- Dividends are a critical component of total returns for equity investors. Over the last 20 years, they have accounted for around a quarter of the average monthly total return to the MSCI World index, while contributing practically no risk.
- There is no discernible relationship between the two components of total returns – price returns and dividend returns. Dividends are stable over time, and operate independently of price returns.
- The distribution of dividends is approximately normal, but the yield has always been positive (as one would expect), and shows some evidence of skewing to the right. Surprisingly, this skew does not appear to coincide with lower price returns.
- Finally, although inflation is uncorrelated to both dividend returns, and price returns in the short run, longer run evidence suggests that the Value equity style outperforms the Growth equity style in moderate to higher inflationary periods.
We hope they are not your only source of pleasure, but Rockefeller was right about one thing – dividends matter. In this paper, the first of a series of four, we will look at the relationship between dividends and inflation over the last twenty years, and then longer, sharing our insights for investors as we live through what could be, for many, their first experience of a higher inflation regime. In subsequent papers we plan to look at the relationships between dividends and interest rates, sectors, and factors. We hope these four papers will equip investors with some new insights, some discussion of known dividend theory, and of course with a clear understanding of the interplay between dividends and these four important financial topics. Our view is that better understanding can lead to better investing. And better understanding starts with the data, and the facts.
“Do you know the only thing that gives me pleasure? It’s to see my dividends coming in.”
– John D. Rockefeller
The Data
Whenever one looks at long run financial relationships, one must choose between having a large enough data sample, i.e., going back as far as possible, while at the same time recognizing that the world changes, potentially making the older data less relevant. There’s no right or wrong answer to that problem, but, in this paper, we take a reasonable approach by first looking at the monthly returns of the MSCI World index over the last twenty years (this is a market cap index of global developed markets, effectively the US, Europe, Australasia, and the Far East). We think this has the merits of capturing enough business cycles, financial crises, and inflation regimes to make our analysis both sufficiently rigorous and relevant. We later contrast this to a longer data series in the US, and draw some interesting distinctions.
Our original data had price returns and total returns for the MSCI World index, and we use the difference between these two as the income or dividend component of returns. We note there are many ways to define and calculate dividend yields, but, for us, taking the difference between price and total returns is simple and reasonable. Finally, our inflation data comes from Eurostat and FRED, and we use month-on-month price changes in both the US and Euro Area (EA). When we convert from nominal to real, we simply deduct the relevant inflation rate for that period, though, again, we recognize this can be done in different ways.
The Facts
Figure One presents some of the summary statistics from looking at the monthly returns of the MSCI World index over the last twenty years. For us, there is a critical insight from this table. Looking at just the nominal returns in the three columns on the left, one can see that the average monthly total return over this period was 0.79%. In addition, the next two columns show the average breakdown of this number, with around 0.59% coming from the price movement of the stocks, and the remaining 0.20% coming from dividends. So, on average, over the last twenty years or so, price returns have accounted for around 74% of an investor’s total return, with dividend or income return (we will use the two terms interchangeably), the other 26%.
This return attribution may not come as too much of a surprise to readers, but the risk contribution of those returns might. Note that effectively all of one’s risk from investing in the stock market over this time came from price variability, with the risk of dividends so dwarfed by that risk as to be effectively inconsequential. The monthly volatility of the dividend return is 0.10%, which compares to 3.75% for the price return (note for risk decomposition we convert to variance so that the numbers sum to 100%). This result is easier to see visually. Figure Two shows the actual monthly price and dividend returns over the time in question, and then Figure Three shows how those returns have compounded over time.
In Figure Two, one can see the stark difference in risk from the two components of return, with the price return swinging around in quite a volatile profile – though still providing a positive outcome on average – equity investors only accept risk for a good reason! The dividend on the other hand is much more stable. It wasn’t negative at any point over this period (in theory one could witness a negative dividend yield, but it would imply a vast amount of equity issuance from many companies at the same time), and its relatively low volatility is clear to see.
Figure Three shows the impact over time of these two quite distinct types of return. The dividend component accretes slowly and steadily over time, whereas the price component swings around more wildly (even underperforming the dividend return after several years).
However, the end results are as expected – in the last twenty years, the total return has been positive for the investor, the riskier component of those returns has delivered the lion’s share of that return, but the dividend stream has played the key role of a potentially safer port in the stock market ocean.
