With varied forecasts from the experts - who should you believe?
There are reliable bootstrap methods to estimating long-term returns to the S&P 500, such as the CAPE ratio or market cap to GDP ratio (in a previous article we explained how the 16-year CAPE is a good predictor of returns for the next 16 years), but to really understand future returns, we need to analyze the individual components impacting future returns.
There are varied estimates and time horizons provided by respected sources, but they don't really give the average investor a sense of how it actually functions, and with no consensus from experts, how do you decide which one more realistic?
Making 10-year return projections more accessible
Given this, I thought it would be useful if investors could visualize how returns get determined and also be able to experiment with their own forecasts, as a useful learning experience. We have created a simple discounted cash flow analysis spreadsheet (you can request it on our website) which contains everything you need to understand where future returns come from. But before we show the spreadsheet, let's briefly review the components of future returns.
Components of future returns
When you purchase an index fund you receive two streams of cash flows; 1) dividends and 2) proceeds when you sell the fund in the future.
Total return = Dividends + Terminal sale value.
To estimate these two items, we need to know three things:
- Dividends = ƒ (current dividend payout ratio and change in dividend payout ratio)
- Earnings = ƒ (current earnings and earnings growth rate)
- PE ratio = ƒ (current PE and estimated PE change)
We have the current values for all three but need to estimate how they will change over our forecast horizon - the three bold items.
This is what the discounted cash flow analysis looks like. It's a 10-year spreadsheet. There are four main sections; the first section calculates corporate earnings, the second calculates the P/E ratio, the third calculates dividends, and the last section consolidates the three sections into a net cash flow forecast from which we can calculate the projected compound annual return.
Explaining the mechanics of the spreadsheet
So, working backward, our end goal is to estimate the cash flows going forward for the next 10 years, line 38, which will come from dividends plus the sale proceeds at the end of year 10. Once we have those and our current price we can calculate the internal rate of return or projected compound annual return, line 40.
To get dividends and sale proceeds we need the three lines shaded in gray.
So the first gray line item, line 15, is S&P 500 earnings estimates. We use the economic data to generate an earnings growth rate for all corporations in the economy and apply that growth rate to S&P 500 earnings. We first need to estimate nominal GDP. We have the starting nominal GDP value, line 7, and going forward it is a function of real GDP growth rate, line 4, plus the inflation rate, line 6. Next, we estimate the change in profit margins, line 9, and with nominal GDP and profit margins, we can calculate corporate earnings, line 12, and the growth rate of those corporate earnings estimates going forward, line 13. We then apply that growth rate to S&P 500 earnings on line 15.
The second gray line item is the P/E multiple or earnings yield, on line 23. The earnings yield is the risk-free rate, line 19, plus the equity risk premium, line 21. For the risk-free rate, we use the 10-year government bond because we are looking out at a 10-year horizon. We calculate the current equity risk premium, line 21, as the difference between the current earnings yield and the current 10-year treasury yield, so 4.14% minus 1.6% which is 2.54%. We estimate the change in the equity risk premium, line 20, and the change in the risk-free rate, line 18, for the next 10 years, and from there can calculate the estimated earnings yield or P/E ratio, line 23, which is applied to the earnings estimate on line 15, to come up with our value estimate for the S&P 500 index on line 25.
Lastly, dividends, we know the current dividend yield, line 30, is 2.14%, and so we apply that to today's price to get the current dollar dividend of $44.94 for the S&P 500, which equates to a payout ratio of 51.7% of earnings on line 29. Then all we need to estimate is the change in the dividend payout ratio, line 28, and that will calculate our stream of annual dividends on line 31.
Finally, we put all the cash flows together. We have the current S&P index at 2,100 which is an outflow when we purchase it. And then our cash inflows, on line 35 and 36, will be the dividend stream going forward plus the sale of the fund at the end of year 10. Line 38 shows our net cash flows for the 10 year period, and the internal rate of return is calculated on line 40. And so that's how this works. You can see all the factors that impact future returns in one simple sheet.
Get the spreadsheet, it's simple to use
If you want to experiment with your own estimates it's very easy to use. There are two simple steps.
- Enter the current values highlighted in green - we've entered them for September 2016 - so you won't need to change those if you're using it close to that date.
- Enter the estimates for the six blue lines going forward 10 years.
And that's all you need to do and it will automatically calculate the ten-year total return estimate for you.
The values we have entered for the six blue lines represent what we think are reasonable estimates that you could expect, but of course, nobody actually knows. We can use history as a guide for what is a reasonable range of values for each variable as well as common sense based on where the current values are. Our estimates result in a projected 10-year total return for the S&P 500 of 1%.
Real GDP growth, inflation, the 10-year treasury rate and the dividend payout ratio are reasonably stable over time, and profit margins and the equity risk premium are more volatile and thus more difficult to estimate - and future returns are more sensitive to these two drivers.
Estimating 10 years out is difficult
It is difficult to estimate variables going out 10 years, but there are some resources to give you a helping hand. For example the Office and Management & Budget provides 10-year economic assumptions that they use for the government budget, so those are a good starting point, although not always realistic; for example, they currently have the 10-year treasury rate at 3.5% for 2017 which implies a doubling over current levels.
The Federal Reserve also attach their long range estimates to their minutes.
Lastly, look at the historical charts to get a sense of the likely range of possible values. Here is an example of the dividend payout ratio.
Experiment with different values for variables to test sensitivity of outcomes
For example, here's a very optimistic scenario:
Real GDP growth goes up to3%, inflation at 2%, and no change in any of the other variables from their current values. That would give a 10-year total return of 7%. It seems like wishful thinking, but the point is that even in a very optimistic scenario, you only get a 7% total return well below the historical average of around 11%.
The spreadsheet makes projecting future returns simple to understand. We provide some guides for estimating the variables into the future, and testing sensitivity under different scenarios will give you a lot more confidence in your expected outcome, and is a good learning experience.
I hope you find it useful.
Disclosure: I am/we are long SPY.
I wrote this article myself, and it expresses my own opinions. I am not receiving compensation for it. I have no business relationship with any company whose stock is mentioned in this article.