18 Chart Crimes and Misdemeanors
Truth is relative, falsehoods are seductive, and shortcuts are commonplace.
Chart crimes are a clear manifestation of this. These badly built graphics are blunders at best.
Most commonly, they are financial fake news. It’s not an antidote, but I think awareness will help counter this subtle phenomenon. As far as I can tell, the expression “chart crime” entered finance Twitter’s orbit after Reuters journalist Eric Burroughs began using it in February 2015.
Connoisseurs always have their favorite category, and I’m no different.
Fed balance sheet pic.twitter.com/lJopxYhyny
— zerohedge (@zerohedge) May 7, 2018
This is, like, the Macbeth of chart crime.
Everything about it is perfect. The Y axis is scaled so that a 3% contraction of the US Federal Reserve’s balance sheet looks like a fire sale.
It may well be that the person who made it was just busy. Anybody who produces analytical work has created something that doesn’t quite come across the way it was intended. That’s the nature of taking intellectual risk. Whatever else this graphic is, it’s not helpful. It leaves the reader believing in an apparition.
.Zooming out for pre-crisis context. pic.twitter.com/SSAIvemytT
— Nii A. Ahene (@Nii_Ahene) May 7, 2018
Thankfully, the nature of social media allows observers to check these behaviors in the marketplace of ideas. It’s both helpful and appropriate to do this, but it doesn’t quite generate the same level of engagement as the misrepresentative masterwork that preceded it.
And it also doesn’t mean people will stop making them.
https://twitter.com/OccupyWisdom/status/992775996087652352
Like Macbeth, this same story just keeps being told with a different cast. Whether you’re watching a version starring Patrick Stewart, Michael Fassbender, Peter O’Toole, or Toshiro Washizu, pretty much everyone still dies at the end. And yet it continues.
This innate desire for big, newsworthy charts can contort some commentators’ output, turning them into brands built around bad news. It can be hard to contravene as an observer, especially because the corpus of crimes only gets more subtle with additional study.
This crash looks more and more like 1987: The resemblance between this week's market and Black Monday in 1987 cannot be more striking. pic.twitter.com/u0gPQl29DB
— Holger Zschaepitz (@Schuldensuehner) February 9, 2018
Indulge me in a thought experiment: Let’s assume the chart above is 100% true. What to do with this information?
One supposes we’d sell everything. When Holger Zschaepitz tweeted this on 9 February 2018, the Dow Jones Industrial Average had just concluded a 2,500-point selloff. It would go on to rally for most of the month, and has since gone back and forth to sit right around the same place.
Aren’t you glad to have hypothetically sold?
The thing about these analog charts is that they raise a specter instead of presenting an argument.
They get brought up relatively frequently, often inspired by a viral clip of Paul Tudor Jones II in the 1987 documentary Trader. In the clip, a youthful Jones and his colleague Peter Borish go short right before the crash. They get a 62% monthly return to show for it, and earn the moniker “legendary” in the process. The analog helped them generate the conviction to put on an aggressive position. Doesn’t that mean it’s a legitimate form of analysis?
Well maybe for some great traders, but for most of us, it’s just a recurring distraction.
Today's stock market bears some resemblance to 30 years ago, just before the #BlackMonday crash. https://t.co/cvNSLDkLP3 pic.twitter.com/hYA7w9TPvV
— FactSet (@FactSet) September 15, 2017
It really is shocking how often these surface, and how surface-level their implications are.
There are #chartcrimes and there are 2-x-axis-chartcrimes. From BAML's actually-pretty-good Bitcoin update. pic.twitter.com/BEsiMk6XS5
— Paul McNamara (@M_PaulMcNamara) July 26, 2017
I could probably find dozens of these overlay charts. But there are so many more excesses to curate.
Say whatever you want about the charts above, at least all of them actually have a Y axis.
today in #chartcrime #chartcrimes we have this truncated #yaxisabuse by $HOME At Home Group pic.twitter.com/lRrE2rCWRZ
— Barbarian Capital (@BarbarianCap) January 10, 2017
Amazing things can happen when information is presented without the grounding influence of a clear vertical scale.
I’ll grant that a flexible capital structure is an advantage, but think about what this chart is really doing. It shows the evolution of leverage ratios over three fiscal years from 4.8x debt/adjusted EBITDA to 3.2x. But just glancing quickly at the area of each entry on the bar chart, you’d think management has entirely changed its attitude towards leverage. The 12-month figure for fiscal year 2017 looks like 10% of the value two years prior. It’s really two thirds.
Massive #chartcrime #chartcrimes here from Colgate… no y-axis so unclear whether it's dollars, % of sales, new projects, existing products, headcount, patents… but it excludes Venezuela for a good measure pic.twitter.com/RFjY7J5vb2
— Barbarian Capital (@BarbarianCap) March 9, 2018
But again, at least the chart has labels.
So far as I can tell, the only one in this classic from my colleagues at Barbarian Capital is “the growth.” And I hate to quibble, but five years of growth in whatever this is from 2013 to 2017 looks like just a lot of running in place to my eye.
And it turns out that’s a theme.
https://twitter.com/HotlantaCapital/status/920024868724228097
Sales-y corporate nonsense is one of the major categories of chart crime. And if you are an investor-relations professional, you ought to be aware that idiotic infographics produce much more than aesthetic objections in the analyst community.
They also heighten skepticism, take readers out of your presentation, and ultimately produce more work.
https://twitter.com/jonkay/status/920399480553275392
This presentation slide reminds me of how Jeff Bezos reportedly banned presentations from his senior team meetings to increase efficiency. There’s even a Swiss political party dedicated to its abolition that argues presentation technology imposes an annual cost of 2.1 billion Swiss francs.
I want to work towards a future where such initiatives are successful.
Surely there’s a better chart (via @tracyalloway)#ChartCrimes pic.twitter.com/5jDPnSZn9U
— Plan Maestro (@PlanMaestro) March 5, 2018
The present is often disconcerting.
Let’s deconstruct this chart: Where did its creator go wrong?
I’d argue the original sin is in trying to squeeze too much information into one place, which it turns out is a common issue.
Hello, is that the chart police? I want to report a #chartcrime pic.twitter.com/DjvjQLWOhs
— Paul McNamara (@M_PaulMcNamara) October 6, 2017
A protip: If you find it necessary to invent a new visualization paradigm for your currency coverage, you might be concentrating on the wrong problem.
How big has this week's China stock sell of been? It's been Six Greeces big! #applesoranges #chartcrimes pic.twitter.com/ENd7qdQ6yJ
— Lorcan Roche Kelly (@LorcanRK) June 19, 2015
The opposite mistake — too little information in too much space — is less prevalent but still problematic.
The main reason this chart is bad is that its creator wanted to present something other than the information in front of them.
Fight that urge.
The trend line on this though #chartcrimes pic.twitter.com/4jRPNRwcV9
— Tomas Hirst (@tomashirstecon) July 13, 2017
Also worth fighting is any impulse to present a conclusion when you only have information.
Most often, it’s the color scheme or our capabilities that fail us. WebMD reckons 5% of men are red-green colorblind. So one in 20 men can’t process our profession’s prevalent color scheme.
Uncertainty also takes art to convey in graphical form. Falling victim to these complexities isn’t necessarily criminal.
https://twitter.com/Birdyword/status/968404105495531520
It is something you want to avoid though.
Not a full-blown #chartcrime but i think a misdemeanour #Axiscrime here from Deutsche pic.twitter.com/ms6VghRb6C
— Paul McNamara (@M_PaulMcNamara) October 13, 2017
When the reaction upon looking at a chart is “What am I seeing here?” it’s clear there’s a problem even if no explicit rules have actually been broken.
https://twitter.com/VegaVandal/status/919936186315046912
So you’ll have to forgive me for not providing a checklist.
The chart above has X and Y axes. They’re labeled. The fonts are somewhat legible, and the eye can trace a process from strategy through logarithmic growth to culture.
The brain still has no idea what is happening inside of it.
and on this note, think i’m gonna log off now pic.twitter.com/0mLn87hEEU
— Sam Ro 📈 (@SamRo) January 18, 2018
The only antidote is: Do less.
In the particularly egregious examples above, the problems arise because extra effort was deployed to the wrong location. Making charts can be simple. I’m not aware of better guidelines than those on the “Data Is Beautiful” subreddit:
- Label everything.
- Make sure it’s readable.
- Include the data and code behind your graphic where possible.
- Be mindful of people who see differently due to colorblindness, disability, and other factors.
Undergirding all of this is simple guidance: Tell the truth in a way that can be understood. Our profession will never fully gain the public’s trust if we can’t speak clearly.
If you liked this post, don’t forget to subscribe to the Enterprising Investor.
All posts are the opinion of the author. As such, they should not be construed as investment advice, nor do the opinions expressed necessarily reflect the views of CFA Institute or the author’s employer.
Image credit: ©Getty Images/erhui1979
Thanks! Made my day. Graphic versions of “If you can’t dazzle them with brilliance, baffle them with BS”.
Those overlays look like good starts on a few conspiracy theories. I can just hear the creators saying, “Coincidence? I think not!”
And anyone that sees a straight line drawn through noisy data needs to be familiar with this- http://www.tylervigen.com/spurious-correlations
Thanks Charles! I agree. Any self-respecting conspiracy theorist will have no shortage of these handy. Appreciate that link to Tyler’s piece — it’s excellent.
The Colgate chart could have been from any number of big brand consumer non-durable companies whose sales and “growth” have flattened in the 21st c. Most are guilty of chart crimes, and you will notice there are never charts of “goodwill”, inventories, or coming long-term debt maturities. The most important role of a chart is as a reminder to look at balance sheet trends. Fun article.
Thanks Sloane – that was great fun to read. If you’re interested in more comically bad charts, you may want to look up a guest chapter I wrote for a colleague’s book, which was reviewed by CFA Institute:
https://blogs.cfainstitute.org/investor/2013/01/22/book-review-the-effective-investor-the-definitive-guide-for-all-south-africans/
I have a large collection of astoundingly bad charts created by supposed professionals. One of my favorites transpired when a chartist switched from Lotus to Excel but failed to realise that the dates of his data were no longer in reverse order. He used his ruler and so on to attempt to forecast trendlines and future prices. The chart he disseminated clearly showed the X-axis going from today (left) to back in history (right), but he simply didn’t notice this. Being an ardent skeptic of all technical analysis, I checked back further in history and guess what – his chart couldn’t even “forecast” the past with any accuracy!