How To Lie With Financial Statistics
It is no secret that various people or bodies within the financial services industry will say or write whatever will sell the products, rather than what is actually true and correct. For this reason, buyers may be more inclined to rely on what they think is hard evidence – in the form of statistics and charts. Regrettably, one can lie with these too.
In 1954, Darell Huff wrote a classic book titled "How To Lie With Statistics," and a German professor by the name of Walter Krämer has had enormous success with a similar book with the same title, except in German "So Lügt Man Mit Statistik" (revised edition, Piper Verlag, 2011). The charts below are also from the book.
In this article, we will see that it is possible to present financial material in such a way as to exaggerate, minimize, distort or generally deceive the reader or viewer. You can say what you want with statistics, just like you can in words.
The Trick Bag
Sadly, it is a pretty full bag. The simplest of all tricks is simply what Krämer calls "the illusion of precision." By using decimals, for instance, a figure sounds more convincing. Instead of writing that you beat the index by 2%, if you put it as 2.35% it has a ring of authenticity. Whether this is correct or sustainable is another story.
And then, there are various other percentage ploys. "We have expanded our team of financial analysts by 50% over the last year," is an impressive claim. But, if it means that instead of only one full-time person, you also have someone coming in during the mornings, the reality is less exciting.
Similarly, you may read that a particular foreign country is a good investment because, among other leading indicators, its unemployment rate has dropped from 10 to 8%. But what if that 2% had simply been removed from the stats and just isn't counted anymore? As for growth and inflation rates, it is just amazing what can be done by focusing on rates of change rather than the actual level of inflation, and so on.
Talking of sustainability, one way of spinning a chart is to choose the starting or end point so as to emphasize only part of the process. Thus, if you want to stress an upward trend, start the graph in a deep trough. If you want to make the graph look generally positive, just leave the downturn out altogether. This is clearly shown in the graphs below.
Where does the curve begin? The impression created is very different if you leave out the first plunging segment.
This chart shows how using bars and adding in one more year (on the left) with an expected value, makes it all look considerably rosier than on the right
Equally misleading are tricks with axes and scales. For example, if you present the sales figures of a stagnating company with the turnover running from nothing to $100, it is obvious that there is no growth or dynamic. On the other hand, if you chop off most of the y-axis, you can exaggerate tiny fluctuations so that things look like they are on the up-and-up. Indeed, this can be combined with the previous trick to double the effect.
This is beautifully illustrated below:
Likewise, you can make a sideways stock market look like a roller coaster, simply by changing the scales (see figure below). By expanding or contracting the axes like a concertina, you can fool a lot of people a lot of the time. Amazingly, both charts below show the same market!
Such tricks can quite literally (but graphically only!) turn a failure into a success, near bankruptcy into a hot stock and horrendous volatility into gently wavering stability. The problem is that reality will eventually prevail, but by then, the seller hopes to have taken the money and run.