Benjamin Felix September 14, 2019 Advanced Investing Personal Wealth The Irrelevance of Dividends: Still a Non-Starter Whenever I make a video or write an article about the irrelevance of dividends, the dividend-loving crowd is quick to respond in defence. Too bad for them, there is still no basis – at all, whatsoever – to prefer dividend-yielding stocks. And by the way, showing me dividend stocks with great past returns doesn’t work either. Dividend Investors Doth Protest Too Much Most of the arguments from dividend lovers boil down to being able to successfully select individual stocks. This is not something most people can do consistently. Even in a supposed stock-picking environment, there is no reason to believe that dividends, or the growth of dividends, would be an indication of a good stock to own. Before I continue, let me clarify: When I say dividends are irrelevant, I do not mean they are unimportant to total returns. They are important. What I mean, is they are irrelevant in determining which stocks may have good future returns. And one more thing. Some dividend investors have suggested online that we should all be nice to each other. These comments go something like this: “We should each invest however we feel comfortable. Neither dividend nor index investing is wrong. There’s room for both.” Nope. That’s not what I’m here for. I’m not trying to hurt anyone’s feelings, but there is peer-reviewed academic research and robust empirical evidence telling us that there is a right way for most people to invest, which is low-cost index funds. If your goal is a reliable long-term outcome, there is no better approach. Factor Exposures by Any Other Name The basis for dividend irrelevance starts with the 1961 paper “Dividend Policy, Growth, and the Valuation of Shares” by Merton Miller and Franco Modigliani. They explained in their paper, before frictions like trading costs and taxes, investors should be indifferent between a $1 dividend (which causes the stock price to drop by $1) vs. receiving $1 by selling some shares. This is a fact that must be true as long as $1 is worth $1. In support of this theory, we know empirically that stocks with the same exposure to factors like size, value, profitability, and investment have the same average expected returns, whether they pay dividends or not. Unless you believe that the market is irreparably broken, or that capital can be created out of thin air, there is no way to argue against the fact that the distribution of a dividend results in a reduction in share value. That must be true. There is no way around it. Dividend investors will tell you that theory does not extend to reality, and that dividend stocks do indeed do better than the market. I am not denying that dividend growth stocks have beat the market on average. But the dividends are not the reason why. Dividend stocks, particularly dividend growth stocks, have excess exposure to the value, profitability, and investment factors. That is what explains performance differences, which means the outperformance still doesn’t justify trying to pick individual stocks. Dividend investors will also tell you that dividend-paying company managers are positively affected by their dividend policy, which spurs them to produce better long-term results. There is no basis to believe this. For this idea to generate premium returns, the whole market would need to think the company isn’t going to do well, while you believe otherwise. Then you would have to be right. Everything comes back to pricing. All else equal, for future returns to be higher, the current price needs to be “too low,” meaning that the market is mispricing stocks with good dividend policies. There is no basis to believe this is systematically happening across the stock market. A Mathematical Analysis of Irrelevant Dividends With this in mind, let’s walk through some math, starting with an example we will return to later. Let’s say you own 10,000 shares each of two companies: Company 1 pays dividends. Company 2 does not. To keep things simple, we will assume both companies trade at their beginning book value of $10 per share. Keep in mind, following valuation theory, the market price is based on the company’s book value plus the value of its discounted future profits, discounted at some discount rate. This needs to be crystal clear: If two companies have the same expected future profits, and the market is discounting those profits at the same rate, the two companies would be expected to have the same price relative to book value. Similarly, if two companies have the same profitability and the same relative price, they must also have the same discount rate. The discount rate is the investor’s expected return on the stock. This relationship holds true whether or not the companies pay dividends. It is crucial to understand that we are assuming that Company 1 and Company 2 are the same size, have the same profitability, reinvest at the same rate, and trade at the same price relative to book value. To keep it simple, we also are assuming the price equals the book value. If all of this is true, then these two companies have the same expected return. So far, so good. Next, let’s assume the two companies, with a starting book value of $10 per share, each earned $2 per share. Your Company 1 paid a $1 per share dividend, while my Company 2 paid no dividend. Your Company 1: You’d still own 10,000 shares, and have received $10,000 in dividends. Your shares are now worth $11 each: the $10 starting value plus $2 in earnings minus the $1 dividend. Your total portfolio consists of $110,000 in stock, and $10,000 in cash. My Company 2: I’d also still own 10,000 shares, but they’d now be worth $120,000. If I needed some cash, I could sell some shares. Maybe I would create my own “dividend” equal to yours – but I don’t have to. I could sell more or less shares as needed. I am not allowing the company’s dividend policy to dictate my spending. All of this is true whether the stock market is up or down. If the market is down and a dividend is paid, the value of the company still falls by the amount of the dividend. It has to. This is not up for discussion, unless you don’t believe in mathematics. Receiving a dividend in a down market is exactly – and I mean exactly – the same as personally selling off some shares in the same down market. A Performance Analysis of Irrelevant Dividends Again, if dividends are irrelevant and it is only risk-factor exposures that explain differences in returns, we would expect two funds with similar exposures to have similar returns, regardless of their dividend policies. So, let’s compare the Vanguard Dividend Appreciation ETF (VIG) to two Dimensional Fund Advisors funds, combined to create comparable factor exposure to VIG. Dimensional funds are total market funds that seek exposure to the known factors; they are blind to dividends. I am using U.S. funds, so I can run the analysis using the Match Factor Exposure tool at portfoliovisualizer.com. For the period January 2013–May 2019 (which is when the data for the newest of these funds starts), the annualized return for VIG was 9.8%. The annualized return for the Dimensional funds portfolio was 14.67%. The risk-adjusted returns were also higher for the Dimensional portfolio. The R-squared for VIG was 94.3%, which means that exposure to the factors explains almost all of its returns. The R-squared for the Dimensional portfolio was 99%. The point isn’t that the Dimensional funds beat VIG. It’s that two portfolios with similar factor exposures produced similar results, as expected. Second Verse, Same as the First I don’t know what else to say. Dividends do not explain future returns, and limiting your opportunity set to the stocks that pay dividends cuts it roughly in half, because roughly half of global stocks do not pay dividends. Less diversification means more dispersion, which reduces the reliability of your outcome. I think that dividend investors should just admit they are nothing more than stock-pickers. And picking dividend-paying companies does not make picking stocks any smarter or safer. I will concede that favouring dividend-paying growth stocks tends to tilt a portfolio toward large-cap value stocks from companies that exhibit robust profitability and conservative investing. So, picking dividend stocks is probably better than picking penny stocks. But you can’t get away from the fact that you’re still taking on a ton of unnecessary risk based on selecting individual securities. Again, to be clear, I have nothing against dividends. They are an important component of returns. I just think it’s egregious to think you can use dividends to pick winning stocks. And again, please do not tell me the names of individual dividend growth stocks with great past returns as evidence that dividends matter. It would be just as relevant to tell me the names of stocks starting with the letter “A” with great past returns. The analogy is sound because, in both cases, the information is meaningless. The Toll of Taxes: Dividends vs. Capital Gains I do have something else to add. In Canada, Canadian dividends are taxed more favourably than any other type of income when you have a low income. This is often used as an argument for Canadian residents to invest in Canadian dividend-paying stocks. I believe the situation is similar in other countries as well. Let’s look at an example. Say you already had $47,630 of taxable income in Ontario in 2019. From there, your marginal tax rate on capital gains would be 14.83% and your marginal tax rate on eligible dividends would be 6.39%. It seems obvious that the dividends are way more tax-efficient. But hold on. Let’s bring back our example from earlier. If you received $10,000 in dividends from your Company 1, you would owe $639 in tax. In contrast, If I received $10,000 in unrealized capital gains from my Company 2 (half of the $20,000 increase in value). I would not pay any tax on those capital gains until I sell. But what if I need some income, and I sell some of my shares? Here is where people get really confused. If I sell $10,000 worth of my shares, I am not paying tax on the full $10,000 capital gain. I am only paying tax on a proportional amount of the gain. Let me explain. I paid $100,000 for my shares. They are now worth $120,000. If I sell $10,000 of my position, the realized capital gain is only going to be $1,666, which is the proportional amount of my total $10,000 gain. The tax due would be only $247. This amount will increase over time as I accrue more unrealized capital gains, but the crossover for dividends being more tax-efficient will not happen for many years. Plus, I may be able to offset gains with losses. Plus, I remain in control of triggering the gains. What if neither you nor I needed the full $10,000 of income this year? I just wouldn’t sell any shares. You, as a dividend recipient, would receive the taxable dividend anyway. Buffetting a Dead Horse I know I am repeating something I’ve covered in the past, but I can’t leave it out. Dividend investors will tell you to look at how much billionaire businessman Warren Buffett loves dividends, so we should love them too. Warren Buffett is happy to collect dividends. But he is also happy to not collect dividends. In his 2012 Letter to Berkshire Hathaway Shareholders, starting on page 19, he explicitly wrote about this to help you, the average investor, understand why. Please go read this before following Buffett into dividend stocks. He has made it clear that this is not a criterion he uses to identify good companies. Buffett likes low prices. He is a value investor. Low-priced stocks will often have high dividend yields. That does not mean Buffett loves dividends. No matter how you cut it, using dividends to pick stocks simply does not make sense. It has no basis in financial theory. It can be disproven empirically, as we have done, by comparing dividend-focused index funds to factor index funds with similar factor exposures. If you want to pick dividend stocks – if that is where you are most comfortable – that is fine. But you have to understand that you are still nothing more than a stock-picker. There is no pedestal for dividend investors to stand on. Thanks for hearing me out. If you think it might help, please share this commentary with anyone who might benefit from it. And don’t forget to tune into weekly episodes of the Rational Reminder Podcast wherever you get your podcasts. Share: Facebook Twitter LinkedIn Email CIRO Advisor Report
David D'Amico 1:09pm July 13th, 2022 Reply There is a small, relatively obvious typo in the following statement regarding 98% – which should state 9.8%: “For the period January 2013–May 2019 (which is when the data for the newest of these funds starts), the annualized return for VIG was 98%. The annualized return for the Dimensional funds portfolio was 14.67%. The risk-adjusted returns were also higher for the Dimensional portfolio.”