“Secular Markets” Are The New Crazy Talk
The Globe and Mail recently ran this article which has some strange quotes. In it, a money manager makes the claim that the Canadian stock market has just entered a 12-year “secular cycle of under-performance”, and investors should focus on other markets.
His arguments for this are that the TSX index has had lower returns than the S&P 500 for the last 26 months with a cumulative difference in returns of 41% in US dollars, and has lost money so far this year (again in US dollar terms). Comparing stock markets on a currency adjusted basis makes sense (even though less-than-efficient markets can get in the way of this), but I don’t understand why he’s telling Canadian investors about returns in US dollars.
Because this under-performance has lasted for 2 years and historical studies show cycles that average 12 years in the last 60 years (that’s only 4 cycles so far), he believes the TSX will continue to do badly over the next 10 years.
Time For Some Actual Facts
That is a very weak argument. While a market that has run up a lot is less likely to continue doing well in the near future, a rapid correction over a year or two could easily reset it to a lower price where it can do well. That’s exactly what happened in 2008. We can look to valuations to get a measure of this. The TSX index currently has a lower P/E ratio than both the US market and the EAFE index (15.2 compared to 18.1 and 18.5 based on recent information from the funds we own). It has a dividend yield well above the US market, and close to that of the EAFE (2.53% compared to 1.97% and 3.15%).
Based on those numbers I have been considering moving more assets to the TSX. The gaps aren’t large and the numbers themselves aren’t that exciting on their own. But a combination of reasonable, competitive prices and the tax advantages that Canadian investors get makes it look like a good choice. If the gap widens for any reason I will be moving more into the Canadian markets.
Follow The Numbers
This has been the basis for many of my investment decisions and it’s worked well for me. I was just reflecting on this this morning when I noticed that I have had near-perfect timing with bonds over the last few years. I had a significant allocation a few years ago back when P/E ratios and bond yields were higher, which produced returns on par with stock markets as yields fell.
Then I cut back as dividend yields on stocks rose above the yield on bonds. It turned out this was just before bond yields started to bottom out and the total returns fell. I finally sold out the remaining bond funds 2 months ago because they weren’t improving and I was tired of them (a fortunate accident sealed the deal). This happened to be just before the bond market dropped, pushing them close to a negative 1-year return.
This is too good to be skill. I got lucky on a few things. But luck favors those who are prepared, and I prepare by buying things that have an attractive price and yield rather than following forecasts that are wrong more often than they are right.
The Dangers of Secular Forecasts
Secular cycles obviously exist. You can’t deny that there have been stretches of a decade or longer where certain markets have been either completely flat or going up steadily. The problem is that you’re only in a secular cycle until you aren’t. The most precise forecast, if someone happens to get it right by accident, is likely to be off by a few years. That could be enough to completely change the numbers if you buy into it.
Most people who talk about secular cycles are either doing this type of foolish extrapolation or doing multi-decade technical analysis, usually by showing how the last few years are similar to a chart from 60 years ago, and predicting that we will follow the rest of the chart. If technical analysis has any value at all it’s at a level of hours or days. Trying to do the same thing with a few decades of data is misleading because that will always include a few one-time world-changing events that won’t happen again, especially not on the fixed schedule of a technical analyst.
Another problem with this idea is that it’s based heavily on data-mining. The return over the last 26 months only depends on the prices on 2 specific days. Same with the returns for the next 10 years, and any other “proof” you want to throw in. A random change on those days, or shifting it by 1 year, can completely throw off the results which makes them meaningless. As Nassim Taleb wisely points out, if a small change in the inputs completely changes the results then your formula or forecast is worthless.
Even if this forecast is true it might be meaningless for the average investor who could pick up a lot of shares at good prices by buying a little every month for the next decade, and get dividends along the way too. If this forecast is true and dividend payouts hold up or rise, it could actually mean that the Canadian market is a great deal for long-term investors. Over a long enough period where nothing else changes, falling stock prices are always good. Many investors forget this.
What Can We Expect?
There are a few things I look at when deciding what investments are most attractive to me, including valuation and relative performance over the past decade. I believe this helps me buy more of the things that are unpopular and cheap today, and have the best opportunities for growth. But even that is a very weak prediction. A Vanguard study last year showed that valuations are only 40% correlated with the next decade’s returns. And that’s one of the best signals we have.
The opposite of these signals, which are sometimes useful but not perfect, is predicting that the trend from the last 2 years will continue for the next 10 years. The only redeeming point in this article is when it says that “nimble investors” could take advantage of “brief periods of leadership reversals within these long-term secular trends”. In other words they are admitting that this is a worthless prediction and we have absolutely idea what will happen over the next 10 years. And here I thought reporters were supposed to put the most important point in the headline.