“Eerie”! A “Scary Parallel”! “Haunting”! No, I’m not conjuring accolades for the latest Hollywood horror flick. These are all descriptions cropping up online of a recent chart that’s gone viral in the financial media and blogosphere. A sensational chart comparing the movement of the Dow Jones Industrial Average for the period 1928 - 1929 and the period 2012 – Present. Viewers are to believe the same pattern is developing—implying a great crash looms. Frightening! But I’d suggest there is a different, more accurate description: misleading. And for more than just one reason—here are five. What’s more, after you weigh the five reasons I believe this chart’s purported predictiveness is malarkey, I’ll discuss the real takeaway: This whole episode actually seems much more bullish than bearish.
Exhibit 1: The Misleading Comparison of 1929 to the Present Day
Source: Factset. Dow Jones Industrial Average price level for the periods 02/20/1928 – 12/31/1929 and 07/02/2012 – 02/11/2014. Reconstructed according to the criteria in the chart featured here.
Problem #1: Scale. As several bloggers have already noted, the Y-axes on the chart have been manipulated to make two time series look similar—they zoom in on each at whichever magnitudes necessary to make the comparison “work.” The percentage moves in the two time series—which are what ultimately determine how similar they are—are actually radically different. In Exhibit 2, I’ve plotted both of these indexed to 100 to eliminate this factor by putting all the data on one y-axis. The similarities vanish. Some have refuted this scale issue, suggesting that magnitude isn’t the point. But that statement is false on its face. Let’s try a little reductio ad absurdum: If I hypothetically plotted two lines suggesting that now looks eerily similar to a 2% down move in early 2013, would anyone care? I’m betting not—magnitude matters.
Exhibit 2: The Comparison with the Dow Indexed to 100 at Both Period’s Start
Source: Factset. Dow Jones Industrial Average price level for the periods 02/20/1928 – 12/31/1929 and 07/02/2012 – 02/11/2014 with both series indexed to 100 at the start.
Problem #2: Date range. Why was July 2, 2012 chosen as the starting point for this present-day series in the comparison? It’s not the beginning of a bull market. It’s not the bottom of a bull market correction. No major developments in the eurozone. In the US, talk was still mostly of the Supreme Court’s upholding of the Affordable Care Act from the prior week (July 2, 2012 was a Monday). There is only one reason I can identify: When you select this date and incorrectly plot the Y-axes, the charts look similar. Pure data mining.
The starting point for the data series in the ‘20s is equally odd. It is neither the entire calendar year, the beginning of a cycle nor the entire set of available data. With the Dow having daily price history dating back to the dawn of the 20th century, one clearly could have gone further. Yet this graph seemingly begins in mid-February. As Exhibit 3 shows, if you use all of calendar 1928 and 1929, the comparison again falls apart. The date ranges here are purely arbitrary and selected to make the comparison seem direct.
Exhibit 3: Illustrating the Selectivity of Start Dates
Source: Factset. Dow Jones Industrial Average price level for the periods 12/31/1927 – 12/31/1929 and 07/02/2012 – 02/11/2014.
Problem #3: It’s the Dow. The Dow Jones Industrial Average is a flawed index. We’ve discussed why many times on this website (like here). It’s price weighted and consists of only 30 stocks, chosen in a rather unscientific manner. The S&P 500 is a better, broader gauge constructed based on market cap—meaning, the largest firms impact market movement most. Exhibit 4 compares the S&P 500 Price index over the two arbitrarily selected time periods, indexed to 100 to avoid the aforementioned double y-axis issues, and compares to the pattern to the Dow. As with the Dow, when eliminating the y-axis issue, the comparison dies. The only major difference between S&P and Dow plots is magnitude.
Exhibit 4: S&P 500 Price Index Level Comparison, Both Series Indexed to 100 at the Start
Source: Factset. S&P 500 Price Index and Dow Jones Industrial Average level for the periods 02/20/1928 – 12/31/1929 and 07/02/2012 – 02/11/2014. All four series indexed to 100 at the first date in each.
Problem #4: Past performance isn’t indicative of future results. History is a useful input in understanding fundamental and sentiment drivers for stocks, but assuming a past chart predicts the future is faulty logic. A chart only depicts the past, pure and simple. And yesterday’s price movement has absolutely no bearing on today’s, tomorrow’s, next week’s or next month’s. Maybe we see some volatility. Maybe we don’t. But whatever the outcome, rest assured it will not be because past returns led to the result.
Problem #5: You can run this type of comparison with virtually any time period. Once you decide to play loose and fast with actual numbers, scale, indexes and time frames, you can literally compare any two line graphs. And the punditry often has! In 2010, it was en vogue to claim the S&P 500 was forming a shape similar to 1937-1938. If you know your market history, this is 1929’s sister crash. Those charts didn’t necessarily suffer all the same problems as this current one, but suffice it to say heeding its advice would have been very costly indeed. This is not an isolated scenario, either.
In my view, the chart that has gone so viral on Wall Street is basically a behavioral error, illustrated. It’s the finance version of color-by-numbers. Fun! But not actionable for investors. But there is one very pertinent, relevant thing the misleading-chart-gone-viral episode shows us about markets today. When bull markets approach their typically euphoric peaks, most folks will latch on to virtually any reason to remain bullish, no matter how ridiculous. (Like commonplace suppositions in 1999 that it was a “new economy” where cash mattered less than clicks.) But an irrational reason to be bearish going viral—and being referred to seriously in major financial media sources—would seem to lay bare the claim we’re in a euphoric bubble. In that way, a terribly misleading, supposed-to-be-spooky graph is actually a sign there is likely more bull market ahead.