The Cubs are in the World Series—if that doesn’t give a guy license to make some bold predictions, what does? Here is what I offer to you tonight: Between October 26, 2016 and October 26, 2026, the total returns of IBM will exceed the total returns of the S&P 500 Index.
It is a theory of mine that general investor behavior mirrors the “general fighting the previous war” metaphor in that the most recent successes are leaned upon to create the next round of successes, and the recent failures are abandoned because, hey, the definition of idiocy is doing the same thing over and over again and expecting a different result, right?
Admittedly, I base my theory on a collection of anecdotal evidence. Nearly every person who has ever talked to me about retirement investing has indicated that they have chosen their retirement funds exclusively on fees and a look at recent performance (with five-year investment performance history seeming to carry the most weight.) That may be a flawed strategy under ordinary conditions, but I find it especially flawed now because the past five years have not included a recession so some high-risk junk has been topping the charts of some investment returns.
But a reason why you cannot base your investing decisions on previous performance is due to valuation. Investing would be a lot easier if it only required identification of companies with bright earnings growth prospects. Instead, you have to figure out: Do I want this high growth company at 27x earnings, or this company trading at 14x earnings at mid-single digit growth? Not only do you have to figure out growth rates, but you have to reconcile it with the current valuations.
This is why recency bias can be unhelpful in investing. When a business has nice earnings growth, it often sees its P/E ratio expand as investors develop a more favorable sentiment regarding future growth. Likewise, when business results suffer, the P/E ratio also contracts to reflect diminished future expectations. Over medium-term investing periods, the stock price changes not only reflect the changes in business results but also how investors value those results.
The subsequent danger is that recent performance can be unduly affected by a valuation shift. IBM is one such example. The past five years have seen IBM’s valuation pivot from 16x earnings to 10x earnings. This says nothing about changes in profits—it tells you that each dollar of IBM’s profits is valued 37.5% less in 2016 than 2011.
This valuation shift has taken IBM from the highest end of its historical fair value to a bit below the low end of its historical fair value. Imagine if IBM sees some reversion to the mean improvements and get 3% core earnings growth for the five years. Coupled with buybacks that reduce the share count by 3.5% annually, and you are getting about $17 per share in earnings in 2021.
And if the dividend grows at a rate of 6.5% for the next five years? The $5.60 per share payout will increase to $7.74, and shareholder will collect $34.16 in cumulative dividends over the next five years.
And if the valuation increases to 13x earnings, IBM stock will trade at $221 per share. With the $34.16 in dividends, the total value created by IBM will be $255.16. That is 11% annualized compounding on the back of growth only half of that.
Plus, I left a safety valve in my calculations. I assumed that the dividends did not get reinvested and just built up as cash. Well, you’re probably not going to let a cash payment that is received during the first quarter of 2017 remain dormant until 2021. Assuming it is put to work at a rate above 0%, my calculations are conservative regarding the dividend reinvestment component.
The expectations for IBM have become so low that even low single digit core earnings growth for five years will lead to double digit returns. IBM only needs 3% core earnings growth to match the 6% growth of the S&P 500 because it is retiring over 3% of its stock each year, plus its dividend is almost doubled that of the S&P 500. Meanwhile, the S&P 500 trades at overvaluation compared to its historical norms while IBM trades at a discount to its historical figures.
It is really weird to think of IBM as a dividend stock. Heck, it is close to offering the type of future performance that is analogous to the cash-generators I point out. I mean, you’re going to collect about $34 per share without reinvestment, and with reinvestment, that figure will be closer to $40 per share. Someone that purchased 36 shares of IBM in a shelter like a Roth IRA might collect close to $1,500 in cumulative cash dividends over the next five years. This isn’t because IBM’s future growth is so superior to everything else you’d find, but because, the profitability base is being deployed in a shareholder friendly manner and the current valuation is primed for future success. This change in starting valuation is why the 2016-2021 period for IBM shareholders will be superior to the 2011-2016 stretch, and also illustrates the dangers of raw hindsight bias without critical thinking attached.