Dr. Jeremy Siegel of The Wharton School of Business performed an important study on the long-term returns for investors that made lump-sum investments during the peak of a bull market that lasted at least four years. He measured lump-sum market investments in June 1901, September 1906, September 1929, February 1937, April 1946, and December 1968 and tallied these results over a ten, twenty, and thirty year period.
During the subsequent ten years, the results are disappointing: A $1,000 investment in these peaks collectively grew to $1,250.
During the subsequent twenty years, the results are still somewhat disappointing: A $1,000 investment grew to $2,150.
However, once the measurement period becomes thirty years, the returns are more moderate and come close to meeting basic investor expectations: A $1,000 investment grew to $5,900.
This, incidentally, is why I treat buy and hold as distinctly different decisions and do not subscribe to the “If you wouldn’t buy it, why would you hold it?” school of thought. Not only would the latter be a logistical nightmare–how could you ever build a diversified collection of assets for your family or avoid getting eaten by transaction costs?–but I also try to treat hold decisions as a recognition to the excellent possibility of strong returns over 30+ year time horizons while buy decisions try to avoid placement into those ten and twenty-year returns cited by Siegel.
For investors contemplating purchases right now, I think of it as a two-step process:
First, make a list of the top fifty or so stocks that you would want to own for the long term. You don’t have to start from scratch–I have provided a “Master List of Stocks” page on the site to aid in the initial discovery process. Some companies, like Johnson & Johnson and Colgate-Palmolive, tend to show up on every investor’s list. Others, like McCormick and Becton Dickinson, tend to vary with stock market experience and a history of paying close attention to the companies’ business models (e.g. did you know that McCormick has grown its global market share in the spice industry from 15% in 1970 to 40% today?).
Once you have your list compiled, the next step is to check for signs of overvaluation. For each company that you are seriously considering, I would look at the twenty-year P/E ratio history for the stock (provided it is not a cyclical) and then figure out what the average historical valuation for the company is. If the current price of the stock is more than 10% above the historical valuation, I would be hesitant to make the purchase.
It’s not meant to be a hard rule, but rather, a strong presumption. I’d rather pay 10% above a historical P/E ratio for Visa than pay 15% under a historically normal P/E ratio for Campbell Soup because of the superior growth characteristics of the former. At no point in the exercise are you excused from exercising independent judgment. But still, you should have a general wariness about overpaying for a stock based on historical P/E markers for the company.
For a lot of people, a ten or twenty year period will encapsulate a fourth or even half of their investing life. You don’t want to spend many of those years burning off the overvaluation, as the investors that came before you benefitted from more wealth than the earnings growth justified. You want to make sure that you capture as much of the company’s long-term earnings growth as possible, and you achieve this by being disciplined about valuation.
The good news is that there are over 15,000 companies to choose, and you only have to find one at a time with available capital to engage in successful investing. Even with large American stocks, there are companies trading at fair value. You’re not going to go wrong paying $90 for Hershey or $25 for General Electric. You stand to achieve even better results if you focus on energy stocks, though you must make sure that you can withstand the added short-term volatility associated with them (no strategy is worth pursuing if your emotional temperament would lead you to sell low.)
The difficulty of market peaks is that government bonds and T-bills still underperform stocks even during stock market peaks. Using those six data points of high stock markets, an investment at the peak still beat bonds and T-bills over all ten, twenty, and thirty-year levels. If someone wants better returns than those ten and twenty-year periods cited, it seems that the options are (1) hold cash and cash equivalents until a better opportunity to buy, or (2) look for individualized opportunities within the overvalued class that offer better ten and twenty-year prospects than the class as a whole.
I tend to prefer option number two because it provides income through dividends, rents, or interest at a rate that exceeds the short-term rate of holding cash. During periods like 1982 through 2000, you might be on the sidelines for a while. You had a decent opportunity in 1987, 1990, and 1991 to make investments, but they were unique blips of a few months or less that aren’t clearly buy points to me without the benefit of hindsight (especially the 1990 and 1991 markets).
That said, this is the art part of investing and I do not argue that my favored approach leads to the best long-term results. If you invest in fairly valued or undervalued stocks in 2007, you are unlikely to achieve better results than someone that held cash and actually deployed the capital in 2009, 2010, and 2011. The problem is that executing a strategy of waiting for a crash is hard to do in real life and incorporate as part of a lifetime strategy.
The post thing I’ve encountered on the subject of this type of market timing comes from Tweedy Browne, which recommended investing in three laddered installments. If you hold 20% of your portfolio in cash and seek to maintain a 5% cash minimum at all times, it might make sense to deploy five percentage points of your cash position in response to each 10% decline in the price of stocks transitioning from fair value to undervalue.
The downside of this strategy is that it leaves you with little capital when stocks fall more than 30% and presumably represent the best deals, but also lets you take advantage of the typical period when recessions are moderate in nature (e.g. if you were waiting for a 40% or more decline to start investing, you would only be able to invest in twelve years of the past 120 years. That kind of timing necessary to execute the strategy is off-putting enough to be a non-starter, though is quite lucrative if you could actually pull it off.)
Mean reversion is a powerful concept. When P/E ratios get higher than usual, there is a consequence in the form of lowered future returns. These lowered future returns are not enough, however, to make bonds a superior asset class to stocks. And once the holding period stretches out to thirty years or more, the high starting valuation of the stocks starts to become immaterial as the earnings growth of the company and the dividends paid out exert a disproportionate effect on stock market returns compared to P/E changes. All buy decisions warrant extra scrutiny of valuation at this point in the market cycle, and the good news is that it only requires the identification of one company at the time you have available capital to be successful over the long term.