How to Analyze Alternative Facts and Ferret Out Fake News

Fake news! By now you've likely heard it's everywhere, the scourge of our times. Facebook, considered the main conduit, is on the warpath, working with outside fact-checkers to discover and label fakery. While the concept of fake news is relatively new to the mainstream, it has had a presence in the financial media for decades. It is the stuff of newsletters warning of the end of America, the great currency reset or hyping penny stocks, all urging you to buy gold/silver/life insurance/some proprietary product. But even within the respectable financial news universe, it is increasingly hard to separate fact from fiction. The more media outlets arise, the more they compete for eyeballs through sensationalism. As Robert J. Martorana put it a while back on the CFA Institute's website, "investment blogs have embraced the golden rule of tabloid journalism: simplify, then exaggerate. Pseudo news and pseudo analysis clutters the web, making it harder to stay well informed." As my colleague Michael Hanson wrote last week, "reading comprehension and testing narratives is a lost art, and it's one that's gaining primacy." So, how do you do it?

The first step is understanding what the news media really is: entertainment, targeted to a specific audience. If you've ever flipped through cable news, you probably get this. When you find a channel whose editorial slant matches your own political views, it makes you happy-you're their target audience. When it doesn't match your political views and raises your ire, you aren't their target. Someone else is. People often self-sort, picking the outlet aligned with their own views, and revel in the echo chamber. The editors tailor the message to keep viewers and attract more, with more sensationalism-entertain more people, get more ad revenue! It's a for-profit business, after all.

Newspapers have a more venerable reputation, but they too are entertainment-just packaged differently. Some recent New York Times navel-gazing laid it bare. One year ago, the executive editor asked seven staffers "to conduct a review of the newsroom and determine a blueprint for its path forward." Among the recommendations were "reducing duplicative layers of article editing, and having visual experts play 'the primary role covering some stories' - part of an urgent call for more visual journalism. The report also calls for a renewed focus on diversity within The Times as a way of ensuring that the paper's journalists 'reflect the audience we seek.'" Boldface mine, because that last bit is the kicker. They have a target demographic in mind and plan to tailor their approach to get it. The goal here is to attract, entertain and retain customers, not inform the populace in an objective manner.

This is why doing your own independent review to gauge a story's validity, regardless of source, is vital. It might be overwrought to say the people responsible for labeling fake news on Facebook will be the same who blur fact and fiction in their own writings, but only just. As The Wall Street Journal's James Taranto documented in his long-running column, Best of the Web, fact-checking outfits have a spotty record whether they are independent or units of major publications, and the checkers' biases and presumptions regularly bleed through. Sometimes, their "fact-checking" amounts to a long-winded examination of the source's reliability or trustworthiness-using the ever-dubious ad-hominem argument rather than evaluating the statements on their own merits, using actual data.

Even articles appearing to be straightforward reporting are often more opinion than fact. Traditional journalism relied on the Five Ws: Who, What, Where, When and Why (or How). There was always some opinion and sensationalism. But over time, cutthroat competition for eyeballs, and hence, ads, has caused more of it. This is readily apparent in reports on market movement or economic data. The actual facts-the market's return or economic indicator's latest reading-takes perhaps a sentence. The rest-color, quotes, explanations-are all opinion. Adjectives and adverbs often signal that a statement is an opinion. "The US economy grew in Q4" is a fact. " The US economy grew too slowly" or "too slowly to create jobs" is an opinion (the latter an opinion and a forecast). Opinions may be right, they may be wrong, but they are not facts.

Hence, separating fact from opinion is one of the first things we do when reading an article. Facts are things like economic data, market returns, interest rates, changes in Fed policy, the text of new laws and regulations-dry information. Opinions include explanations of why markets did A or some economic data point did B; interpretations of whether A and B are good or bad; analysis; forecasts; qualitative statements like "Stock XYZ is oversold;" explorations of whether a company/sector/country is a "buy" right now. There is nothing wrong with opinions-heck, MarketMinder is an opinion website-but determining whether an opinion is actionable or sheer noise requires careful evaluation. We do this daily in our Headlines section.

I said earlier that we aren't fans of ad hominem arguments, but when you evaluate an opinion piece, it is crucial to understand the writer's biases, qualifications and expertise. For example, some noted economic opinion writers are strong partisans, and they often interpret data and evaluate policy through a partisan lens. That doesn't invalidate their opinions, but it does put the onus on you to see where bias might be influencing their take. When you read market forecasts, check whether the person issuing them actually manages money. Not only does a professional likely have more expertise, but they also have an extra layer of accountability. You can see whether they put their money where their mouth is. If someone is warning of imminent collapse, yet they or their firm remains invested in stocks, how seriously should you take their prediction? On the other end of that spectrum, with the least transparency and accountability, are anonymous bloggers. They have no credentials, no editorial filters and no way for you to evaluate them as sources.

As for evaluating the argument itself, we recommend using the scientific method and testing the evidence presented. We often reverse-outline articles we read, noting their thesis and evidence. Are the opinions based on fact-or speculation and more opinions? We've seen too many articles with a strong warning of Bad Thing X, with the only evidence being the possibility that Jimmy Politician does something. There is no discussion of the probability that Jimmy Politician does this horrible thing. Markets move on probabilities, not possibilities.

Articles that include data should cite their sources, giving you an avenue to check the numbers-not only for accuracy, but to see what data the article might have omitted. People cherry-pick numbers all the time, including those that support their argument and discarding those that don't. When we analyze articles for our Headlines section, we regularly re-run the relevant data and look for things the article might have overlooked, and often we find nuggets that (in our opinion!) render the argument null and void. We're always asking critical questions: Did they take one month out of context, ignoring a contrary longer-term trend? Are the data seasonally adjusted? Does a more detailed breakdown of the data show something the article missed? (We saw this regularly as inflation picked up last year-very few tied it to oil or noted that "core" inflation, which excludes oil, was stable.) Are market returns from multiple countries quoted in US dollars, or each country's local currency-and if the latter, does the argument fall apart when you make them all match? Now, we have professional data tools that make this easy, but most market and economic data are publicly available. We rounded up a few dozen of them on the data page at www.FisherInvestments.com, which we hope you'll bookmark.

Some articles feature charts, which are both wonderful and dangerous. Wonderful because you can view long-term trends. Dangerous because it's pretty easy to manipulate a chart to make it look like two things are related. A couple years back, we dissected a chart that went viral because it purported to show the Dow's rise in 2012-2014 tracking the Dow in 1928-1929. The implication was that stocks would soon suffer an epic crash. Scary! But the chart was mere art. The Y-axes were manipulated, with mismatched scales, to make the series look alike. When scaled properly, it was abundantly clear there was no similarity. More recently, we've seen similar graphical funny business trying to relate the Fed's balance sheet to the S&P 500. Same trick-fudged Y-axes-and when you plotted them correctly, the argument that stocks correlated with Fed bond holdings fell apart.^[i]

We could write a whole book on evaluating data properly, but thankfully we don't have to-someone already did. Darrell Huff's classic, How to Lie With Statistics, should be required reading for anyone with a stock portfolio and an Internet connection. Used copies go for a couple bucks.

One other biggie: Avoiding confirmation bias, which is the impulse to accept commentary you agree with and ignore anything you don't. This rears its head in political writing all the time, but it's an issue in financial writing as well. If you think stocks are set to fall, you're naturally inclined to think articles saying stocks will fall are very smart, and bullish articles are wrong and uninformed. Maybe! Or maybe your bias is blocking you from important information! Always take time to read stuff you disagree with, as well as stuff you agree with. And don't just write it off. Explore the argument-identify the thesis and evidence, and test test test. In a recent Bloomberg piece, economist Tyler Cowen suggested taking time periodically to write a defense of an argument you disagree with, and we think that's a dandy idea. It trains the mind to think more objectively, and will increase your total understanding of the subject matter at hand.

Perhaps you're now sitting there thinking, "golly gee whiz, this stuff is hard!" You're right! Thankfully, there is a growing number of resources devoted to helping investors read better. Here are some of our favorites:

How to Read Financial News Headlines
(Ben Carlson, A Wealth of Common Sense)

How to Read Financial News: Tips From Portfolio Managers
(Robert J. Martorana, The CFA Institute)

How to Read
(Morgan Housel et al, The Collaborative Fund)

Ways to Burst Your Filter Bubble
(Tyler Cowen, Bloomberg)

How to Read The Wall Street Journal
(Michael Hanson, MarketMinder)

Market Insights Podcast: How to Read the Modern Financial News
(Todd Bliman, MarketMinder)

Memoirs of a Markets Reporter
(Chao Deng, Columbia Journalism Review)

Use the News: How to Get the Most Out of Financial Media
(Barry Ritholtz, The Washington Post)

Five Tips for Savvier Consumption of Financial Media
(Joshua M. Brown, The Reformed Broker)