Just as sure as a hurricane brings TV news footage of a reporter standing on a boardwalk with waves crashing in the background, stormy times in the market also bring out journalistic cliches. Breathless and overhyped commentary and pictures abound – in print, on the air and online.
One of my pet peeves is a subspecies of exaggerated reportage: graphs that deceive the eye. I’m not saying it’s intentional, but some graphs depicting the latest carnage create an erroneous perception that overstates the “meltdown.” (The graph shown here is from Page One of today’s Wall Street Journal, though I’m not singling them out.)
The problem is – I know this will sound nerdy – the lack of a proper scale on the Y axis. When a graphic designer sets the base of the scale not at zero but at some higher point, it creates a visual exaggeration. In this case, a Dow Jones Industrial Average index of 10,900 is the base. The truncated scale makes a more dramatic picture, but the reader gets a quantitative impression that is out of context. Instead of a 4.4% drop, our eyes see the stock market plunging more than 95% – very close to “zero” on the graph.
Of course, an editor might say readers are smarter than that; anyone sophisticated enough to read the WSJ can tell the difference between zero and 10,900. Yes, but images do influence our thinking – and more on an emotional level than a rational one.
The representation of numbers, as physically measured on the surface of the graphic itself, should be directly proportional to the numerical quantities represented.
Tufte even offers a formula for a “Lie Factor” to gauge how far out of proportion a graph is. (This calculation is off the scale for most stock-price charts in the media, which are vertically truncated.)
Moving beyond whining about a pet peeve, I might suggest a lesson for investor relations professionals: We should always look at our graphs – those bar charts that fill PowerPoint presentations and some say are eye candy in annual reports – and test them for visual integrity.
The classic bar chart might show EPS rising from $3.00 to $3.25 to $3.50 over three years. If the Y-axis scale starts at zero, the eye sees a 17% total rise – looks steady, not too bad. But if you draw the scale starting at, say, $2.00, the increase in EPS looks like a more dramatic 50% – much more “growthy.” Try it both ways in Excel or PowerPoint; you’ll see. My Excel sets “zero” by default at $2.70 – which really makes for skyrocketing growth.
Financial communicators of all sorts, corporate or journalistic, should be careful to present information not only accurately, but in context and with perspective … which I think means graphs drawn to scale.