus Horizontal Ray

A full horizontal level visually implies the price mattered across the entire displayed period. A horizontal ray expresses causality better when the level was created by a particular event—an opening range high, break-even threshold, or newly confirmed swing. It begins at the clicked time and extends into the testable future.

Price labels

To reduce clutter, most drawings show detailed anchor price tags while selected or being created. Horizontal levels retain their right-edge price reference. Select an object when you need to audit its precise anchors; click empty space when you need a clean reading view.

Chapter 21

Represent zones with Rectangle, Ellipse, and Parallel Channel

Markets rarely react to a perfect one-pixel level. Shape tools let a protocol describe an area, but the area still needs objective boundaries and a known creation time.

Rectangle

Click opposite corners to define both a price range and time span. The object fills lightly with its stroke colour and shows top/bottom prices while selected. Use it for a supply/demand region, consolidation box, gap study, or time-bounded opening range. Drag the body to move the whole zone or drag anchors to change its boundaries.

A rectangle should not grow after a losing trade unless the strategy contains a written update rule. Record whether a touch means wick entry, close inside, or full candle acceptance; those interpretations create different samples.

Ellipse

Click opposite corners of the bounding box. The ellipse is primarily an annotation tool for circling a rejection cluster, liquidity sweep, or review error. Because its curved boundary is not a natural order level, avoid using “inside the ellipse” as an entry rule unless the geometry is defined independently.

Parallel Channel

The first two clicks define the base line. The third click sets the vertical offset to a parallel copy, producing a shaded channel between them. Use the same anchor convention every time—for example, first low, second low, then the intervening high. Reversing the anchor order or choosing extremes after seeing the breakout changes the channel.

Shapes survive interval changes

Anchors are stored as time and price, so a rectangle drawn on H1 remains at those coordinates when you switch to M15. Its edges may align differently with the smaller candles because the finer view contains details the H1 candle compressed. That is useful for auditing but does not authorize moving the original boundary unless the protocol says so.

Chapter 22

Use Fibonacci, Measure, Arrow, and Text precisely

These four tools answer different questions: proportional price location, exact distance, visual direction, and written context.

Fibonacci

Click the first price-time anchor and then the second. FXAbsolute draws full-width levels at 0%, 23.6%, 38.2%, 50%, 61.8%, 78.6%, and 100%, interpolated from the first price to the second. Direction matters: low-to-high and high-to-low place percentages in opposite semantic order even though the same prices appear.

Define which swing qualifies before drawing. “The latest obvious swing” lets the tester choose the anchors that produce a desired retracement. A stronger rule might require the highest high and lowest low between two vertical session markers.

Measure

Click start and end. The label reports signed pip distance, percentage change from the first price, and elapsed bars on the active timeframe. Use it to verify stop distance, range size, impulse size, or holding length. Because bar count depends on timeframe, record the interval with the measurement.

Arrow

Click tail and head to create a directional annotation. An arrow does not calculate a trade. Use it to distinguish planned path, observed impulse, or review comment through colour. Do not let a decorative arrow substitute for an entry marker that the trade history already records precisely.

Text

Click a location, type, and press Enter. Double-click existing text or use its edit icon to revise. Text can label setup IDs, regime, exclusion reason, or mistake category. Keep it short enough to stay readable. The Journal is better for full reasoning because it records chart time and mood.

Chapter 23

Plan and execute with the Long and Short Position tools

The pinned Buy and Sell tools create an interactive entry, stop, and target box. Unlike an ordinary drawing, the draft can create a real simulated market trade or pending order.

Create the draft

1. Select Buy for a long plan or Sell for a short plan.
2. Click the intended entry price on the focused chart. The app creates an equal-distance stop and target around entry, initially 1:1.
3. Drag the entry, red stop, green target, or right edge. The price levels have wide invisible grab areas so you do not need to hit a tiny dot.
4. Read SL pips, TP pips, R:R, risk dollars, reward dollars, lot, and the panel's internal pip-value note.
5. Set risk with the 0.1%–5.0% slider, or edit the lot field. The default remembered risk is 1% and minimum calculated lot is 0.01.

How the default distance is chosen

The initial stop is scaled to roughly 0.085% of the clicked price and rounded into the instrument's pip unit, with at least one pip. The target opens at the same distance. This only creates a visible starting box; it is not a volatility recommendation. Move both levels to prices justified by the test rule.

Buy/Sell Now versus Place Order

Buy Now or Sell Now opens a simulated position immediately at the box's selected entry line. It does not force that line to equal the current candle close. For a realistic immediate entry, align it with the current displayed close before pressing the button. Place Order leaves the selected entry as a pending touch order and waits for replay to reach it.

After execution

The position box is saved as a drawing linked to the trade. It remains on the market where it was placed. Closed boxes are dimmed. Saved boxes can still show their planned geometry, but active trade SL/TP management should use the live trade lines described in Chapter 26 so the order record and chart agree.

Chapter 24

Place a market order from the Trade ticket

The ticket is the direct path for an immediate simulated fill. It uses the focused market's current replay candle close and accepts lot size plus optional absolute-price stop and target.

Read the fields correctly

• Price: the current focused candle close, formatted for that instrument.
• Lot: simulated lot size; the field starts at 0.10. Enter a positive numeric value.
• SL: optional stop-loss price, not a pip distance.
• TP: optional take-profit price, not a pip distance.
• Buy / Sell: direction. Buy gains when exit is above entry; Sell gains when exit is below entry.

Validate direction before submitting

For a conventional long, SL belongs below entry and TP above. For a conventional short, SL belongs above and TP below. The ticket accepts numeric prices; it is your responsibility to place them logically. Inputs clear after a successful order except lot, which remains for the next trade.

One open ticket position per market

Once the focused market has an open position, the new-entry form is hidden for that market. Positions in other markets remain visible and can coexist. A pending order is not the same as an open position; cancel stale or duplicate pending rows deliberately so they cannot create an unintended later fill.

What a market fill means here

The entry is the current historical candle's close. There is no bid/ask spread or next-tick delay. If your real rule enters at the next candle open, advance to that candle and use its observed decision point consistently, or record the model difference. Do not call a close-price fill a next-open test.

Chapter 25

Create, move, fill, and cancel pending limit orders

Set Limit Order creates an orange draft line. When saved, the order remains pending until a later evaluated candle's high-low range contains the entry price.

Place from the ticket

1. Press Set Limit Order. The orange draft begins at the current close.
2. Drag the orange line to the intended entry.
3. Enter lot, optional SL price, and optional TP price.
4. Choose Buy Limit or Sell Limit. The draft disappears and a Pending Orders row appears.
5. Use the X on the row to cancel it before fill.

You can also build a Long/Short Position box and choose Place Order. That creates the same pending trade state while retaining the planned box drawing.

Move an existing order

The saved entry is an orange dashed line. Press and hold it for about 400 milliseconds, then drag vertically. Releasing saves the new price immediately. The separate SL and TP lines can be dragged with their own save confirmation.

How a fill is detected

On replay movement, the engine checks whether low ≤ limit price ≤ high. If true, the pending record becomes open at the exact limit price. The app does not enforce conventional placement geometry. For a genuine buy-limit test place entry below current price; for a genuine sell-limit test place it above. A buy above current or sell below behaves like a touch order in this engine, not a broker stop order.

Chapter 26

Manage open positions, stops, targets, and history

The Trade tab lists every open position in the session, even when its market is not focused. Each is valued from its own loaded price at the current historical timestamp.

Open-position banner

The banner shows direction, lot, entry, market label when it differs from the focused pair, floating USD result, floating pips, SL, TP, and a Close button. Floating values can be absent for an off-screen pair that has not been loaded during the page session. Closing uses that pair's price at the current timestamp if available.

Move SL or TP on the chart

Open trades draw an entry line plus red dashed SL and green dashed TP lines. Drag a stop or target vertically. On release, the chart shows a Save control and an X. Press Save to update the trade record; press X to abandon the proposed level. Merely dragging without Save does not complete the change.

Close manually

Press Close [symbol]. The position exits at the current candle close for its market, records exit time and pips, and moves to History. This is not a bid/ask or next-tick fill. If your protocol permits discretionary closes, journal the reason before pressing the button.

History and deletion

History shows closed trades newest first with direction, lot, P&L, entry-to-exit prices, and pips. Its header totals closed P&L. The X on a history row permanently deletes that trade, which also removes it from statistics. Use deletion only to correct an accidental simulator action, never to improve the sample. The X beside a pending order cancels that order rather than recording a losing trade.

Do not replay backward across a live trade

Backward chart movement does not undo database events. An open, closed, deleted, or edited trade remains in that state when you step to an earlier candle. For a valid run, avoid moving backward after entry. If an accidental click requires correction, document it immediately and restart the sample window if the later outcome was seen.

Chapter 27

Know exactly what the execution model assumes

FXAbsolute has OHLC candles, not the ordered ticks inside each candle. The engine therefore uses deterministic, conservative rules whenever the path is unknowable.

Market entries and manual exits

A ticket market entry uses the visible candle close. A position-box immediate entry uses the box entry level you selected. Manual Close uses the current candle close. None includes spread, commission, slippage, latency, rejection, or partial fill.

Automatic SL and TP

For each later evaluated bar, a long stop is touched when the low is at or below SL; a long target when the high is at or above TP. A short stop is touched when the high is at or above SL; a short target when the low is at or below TP. The entry bar is skipped so a level inside the candle used for entry does not instantly exit the new trade.

When both levels are inside one candle

The OHLC bar proves both prices traded but not which came first. FXAbsolute records the stop. This pessimistic ordering avoids systematically flattering results. It can still differ from the real tick path, so ambiguous bars should be tagged and, where finer data exists, reviewed on a smaller timeframe without changing the original result silently.

Pending entry sequence

A pending order fills when its price lies between the bar's low and high. Newly filled orders are not evaluated for SL/TP until a subsequent replay event. One bar can therefore contain an unseen adverse excursion that the simulator does not charge to that new order. Very short-stop pending strategies require separate lower-timeframe or tick validation.

Chapter 28

Understand pip units, lot size, and simulated P&L

The app first converts price movement into instrument-specific pips, then applies one simplified dollar formula. This makes session statistics internally readable but is not a universal contract-value engine.

Pip units used by FXAbsolute

Direction and pips

A positive value is a winning price move for that direction. Zero-P&L closed trades are included with losses in the current win/loss count, so break-even handling should be noted when interpreting win rate.

The simplified dollar formula

At 0.10 lot, the app therefore records $1 per app pip; at 1.00 lot, $10 per app pip. For EURUSD in a USD account this resembles a common convention. It does not automatically reproduce each broker's metal, crypto, index, JPY-cross, share, ETF, or AED contract value and conversion.

How to use the number responsibly

Keep lot size constant when comparing setup quality, or use one consistent risk-sizing method. For cross-asset research, pips and R-multiples are often more portable than the app's dollars. Before using any result for capital decisions, rerun the arithmetic with the intended broker's tick value, account currency conversion, commissions, typical spread, swap, and expected slippage.

Chapter 29

Turn each decision into a useful journal entry

The Journal tab saves a note with the current primary-chart time and an optional mood. It is most valuable before the outcome, when your reasoning cannot be rewritten by what happens next.

Create a note

Open Journal, type in the text area, optionally choose Confident, Neutral, Anxious, FOMO, or Tired, then press Add Note. You can also save with Ctrl+Enter. Selecting the active mood again clears it. Empty notes cannot be saved.

The pin line above the editor shows the chart timestamp that will be attached. In split view, journal timestamps currently come from the primary chart, so include the secondary symbol in the text if the note is about the right pane.

A compact note template

Review and delete notes

Saved notes appear newest first with chart time and mood. The X permanently deletes a note. Correct a genuine input mistake promptly; do not delete evidence of a poor decision because the trade lost. A useful review separates process (“did I follow the rule?”) from outcome (“did price reach target?”).

Use mood as a test variable, not a diagnosis

Mood tags are self-reported context. After a meaningful sample, compare rule violations or impulsive entries across tags, but do not infer causation from a handful of trades. A “Confident” loss is not proof confidence is harmful, and a “FOMO” win does not validate breaking the plan.

Chapter 30

Read the statistics without overreading them

The statistics panel updates from closed trades in the current session. Open and pending orders do not enter the result calculations until a position closes.

Profit factor and expectancy are not the same

Profit factor compares total profit with total loss. Expectancy asks what the average trade contributes. The current compact statistics panel does not display expectancy directly, but you can derive app-dollar expectancy as total P&L divided by number of closed trades. A strategy can have a low win rate and positive expectancy if winners are sufficiently larger; win rate alone is not an edge.

Colour is a cue, not a verdict

The UI colours win rate green at 50% or above, yellow from 40%, and red below; profit factor green at 1 or above. Those thresholds aid scanning but cannot decide whether the strategy is acceptable. Required win rate depends on reward/loss distribution, costs, drawdown, and sample uncertainty.

Optional report prompt

After five closed trades, an optional “Email it to me” card may appear below statistics. It asks for email and phone to send the report and occasional updates. Choose No thanks to dismiss it; it is not required to continue and is remembered on that device.

Chapter 31

Resume sessions and protect browser-local work

FXAbsolute stores sessions, trades, notes, and drawings in an IndexedDB database named fxabsolute inside the current browser profile. The account and Pro status are separate server-side records.

What is saved automatically

• Session instrument, start and end indices, start balance, creation/update time.
• Current replay index and timestamp, selected timeframe, and playback speed.
• Pending, open, and closed trades with levels and outcomes.
• Journal notes and moods.
• Drawings, styles, pair ownership, and position-box links.

Replay writes are throttled to avoid a database operation on every animation frame, then flushed when the page hides. At worst, an abrupt browser kill can lose approximately the last second of cursor movement; committed trades and notes are separate immediate writes.

What local browser storage also remembers

The chart timezone, active indicators, drawing favourites, collapsed panels, risk percentage, usage counters, and pairing code state use localStorage or sessionStorage. Some survive browser restarts; a pairing code is scoped to the tab session. Private-browsing behaviour varies and may discard everything when the window closes.

Resume paths

The home page shows three recent sessions. A signed-in profile lists all sessions currently stored in this browser. A backtest URL may restore a local session ID. If that row no longer exists, the URL cannot reconstruct it from the server.

If sessions seem missing

1. Check that you opened the same browser, OS account, and browser profile.
2. Exit private/incognito mode and inspect the normal profile.
3. Do not clear cookies/site data while diagnosing.
4. Check whether storage is blocked; the Profile page explicitly reports when IndexedDB cannot be read.
5. If the device was reset or site data was cleared, authentication can return but local sessions cannot be restored from the account.

Chapter 32

Connect your own AI coding agent to the chart

The AI tab does not call a built-in model. It pairs the browser with a compatible agent you already use and sends a structured chart snapshot with each question.

Install once

Your local environment needs Node.js 18 or later and an MCP-compatible client such as Claude Code, the Claude desktop app, Codex, or Cursor. Copy the command shown in the AI panel:

The command is written for Claude Code's MCP configuration. Other clients use their own MCP server setup screen or configuration format; the standalone setup page covers variations at Connect your agent.

Pair the current chart

1. Open AI in the session side panel.
2. Copy the six-character command shown under “Give it this code.”
3. Tell the configured agent: connect to fxabsolute with code ABC234, replacing the example with the live code.
4. Wait until the status says Connected to [agent].
5. Ask a specific question, such as “Is the marked trendline respected by closes or only wicks?”

What is sent

Each message can include instrument, timeframe, replay bar/time, current OHLC, visible time range, numerical drawing anchors, open trades, closed-trade count, and session performance. The context tells the agent not to reveal future bars unless explicitly asked. Drawings are transmitted as price/time data rather than a screenshot, so the agent can calculate distance and slope instead of estimating pixels.

Voice input

If the browser exposes the Web Speech API, the microphone button transcribes English speech and sends the completed dictation automatically. The browser may ask for microphone permission. If support is missing or permission is denied, type the question; chart pairing still works.

Privacy and disconnecting

FXAbsolute does not ask for your model API key. The browser and agent exchange ephemeral messages through a pairing channel. Anyone who knows the active code could attempt to pair, so do not post it publicly. Press New code to rotate it, close the panel/page to disconnect the browser, and follow your agent client's own removal process if you no longer want the MCP server installed.

Chapter 33

Use Profile, Community, Messages, and feedback

Select your avatar or initials to open the signed-in profile. Its three tabs separate browser-local trading history, the shared community room, and your support conversation.

Your trading

The tiles aggregate sessions and closed trades stored in this browser: number of backtests, closed trades, non-closed items, win rate, profit factor, net simulated P&L, net pips, and most-used session market. Beneath them, each local session shows bars advanced, closed-trade count, result, and a Resume action.

These totals can differ from another computer even under the same login because the underlying trading rows do not sync. They also combine instruments whose pip units and simplified dollar economics differ; use pair-specific samples for research conclusions.

Community

The community displays numbered participants rather than public names or emails. You must sign in to post. The number is assigned when an account first posts, so it acts as a stable room identity without exposing the account address. Opening the room marks visible new messages as seen.

Messages

Messages is the private support thread for the signed-in account. Send a clear problem description with the affected page, pair, timeframe, approximate time, browser/device, and what you expected. Support replies appear in the same thread and can trigger the unread badge. Opening the thread marks retrieved replies read.

Footer feedback

“Send us whatever you feel about our product” opens a low-friction feedback box. A signed-out submission is anonymous unless you include a contact method in the text. For payment or account issues, sign in first and use Messages when possible so the team can associate the conversation with the correct account—never send passwords, access tokens, private keys, or recovery phrases.

Chapter 34

Understand what Absolute Tracker currently does

Absolute Tracker is a coming-soon product page, not an active account connector. It describes the intended read-only tracking service and lets visitors prebook interest.

Current public function

Open Absolute Tracker from the top navigation. The page explains a future system for MetaTrader, cTrader, crypto exchanges, prop-firm accounts, and statement uploads. Today, it does not request broker credentials, ingest trades, verify fills, or change your backtest statistics.

Prebooking

Choose a platform, enter an email if signed out, add an optional note, and press Prebook my place. A signed-in email is filled automatically. The form sends interest and requirements to FXAbsolute; it does not create a subscription, take a card, or commit you to the future product.

Backtester versus Tracker

Chapter 35

Free access, Pro access, checkout, and delayed payment confirmation

There are two separate thresholds: a free account prompt after initial use, then a Pro offer after meaningful signed-in use. Payment is attached to the authenticated account, so sign-in method matters.

The account threshold

The first session needs no account at all. Starting a second one asks for a free account, and the modal remains until sign-in or sign-up succeeds. Nothing here runs on a clock: time spent reading, thinking or with the tab in the background costs you nothing, because what is counted is sessions rather than minutes. Creating an account is free.

The Pro threshold

The free plan is two sessions, each replaying up to 14 days of market data from any start date across the five years. Everything else — every instrument, timeframe, indicator, drawing tool, the journal and the statistics — works exactly as it does for somebody who has paid. The offer becomes due when a third session is started. Existing active Pro access bypasses the wall. Countries the checkout cannot serve remain exempt rather than being trapped behind an unavailable payment route.

Current offer

At this guide's 7 September 2026 review, the in-app offer is $4.99 USD for six months, paid once with no automatic renewal. It was $13 for twelve months between 31 August and 5 September. Two rails are offered: card — which also covers Apple Pay, Google Pay and PayPal — and crypto, where a hosted invoice presents USDT, BTC and around two hundred other choices. Pricing and available methods can change; the amount shown inside the authenticated checkout is authoritative before you send funds.

Pay step by step

1. Confirm the account. Check the avatar email. If you registered through Google, there is no separate FXAbsolute password; keep using Continue with Google.
2. Open checkout once. Press Pay by card or Pay with crypto. The server verifies the signed-in token, creates an order and redirects to the provider. Buying again while access is still live adds to what is left rather than replacing it, so an early renewal never costs you the remainder of the term you already paid for.
3. Read the invoice. Confirm amount, currency/network, destination, and any expiry. Send exactly what the provider requests using the specified network.
4. Wait for final status. A transaction broadcast is not always a final provider confirmation. Keep the invoice/payment ID for support.
5. Return to FXAbsolute. The app rechecks entitlement every three seconds for roughly 30 seconds after a checkout return.
6. Reload if needed. If the chain or provider settles later, an automatic reconciliation checks recent unfinished orders every 15 minutes and grants confirmed payments idempotently.

What grants access

A browser redirect alone never grants Pro. The payment callback is signature-checked and the provider status is independently queried. Only a final confirmed or finished status grants a 365-day entitlement to the order's user. Repeated callbacks or reconciliation passes do not stack duplicate years for the same payment.

Common account/payment mismatch

A person can pay while signed into one Google address, then return under another browser profile or email/password account. Pro correctly belongs to the original authenticated user. Sign out, use the same Google button and same Google address that opened checkout, then reload. A password-reset email cannot recover a Google-authenticated account because there is no FXAbsolute password to reset.

What Pro changes—and what it does not

Pro keeps the current backtesting toolkit available past the signed-in allowance for the entitlement period. It does not convert simulated results into live trades, cloud-sync local sessions, or activate the coming-soon Tracker. Your local data rules and simulator limitations remain exactly the same.

Chapter 36

Worked example: an auditable EURUSD London-range backtest

This example demonstrates the complete FXAbsolute workflow. It is a testing template, not a claim that the strategy is profitable. The point is that every discretionary-looking phrase is replaced by a decision another person could repeat.

1. State the research question

Question: On EURUSD M15, during dates in 2024, what happens when price closes beyond the first hour of the London session and then retests that boundary before 12:00 London time? This defines instrument, entry timeframe, historical period, event, and allowed window. The test will estimate behaviour under FXAbsolute's candle model, before real costs.

2. Freeze the rules before opening outcomes

3. Configure the session

1. Set a $10,000 balance and choose EURUSD.
2. Select a chronological start such as 2 January 2024. Do not choose dates because you remember large moves.
3. Start, select M15, and turn on Sessions. Keep other indicators off because they are not in the rule.
4. Use a Vertical Line at the London open. After four candles complete, draw a Rectangle over the range and Horizontal Rays from the high and low beginning at the range close.

4. Replay the decision

Advance +1 through the opening range and any break/retest window. When a break closes, write a journal note with range high/low, direction, break time, and last permitted retest time. If a qualifying retest closes, use the position tool at the current close, drag SL to the defined buffered price, set target to exactly 2R, and press Buy Now or Sell Now. Because the box immediate entry can be moved away from current price, confirm its entry line equals the visible close.

5. Record no-trade days

A backtest that records only trades cannot distinguish “no setup” from “tester skipped a loser.” Add a journal note for each evaluated day: no break, no retest, invalid late retest, stop too wide, both sides broke, or trade taken. These denominators let you measure signal frequency and audit exclusions.

6. Review after a fixed sample

Do not modify rules after five losses. Complete the predeclared number of eligible days—such as 60 sequential trading days—then export results manually into a research sheet if deeper analysis is needed. Separate compliant trades, operational mistakes, and ambiguous bars. Any revised rule becomes a new out-of-sample test, not a rewrite of the first one.

Chapter 37

Worked example: XAUUSD pullback with multi-timeframe context

Gold magnifies execution-model weaknesses because its candles can cover large distances quickly. This workflow uses H1 context and M15 decisions, tags event-like bars, and treats the app's dollars as internal simulation units.

Define the setup

Question: After an H1 close establishes a new 20-bar high above EMA 50, how does the first M15 pullback to the breakout level behave during the London or New York band? The example permits one long candidate before the H1 structure fails. It does not assume a breakout edge exists.

Configure and mark context

1. Choose XAUUSD, $10,000, and a sequential date in the available 2021–2026 range.
2. Open H1 and enable EMA 50 plus Sessions. Mark the highest high of the prior 20 completed H1 candles with a Horizontal Ray only after the breakout closes.
3. Switch to M15 at the same timestamp. The ray remains because its anchors are time/price records.
4. Define the first pullback as an M15 low touching the ray, followed by an M15 close above it while the H1 close remains above EMA 50.

Size in app pips, then separate live economics

FXAbsolute defines one XAUUSD pip as 0.1 price unit. If entry is 2,420.0 and stop 2,415.0, distance is 50 app pips. A target at 2,430.0 is 100 app pips, or 2R by distance. At 0.10 simulated lot the app would display about -$50 or +$100. A broker's gold contract, tick value, spread, and margin can differ; never transfer that lot directly.

Handle news-like candles honestly

The app does not include an economic calendar. If a candle is far larger than neighbouring bars, tag it as “event-like/unknown” rather than inventing a cause. Do not delete it. Run two summaries later: all qualifying observations and a predeclared sensitivity set excluding event-like bars by an objective range threshold.

Use the finer view only as declared

If an M15 result is ambiguous because both SL and TP lie inside one bar, the M1 series can sometimes show which boundary appears first. Decide before the study whether M1 overrides the conservative M15 result or merely annotates it. Never open M1 only for losing ambiguous bars.

What this workflow teaches

It separates regime, location, trigger, risk, and execution resolution. That decomposition is more useful than saying “trade gold breakouts.” It also reveals how many signals disappear once session, first-pullback, and stop-distance constraints are applied.

Chapter 38

Worked example: an EMAAR daily-to-hourly UAE stock study

UAE equities require a different workflow because daily history can span years while intraday tiers cover much shorter periods. Timeframe availability must be part of the research design.

Ask a question the data can answer

Question: During the overlap between EMAAR hourly coverage and its daily history, what happens after an hourly close breaks the prior five-session high, followed by the first hourly retest during DFM hours? This avoids pretending old daily setups can be executed on minute data that does not exist for those dates.

Configure the market

1. Set timezone UTC+4 and choose EMAAR (Emaar Properties) under Stocks.
2. Start on the default 1H view. The UAE launcher prefers hourly because it offers far more context than the short minute tier.
3. Turn on Sessions. The chart shows DFM / ADX 10:00–15:00 Dubai rather than Tokyo/London/New York.
4. Switch to 1D only to label the broader regime, then return to 1H at the same historical timestamp.

Respect unavailable buttons

If the selected old timestamp lies outside EMAAR's M1 or M5 tier, those buttons are disabled. This is not a loading failure. Do not move to a recent month merely to obtain a finer entry after reviewing an older daily outcome. Either define the study on H1, or restrict the entire sample to the minute-tier overlap from the beginning.

Quote and simulation conventions

EMAAR prices display three decimals and one app pip equals AED 0.001. An entry at 12.350 and stop at 12.270 spans 80 app pips. The session's P&L still applies the generic pips × 10 × lot USD-labelled formula, which is not an AED share-position valuation. Use price percentage, pip distance, or R for research; calculate real shares, fees, AED/USD conversion, and exchange constraints outside the app.

Liquidity gaps are information

Some UAE symbols have sparse intraday bars. FXAbsolute leaves intervals empty instead of fabricating flat candles. A “five bars later” rule can therefore span more wall-clock time than expected. When time matters, define exits by displayed timestamp or session boundary rather than bar count alone.

Chapter 39

Build a backtest protocol that can survive review

A long session is not automatically a good test. Validity comes from freezing the decision rules, preserving chronological information, and recording every eligible observation—including no-trade cases.

The minimum protocol

Prevent look-ahead bias

Use +1 near decisions, never scrub ahead, and place levels only when the information needed to define them has closed. A previous-day high is knowable after that day; today's final high is not. A range boundary is knowable after the range window completes; extending it backward visually does not make it historically available.

Prevent selection bias

Test every eligible date in a sequence, not memorable trends. Record “no setup” and every exclusion reason. If you switch instruments, apply the same schedule and rule. Stopping when profit factor looks high biases the estimate; stop at the predeclared sample size or period.

Prevent rule drift

Version the protocol. When you discover that “clean retest” is ambiguous, finish or invalidate the pilot, write a measurable definition, and start Version 2 on unseen dates. Do not relabel old trades using the new interpretation and call the combined total one test.

Account for multiple testing

Trying EMA 20, 50, and 200 across ten timeframes and 53 markets creates thousands of combinations. Some will look excellent by chance. Use a small development period to form a rule, then a separate validation period that was not used to choose it. Treat the validation result as primary.

Chapter 40

Review results as evidence, not a verdict

The statistics panel is a summary of the recorded sample. Your review must add sample size, rule adherence, uncertainty, costs, and the distinction between development and validation.

Start with data integrity

1. Count eligible periods, no-setup periods, taken trades, skipped valid trades, and accidental simulator actions.
2. Confirm every entry existed at the stated evaluation point.
3. List ambiguous bars, pending fills on sparse data, and off-screen markets that may not have been cached.
4. Separate trades placed under different rule versions.
5. Reconcile trade history totals with the summary before analysing subgroups.

Convert outcomes to R when risk varies

For each trade, define 1R as its initial entry-to-stop risk. A $150 win on $100 initial risk is +1.5R; a full stop is -1R. R lets you compare trades with different stop distances or lots more cleanly than raw app dollars. The current panel does not calculate R automatically, so record it in a separate sheet or journal summary.

Inspect distribution, not only averages

Report median and extremes, losing streaks, the share of total profit contributed by the best few trades, and performance by time period. A positive total produced by one outlier is fragile. A smooth average can hide a long unusable drawdown. FXAbsolute's compact panel does not currently calculate maximum drawdown or confidence intervals, so do not claim them from the visible tiles.

Use subgroups only when justified

Session, weekday, volatility regime, direction, and mood can explain variation, but each split reduces sample size. Mark exploratory findings as hypotheses for the next unseen sample. “Tuesdays lost in 12 trades” is not yet a Tuesday filter.

Run sensitivity checks

Recalculate after a plausible spread/slippage deduction, after removing accidental entries, and under conservative ambiguous-bar treatment. For multi-asset tests, use R or percentage rather than summing economically different app pips. If a small cost or one outlier reverses the conclusion, the edge is not robust yet.

Chapter 41

Troubleshoot the app systematically

Start with the symptom, preserve local storage, and change one condition at a time. Reloading is reversible; clearing site data is not.

A safe diagnostic order

1. Write down the URL, pair, timeframe, timestamp, and visible error.
2. Pause replay and take a screenshot if it helps demonstrate state.
3. Check focus, timezone, coverage, and exact numeric levels.
4. Reload the page once; resume the saved local session.
5. Try a current Chromium, Firefox, or Safari browser profile with storage enabled, without deleting the original profile's data.
6. Send support reproducible steps and expected versus actual behaviour.

What not to send

Support never needs your Google password, email password, crypto seed phrase, private key, full card details, Supabase token, or GitHub secret. A payment ID, approximate time, public transaction hash where appropriate, and account email are enough to locate an order.

Chapter 42

Use FXAbsolute on touch devices and with accessibility needs

The interface resizes to the viewport, supports touch panning, pinch scaling, touch drawing, and touch trade-line dragging. Precision chart work is still easier with a larger screen and pointing device.

A practical mobile layout

1. Use landscape orientation when reading or moving exact price levels.
2. Collapse the drawing rail when replaying and collapse the side panel when analysing.
3. Avoid split view unless both charts remain legible; use instrument switching on narrow screens.
4. Reopen the side panel only after confirming the focused market.
5. For an existing pending orange line, press and hold about 400 ms before dragging.

Keyboard actions currently available

Reduced motion and readable structure

This guide honours the operating system's reduced-motion preference and provides semantic headings, a skip link, labelled navigation, visible focusable controls, and print styles. The app uses labelled buttons for major account and chart controls. However, a canvas-based financial chart and drag-only drawing geometry are not fully equivalent to a keyboard/table interface.

If precise dragging is inaccessible

Use the Trade ticket's numeric SL/TP fields instead of drag handles, the date input instead of the scrubber, and the information-bar price readout to verify a level. For formal work, consider using desktop zoom, OS magnification, or a larger display. Send feedback describing the exact blocked action so an alternative control can be prioritised.

Chapter 43

FXAbsolute glossary

These definitions describe how terms are used in this app and guide. Broker or exchange usage can be more specific.

Three terms that should never be confused

Accuracy asks whether the simulator applied its documented rule correctly. Realism asks how closely that rule represents an executable market process. Profitability asks whether results remain positive after costs and uncertainty. A backtest can be internally accurate while its execution assumptions are not realistic enough for a live conclusion.

Chapter 44

Frequently asked questions

Short answers to the questions most likely to block a session. The chapters above contain the operational detail and examples.