Colours, Ribbons & Conditional Candles

Colour is the fastest language a chart can speak. Before you read a single number, a green candle says "up" and a red one says "down"; a band of green along the bottom of the screen says "the trend is your friend right now." A well-coloured chart lets you grasp the state of the market in a heartbeat - and the tools behind it are ones you already know. Everything in this chapter is IIf choosing a colour, plus a few new Plot styles that turn those colours into candles, ribbons and shaded backgrounds.

This is where your charts stop looking like a beginner's and start looking like a trading desk's. And it costs almost nothing - a colour array here, a style flag there.

Colours are just values - and arrays

A colour in AFL is a value like colorGreen or colorRed. The leap that unlocks this whole chapter is that a colour can also be an array - a different colour on every bar - and the way you build that array is the IIf you met in Chapter 8:

upTrend = Close > EMA(Close, 50);

// a colour array: green on uptrend bars, red elsewhere
candleColor = IIf(upTrend, colorGreen, colorRed);

candleColor now holds a colour per bar. Feed it to Plot in the colour slot and every drawing element gets coloured individually.

Colouring candles by a rule

Pass that colour array to Plot together with styleCandle and you get candles painted by your rule rather than by simple up/down. Here is the author's MACD example, which colours the candles by whether MACD is above zero:

imacd = MACD(12, 26);

// green when momentum is positive, red when negative
color = IIf(imacd > 0, colorGreen, colorRed);

Plot(Close, "Candle", color, styleCandle);

Suddenly the price candles themselves carry the momentum signal - no separate pane needed. The same trick works for any condition: trend, RSI zone, volume surge. Swap the rule, keep the structure.

See the colour decided per bar

It helps to watch the colour get chosen one bar at a time. up is a true/false column, and IIf reads it bar by bar and hands back a colour, so each candle is coloured by its own bar's rule - never by what its neighbours are doing:

Green where the bar closed up, red where it closed down - decided independently on every single bar.

A trend ribbon along the bottom

A ribbon is a thin strip of colour running along the bottom of the chart - a constant visual reminder of state that never crowds the price. You build it by plotting the value 1 (so it fills a fixed-height band) with a special combination of styles:

upTrend = Close > EMA(Close, 50);
ribbonColor = IIf(upTrend, colorGreen, colorRed);

// a full-width strip of colour pinned to the bottom of the pane
Plot(1, "", ribbonColor, styleArea | styleOwnScale | styleNoLabel, 0, 1);

Three style flags do the work, joined with |:

• styleArea fills solid down to the axis instead of drawing a line.
• styleOwnScale gives the ribbon its own private scale so it ignores the price axis and does not squash your candles.
• styleNoLabel keeps its value off the y-axis labels.

The trailing 0, 1 set the min and max of that private scale, which pins the strip to a clean band at the bottom. Run it and a green/red bar tracks your trend along the foot of the chart - a glanceable context strip you will want on every layout.

Background shading

Where a ribbon hugs the bottom, background shading tints the whole pane behind the price - ideal for marking a regime like "bullish" or "high volatility." You do it by passing a colour to SetBarFillColor or, more simply, by plotting a tall area in the background. A clean way is a shaded cloud between two levels using PlotOHLC with styleCloud, exactly as the author shades a Bollinger Band:

upper = BBandTop(Close, 20, 2);
lower = BBandBot(Close, 20, 2);

Plot(upper, "Upper Bollinger", colorYellow, styleThick);
Plot(lower, "Lower Bollinger", colorYellow, styleThick);
Plot(Close, "Candles", colorDefault, styleCandle);

// a soft fill between the two bands
fill = ColorBlend(colorYellow, colorBlack, 0.9);   // a faint yellow
PlotOHLC(lower, lower, upper, upper, "", fill, styleCloud);

PlotOHLC here is told to draw a band whose "open/high/low/close" are simply the lower and upper bands, and styleCloud fills the space between them. ColorBlend mixes the bright yellow heavily with black to get a faint tint that shades without hiding the candles underneath.

MACD and RSI colour coding

Indicator panes come alive with the same treatment. Colour a MACD histogram by its sign, or an RSI line by its zone, and the indicator tells its own story:

// MACD coloured by sign, with a zero reference line
imacd = MACD(12, 26);
macdColor = IIf(imacd > 0, colorGreen, colorRed);
Plot(imacd, "MACD", macdColor, styleThick);
Plot(0, "", colorWhite, styleDashed);   // the zero line

// RSI coloured by overbought / oversold / neutral
rsi = RSI(14);
rsiColor = IIf(rsi > 70, colorRed,
           IIf(rsi < 30, colorGreen,
                         colorYellow));
Plot(rsi, "RSI", rsiColor, styleThick);

The nested IIf for RSI is the three-zone pattern from Chapter 8, now doing visible work: red when stretched, green when washed out, yellow in between.

colorXXX constants versus ColorRGB

You have two ways to name a colour. The colorXXX constants - colorGreen, colorRed, colorAqua, colorOrange and dozens more - are quick, readable and cover most needs. When you want a precise shade the constants do not offer, ColorRGB(r, g, b) mixes any colour from red, green and blue components, each from 0 to 255:

myTeal      = ColorRGB(0, 150, 136);    // a custom teal
softGrey    = ColorRGB(60, 60, 60);     // a muted background grey

Plot(Close, "Price", myTeal, styleThick);

ColorBlend(colorA, colorB, fraction), used in the cloud above, is a third handy tool - it mixes two named colours, which is the easy way to get a faded version of a constant for subtle fills.

Try it yourself

• Colour the price candles by whether Close is above its 20-EMA, then switch the rule to MACD above zero and compare.
• Add a green/red trend ribbon along the bottom using styleArea | styleOwnScale | styleNoLabel, driven by the same trend condition.
• Shade a Bollinger Band cloud with PlotOHLC and styleCloud, then lighten or darken the fill by changing the ColorBlend fraction.
• Build a three-zone RSI colour with nested IIf and plot it; tweak the 70/30 thresholds and watch the colours shift.

Recap

• A colour can be an array - build it with IIf(condition, colorA, colorB) and pass it to Plot to colour by a rule.
• Plot(Close, "", color, styleCandle) paints the candles themselves by any condition - trend, MACD sign, RSI zone.
• A trend ribbon is Plot(1, "", color, styleArea | styleOwnScale | styleNoLabel, 0, 1) - a glanceable strip pinned to the bottom.
• Background shading with PlotOHLC(..., styleCloud) (and ColorBlend for a faint tint) marks regimes without hiding price.
• Colour indicator panes - a MACD histogram by sign, an RSI line by zone - so each pane tells its own story.
• Use the readable colorXXX constants by default and ColorRGB(r,g,b) when you need a precise custom shade.

That closes Module C: you can now make a chart that communicates at a glance. Next we open the Analysis window and put AFL to work across the whole market at once - the world of exploration and scanning.