Tuesday, September 22, 2009

Evaluating Equities

While the days of buy and hold may be over, below is a simplified way to think about forecasting equity returns over the long haul that I've found useful. I'll walk through the methodology, show some supporting charts, then detail my expectations for the equity asset class over the mid-to-long term. I'd love to hear all of your thoughts.

Equity Returns: In Theory
The earnings of a company are what you own as an investor and earnings are paid out to investors through dividends. Another way to say this is that the value of a company is the present value of all future dividends. Thus, equity returns = dividend (or corporate earnings) growth + the dividend yield (what isn't returned to an investor in the form of a dividend now, is reinvested, and grows at that same growth rate as earnings).

Quick and Dirty Estimation of Earnings Growth
A quick and dirty way that has been used to forecast future equity returns is to use nominal GDP growth as a substitute for earnings growth because in theory, earnings growth should be roughly equal to the growth of the economy. The reason why this is quick and dirty is because it is just that... quick and dirty. So dirty in fact, that it is wrong. Over the past 60 years, earnings (and thus dividend) growth of S&P 500 companies hasn't kept up with that of the broader economy (AND has recently taken a massive dive). Want an even quicker and dirtier method? Assume 6% (see here).

Historical Data
Below is a chart showing just that, the growth of the economy and that of both earnings and dividends.

Click for Giant Chart...

In theory, since the growth rate of dividends and earnings have been lower than that of nominal GDP, the S&P 500 should have underperformed the nominal growth rate of equities ex dividends. However, as the chart below shows, just the opposite has occurred as investors have flocked to the asset class (think 401k's and baby boomers) driving up the price to earnings ratio (i.e. the price individuals are willing to pay for each $1 of earnings) since 1949.

Equity Forecast
Forecasts should never be thought of as an exact science, but rather a framework based upon assumptions that can / will change frequently. That said, using the quick and dirty method as a starting point, I'd go with 2.5% long-term GDP growth (in real terms) with 2% inflation to get nominal economic growth at 4.5% (low, yes). Add to that the 2.5% current dividend yield and we get a 7% long-term return for equities as an asset class. BUT, based on the data I presented above, I'd reduce this by roughly 1% as dividends have underperformed nominal GDP to get a 6.0-6.5% outlook over the long-term. Less fees associated with transacting (hat tip MAB) and I'll reduce this to 5.5%-6.0% for the broader investment community.

Of course there are scenarios that could play out over both the near-term (continued P/E compression from strong flows into the asset class) and long-term (economic growth reverting back to the long-term mean faster than I suspect) that would mean higher returns, but there are also plenty of surprises that may happen to the downside (one specifically being the possibility of the P / E ratio trending lower in coming years).


Source: Irrational Exuberance / BEA


  1. Jake,

    Interesting charts.

    I have long been skeptical of linear extrapolation. Just because something has corellated well in the past does not necessarily mean it will in the future. And there can be large gaps where there is negative corellation for periods of more than a decade or longer.

    The problem comes with having to make assumptions. We need to assume that:

    1) There will be net positive economic growth over the period of time in question
    2) There will be net positive earnings growth...
    3) Interest rates will be high enough to raise dividends (try buying a dividend paying company in Japan)
    4) Inflation is accurately reflected by the CPI (we're talking asset prices, so why must we use consumer prices to deflate growth stats?)
    5) Risk appetites do not fluctuate based on factors not considered
    6) Regulatory/Structural changes are not too radical to have large effects

    And there are many others. But making these assumptions is what gets pension funds in trouble when they make projections of 7.5% compound annual returns. Ditto for Social Security, hedge funds, quant funds, etc.

    There is nothing that can be done to eliminate risk and nothing that can be done to guarantee returns over any period of time. We are kidding only ourselves to think otherwise.

  2. All great points Matt. Thanks for the comment.

    I think the bigger point isn't to back into the assumed 6-6.5% rate, but to determine how comfortable you are with all the assumptions that feed into that rate.

    For example, I am less inclined to believe that all of those assumptions we both raised will ring true AND based upon the fact the earnings growth has not kept and multiples are high, I believe the 6-6.5% may be a stretch over the long-run.

    All that said it is important to have a basic understanding of what to expect with regards to returns to compare values across assets classes. To me a corporate bond yielding 4.5-5% or a Treasury yielding 3.5-4% has more value in the current environment than equity.

  3. Jake - excellent thinking and framework particularly with the notion that a framework is a weigh of thinking thru a problem and then adjusting to realities as they unfold. And Matt's points are well taken as parameter tweeks for working thru the framework. My own approach is slightly driven and focuses more on real data but our conclusions are similar. This chart on profits, GDP the SP500 might be interesting:
    Which ought to be coupled with this chart on LT PE ratios:
    BtW - have you looked into Graham-Dodd's formula for valuing equities?
    If your e-mail still works and you're interested I'd be happy to forward the recent newsletter that kicks this around, among other things.

  4. please do:


  5. Jake,

    I think the TIPS markets have it right - low real growth (under 2%) and low inflation (under 2%) for at least a decade. What keeps me up at night is the amount of debt in the system. I could easily envision a big debt deflation problem. Heck, it took tens of trillions in guarantees to prevent one already!

    Assuming the Fed and Gov't can stem the debt default tide, my best guess is 2% real growth and 2% core inflation. Oil could very well bounce the CPI well above and below the 2% core rate, but I still think it's a reasonable long run inflation rate given debt levels.

    As for long run stock returns, I think your 6-6.5% annual return estimate is a bit high. I have it pegged at 2% real + 2% inflation + 2.1% dividend - 1% fees & dilution (~ 5%). No way can investors expect 100% of the real growth and inflation. Note that the forward dividend rate on the S&P 500 is just under $22/share (less than your dividend rate).

    Most of our recent "growth" was fabricated with CONsumption debt and Government spending. That can't last for ever. And the payback period portends low growth and perhaps much worse imo.

    I still like my bonds (primarly treasuries) but I'm not a big risk taker anymore. Too much debt in the economy and leverage in the financial markets for my money.

  6. missed fees (how, i'm not sure).

    you and i are generally on the same page. long the long bond again as well.

    my fear is related to kicking the can down the road. i feel like these low rates / monemental stimulus are doing just that with all that debt overhang. what happens when private demand truly needs to take over?

  7. Jake - on the way. Price is feedback of course.

    One more chart to team up with the other two and riff on everybody's points - a Graham-Dodd master table:

  8. In regards to where you think p/e should be, this article might be of value to you

  9. thanks crispycrunch-

    here is a post i wrote back in jan re: the P/E ratio based on comments from dblwyo


  10. the graham-dodd formula assume ROE of 12% based on a historical context (and no growth). has anyone read any material challenging this assumption?

  11. Jake,

    Every year S&P issues a pension obligation report for S&P 500 companies. At the end of 2008, S&P 500 companies had ~ $500 billion in short falls pension and other post employment benefits (OPEB, primarily medical).

    It's a fun yearly read as the delusional projections of pension funds make for great heckling.


    From the most recent report:

    Estimated pension return rates
    declined 7 bps to 7.95% from 8.02% in 2007, marking the eighth consecutive annual reduction.

    I guess these money (mis)managers are all going to be way above average going forward, despite their past dismal performances. Well, at least they are politically correct - we're all above average nowadays!

    I'll eat a bug if they get 7.95% long term annual returns from today's bond and stock valuations. Greenspan and Bernanke never should have allowed banks to issue so much financial debt. The leverage means to much of our future returns are already spoken for. It's a double whammy too. When the returns don't materialize, the majority have to bailout the banks and leveraged speculators. The majority also get smaller pensions than they would have. What a sham(e).

  12. Jake - could you walk thru the arguments/math on that G-D ROE argument? Afraid it escapes me.

    Did the newsletter make it btw?