Stock Market Charts Without Context Only Feed Hysteria
(Bloomberg View) -- Lance Roberts, the Clarity Financial strategist and editor of the Real Investment Advice blog, recently posted the charts below to make the point that market declines are steeper than run-ups. It’s a fair claim that deserves some nuanced discussion, but the charts without text went viral, first on Twitter, and then on the internet more broadly. The reasonable-but-debatable point of the post was lost on the Chicken Little crowd screaming “History demands you sell your stocks now!”
The attention was a bit strange, because the charts are similar to ones produced by doomsayers for many decades. But there are problems, the first being one of scale. Instead of looking only at windows leading up to historic crashes, it’s best to look at the entire data series. The chart below shows the growth in purchasing power of $0.05 in 1871 (which bought about the same as $1 today ). Each horizontal gridline represents a doubling in purchasing power. This chart suggests more calm, more faith in the stocks-for-the-long-run argument.
The second problem is the selection bias of ending each chart at the crash. That’s like ending the Western movie with the hero pinned down and outnumbered in a gunfight while the ingenue is tied to the railroad tracks with the train approaching. Let’s look instead at the same periods and lengths, but center them at their peaks. The orange lines represent cubic fits to the data.
For those inclined to see patterns, the takeaway is that it’s not really the slow-build-up-sudden-crash suggested by the original charts. Rather there are periods of general uptrends, which tend to have relatively low volatility around the smooth orange line, interspersed with level or down periods, with much more volatility both up and down. I’ve never tested that rigorously, but it doesn’t strike me as a bad qualitative description.
That brings us to today. The chart from 2010 appears to show the uptrend losing steam, and an uptrend can only be followed by a downtrend. In the earlier charts, you’d be wise to get out when the orange line turned down, and stay out until it turned up. The problem is that those orange lines were drawn knowing the full period data. To be useful to investors, we must be able to predict moves before they happen.
Suppose you try the following Chicken Little rule. Fit a cubic polynomial (like the curves in my charts above) to the last seven years of data. If it indicates an uptrend, invest in the S&P 500; if not, stay in cash. This rule delivered returns of 4.0 percent per year after inflation versus 6.9 percent for buy and hold. Volatility falls to 8.5 percent versus 14.2 percent, and the Sharpe ratio is identical at 0.48.
The table below shows the 10 biggest drawdowns for both strategies. Chicken Little’s drawdowns were milder than a buy-and-hold investor had to endure. On the other hand, investing 58 percent of your portfolio in the S&P 500 and 42 percent in cash would have similar drawdowns, return and volatility, while saving a lot of trouble and taxes.
I believe that valuation matters, and that stocks are more attractive when cheap. However, equities remain the best choice for most investors for their long-term ability to outperform inflation. It’s tempting to try to avoid crashes, but unless you have a better model than drawing curves to historical data, you’re probably going to cost yourself at least as much return -- especially after taxes and transaction costs -- as you reduce volatility and drawdowns.
This column does not necessarily reflect the opinion of the editorial board or Bloomberg LP and its owners.
Aaron Brown is a former Managing Director and Head of Financial Market Research at AQR Capital Management. He is the author of "The Poker Face of Wall Street."
https://realinvestmentadvice.com/breaking-bull-market-07-28-17/ The data for the charts come from Professor Robert Shiller http://www.econ.yale.edu/~shiller/data.htm I used the same data, although I added in dividend yield.
As closely as the Bureau of Labor Statistics can estimate it.
The full period chart is actually as misleading as the shorter charts, but in the opposite direction. It makes those wiggles around the general upward trend seem unimportant when they in fact cause great financial distress, even to long-term investors. Maybe not to 146-year long-term investors, but to real life to 40-year ones.
Before looking at the charts themselves, consider how the bottom of the vertical axis increases from to then Clearly most of the growth is happening in the (relatively) crash-free periods between these events. So even if the behavior of the S&P500 around peaks is scary, that’s a bad reason to avoid stocks.
I don’t endorse predicting the market with cubic polynomials, but it’s a fairer way to describe market trends than just drawing a U-shape by eye at the end of each chart as in the original.
And I’m not. Your eye is really good a picking out patterns whether or not they exist. I prefer formal quantitative analysis.
So if you look at the orange lines, abstracting from shorter-term volatility, the up moves are steeper than the down moves. But greater volatility makes the actual blue-line price moves steeper both up and down in down periods.
Which I will assume earns the inflation rate.
Duck and cover completely avoided the Crash of the tech crash of the financial crisis, the crash of the Nixon shock among other unpleasant events.
For more columns from Bloomberg View, visit http://www.bloomberg.com/view.