Wednesday, August 10, 2011

Why David Trainer is an Idiot

I was reading Don't Fall For the Market's Head Fakes by David Trainer on Forbes.com this morning. At least I was until I reached that graph. I have a major problem with this graph: the scale. If you look at the left-hand side of that graph, you'll see the levels for large-cap stocks in increments of 500, each of them evenly spaced, which I've reproduced in Figure 1.

 Figure 1: Scary-looking, huh? Not really--it's blatantly misleading.

The problem with that scale on the left is that an increase from 500 to 1000 means it doubled in value. The difference between those numbers is 100%. At the top of the scale, from 3000 to 3500, the value went up a mere 16%. A 500-point change in 1926 was a lot more than a 500-point change in 2010, and not at all comparable. The scale should have been logarithmic for apple-to-apple comparisons between 1926 and 2010.

If you look closely at the graph, you can even "sense" that this graph is bull\$#!! The Great Depression started in 1929. The Dow fell from a high of 381.17 in 1929 to a low of 41.22 in 1932--a staggering 89% drop in value. But look at this graph during that time period. That black line representing the stock market looks absolutely flat. If the Dow fell the same number of points today--339.95, it amounts to a 3% drop in value. How useful can a graph possibly be when a 3% drop in value today is equivalent to an 89% drop in value in 1929-32?

And that "trendline"--how is it possible that a "trend" can overshoot the value it's supposed to be trending (and by such an enormous margin!) from 1926 through 1994? To show how ridiculous this trendline is, I took a segment of the graph above for use in Figure 2.

 Figure 2: The trend--from 1926 through 1980, seems to be quit a bit off from the  actual values it's supposed to be trending. Could it really be off by this much for over 50 years?!  Is there anyone during these 54 years who ever showed such a graph to explain why stock  prices were so incredibly undervalued for over 50 years? No--because that "trend" wasn't created until 2010!

But what would a properly drawn graph look like? This is the second problem I have with the graph: The small print. Sources:   New Con­structs, LLC and Ibbot­son Ibbot­son, 2010 Ibbot­son Stocks, Bonds, Bills and Infla­tion Val­u­a­tion Year­book, (Chicago: Morn­ing Star, 2008), 228–229. *Large Cap Stocks as defined by Ibbot­son are the best com­par­i­son for the S&P 500, which did not exist as it does today in 1926.

So basically, this scale is completely made up using a source I don't readily have access to. It's true, the S&P500 did not exist today as it did in 1926, but the Dow Jones Industrial Average did. Why not just use that? The Dow has always been made up of a large cap stocks, and it's an index I can look up values for and graph with ease on Yahoo Finance.

 Figure 3: A properly scaled graph of the DJIA from 1926 to today.

Looks a lot different like this, huh? You can quite clearly see the stock market crash from the Great Depression early in the graph. Now that must have been painful! A similar crash today would send the Dow plummeting nearly 10,000 points to about 1,000.

And what about the trend line? Unfortunately, Yahoo doesn't have a slick option to "insert trendline" into their graphs, but I figured I could eyeball one myself and add it in. Can't be any worse than the graph provided by David Trainer. So I added a trendline for Figure 4.

 Figure 4: My custom-made trend line. Fits a lot better than David's trendline, doesn't it?

There are statistical methods available to get a statistically correct trendline, and while I don't claim my trendline is 100% accurate, just looking around, it feels intuitively correct. Before the Great Depression, at the end of the go-go '20s, the stock market is now widely considered overvalued and speculative. (At the time, it wasn't so obvious, but in hindsight, it always is!) And my trendline shows that. At the greatest depths of the Great Depression, stocks were considered cheap, and my trendline shows this.

For pretty much all of the 1980s, my trendline shows stocks to be undervalued, and had you loaded up then, you'd have done pretty well for yourselves. But by 2000, stocks were wildly overpriced at their worst values since the Great Depression, and again, my trendline shows this.

And during the market meltdown that hit bottom in 2009, my trendline shows that stocks fell well below normal values and only recently returned close to the trendline.

So while my trendline may not be scientifically precise, it does seem to be more-or-less correct. And really, when it comes to the stock market, it's better to be approximately correct than precisely wrong. =)

So what does this mean for the future? I won't make any claims about what will happen next week or next month or even next year. I have no idea what will happen. But I do know this: That trendline is going up. Through a Great Depression, two world wars, through 9/11--it keeps going up. If the trend continues, the Dow could hit 20,000 in the next 10 years. Maybe more, maybe less, depending on the state of mind of investors at the time. But whatever the actual value turns out to be, it's almost certainly going to be higher.

And that's why I think David Trainer is an idiot. He can't even look at a simple graph and say, "You know, this graph is misleading and deceiving." All those clues in the graph never made him sit up and think critically.

But that's okay--it's people like him that scare others out of the stock market, driving prices down, and allowing me to purchase stocks that I believe are significantly undervalued. Just because I think he's an idiot doesn't mean I don't like him or appreciate his efforts in allowing me to buy stocks on the cheap. Thanks, Dave! =)