Applying Growth Analysis to Capital Market Expectations

The CFA Level III CME Module 1.04 reading addresses one of the most practical questions in long-horizon investing: how do you translate a trend growth forecast into expected returns for stocks and bonds? This article walks through the four reasons trend growth matters for capital market expectations, the equity-value identity that decomposes long-run returns, the Bjornsdottir worked example, and the famous puzzle of why high-growth countries do not reliably deliver high equity returns.

Symbols used in this article

Symbol	Meaning
$g_Y$	Trend real GDP growth
$\pi$	Trend inflation
$V_e$	Aggregate market value of equity
$\text{NGDP}$	Nominal GDP
$S_k$	Capital share of income (earnings / GDP)
$\text{P/E}$	Price-to-earnings ratio
$D/P$	Dividend yield (annual dividends / market value)
$D/E$	Dividend payout ratio (dividends / earnings)
$r_e$	Expected nominal equity return
$y$	Default-free nominal bond yield

Why trend growth matters for CME — the four reasons

Loading diagram...

1. DCF discipline. Earnings and cash-flow forecasts in any DCF model must, in the aggregate and in the long run, grow at the trend rate. A model that projects 10% trend earnings growth in a 2.5% trend-growth economy is internally inconsistent — corporate profits cannot consume an ever-growing share of GDP indefinitely.

2. Already-priced growth. A country with a higher trend rate of growth may offer good equity returns IF that growth is not already priced in. The market may have already discounted the trend, in which case no excess return is available.

3. Inflation headroom. A higher trend growth rate allows actual output growth to be faster before accelerating inflation becomes a concern. This shapes the path of monetary policy and bond yields.

4. Real bond yield anchor. Theory and empirical evidence link the average level of real default-free bond yields to the trend rate of real growth. Faster trend growth implies higher average real yields.

Anchoring bond yields to trend growth

Trend real GDP growth pulls real bond yields toward it over long horizons:

$$\text{Real long bond yield} \approx g_Y \quad \text{(long-run mean reversion)}$$

The nominal yield decomposes as:

$$y = g_Y + \pi + \text{term premium}$$

For a US economy with $g_Y \approx 2.4\%$ trend real growth, $\pi \approx 2.1\%$ trend inflation, and a modest term premium, the long-run nominal 10-year Treasury yield should anchor around $4.5\% - 5\%$.

Loading diagram...

This anchoring matters for intertemporal consistency — even your short-horizon forecasts must be consistent with the eventual reversion to this anchor.

The V_e identity: the central equity decomposition

The CFA framework decomposes aggregate equity market value into three multiplicative factors:

$$V_e = \text{NGDP} \times S_k \times \text{P/E}$$

Where:

• $\text{NGDP}$ = nominal GDP
• $S_k$ = capital share of income = $\frac{\text{Earnings}}{\text{GDP}}$
• $\text{P/E}$ = price-to-earnings ratio = $\frac{V_e}{\text{Earnings}}$

Take logs and differentiate to get a growth-rate decomposition:

$$g_{V_e} = g_{\text{NGDP}} + g_{S_k} + g_{P/E}$$

Loading diagram...

The key insight: in the long run, both $S_k$ and P/E must be approximately constant. Capital cannot continually take a rising share of national income (it would eventually approach 100%), and the P/E multiple cannot rise without bound. Therefore:

$$g_{V_e}^{\text{long run}} \approx g_{\text{NGDP}} = g_Y + \pi$$

Long-run equity CAPITAL APPRECIATION grows at the rate of nominal GDP. To get TOTAL return, add the dividend yield:

$$r_e^{\text{long run}} \approx g_Y + \pi + g_{S_k} + g_{P/E} + \frac{D}{P}$$

Over finite horizons, $g_{S_k}$ and $g_{P/E}$ can be nonzero — if profit margins are expected to widen or multiples to compress, these terms matter.

The dividend yield identity

A useful complementary identity for the dividend yield:

$$\frac{D}{P} = \frac{D/E}{P/E}$$

Where $D/E$ is the dividend payout ratio. Given any two of these three ratios, you can infer the third. CFA-style problems often give you a target P/E and a payout ratio and ask you to compute the implied dividend yield.

The Bjornsdottir example — decomposing post-WWII US returns

In January 2021, Alena Bjornsdottir (the curriculum-named CFA analyst) examined US equity returns from 1946–2020 to test whether they were consistent with economic growth. The continuously compounded annual return was 10.7%, decomposed as:

Component	Value	Contribution
Real GDP growth $g_Y$	2.9%	2.9%
Inflation $\pi$	3.5%	3.5%
Change in EPS/GDP ($g_{S_k}$)	0.0%	0.0%
Change in P/E ($g_{P/E}$)	0.9%	0.9%
Dividend yield $D/P$	3.4%	3.4%
Total		10.7%

Question 1: What conclusion does Alena draw?

The post-war stock return EXCEEDED what would have been consistent with economic growth. The rising P/E added 0.9% of "extra" return per year for 74 years:

$$\text{Cumulative P/E contribution} = 74 \times 0.9\% = 67\%$$

This left the market roughly:

$$e^{0.67} - 1 = 1.95 - 1 = 95\% \text{ above "fair value"}$$

Question 2: Baseline projection for long-run continuously compounded US equity returns

Alena assumes:

• Labor input growth: 0.9%
• Labor productivity: 1.5%
• Inflation: 2.1%
• Dividend yield: 2.25%
• No further P/E growth

Real GDP growth $g_Y = \text{labor growth} + \text{productivity growth} = 0.9\% + 1.5\% = 2.4\%$.

$$r_e^{\text{baseline}} = g_Y + \pi + g_{S_k} + g_{P/E} + \frac{D}{P} = 0.9\% + 1.5\% + 2.1\% + 0\% + 2.25\% = 6.75\%$$

Notice: this projection is roughly 400 basis points BELOW the 10.7% historical average. The historical return was inflated by (a) higher historical inflation than projected, (b) P/E expansion that cannot continue, and (c) higher historical dividend yields than the current 2.25%.

Question 3: P/E reversion adjustment

If the 95% over-valuation is expected to mean-revert over 30 years, the baseline projection should be reduced by:

$$\text{P/E drag} = \frac{67\%}{30 \text{ years}} = 2.2\% \text{ per year}$$ $$r_e^{\text{adjusted}} = 6.75\% - 2.2\% = 4.55\%$$

Loading diagram...

Alena adjusted projection of ~4.55% is dramatically below the 10.7% historical average — by more than half. This illustrates how dangerous it can be to extrapolate historical returns without testing whether they were consistent with sustainable growth fundamentals.

The "growth ≠ returns" puzzle

The curriculum cites a striking empirical finding: countries with higher economic growth rates do NOT reliably generate higher equity market returns. Two explanations:

Loading diagram...

1. Already priced in. The market may have already discounted the higher growth rate. Investors buy in at high valuations, and the return on capital is normal even if the underlying economy grows fast.

2. ROIC compression. If the capital stock grows rapidly (capital deepening), the return on invested capital falls. Recall that fast capital growth runs into diminishing returns — same insight from the TFP framework.

3. Capital share compression. Faster growth often comes from policies that favor labor (e.g., minimum wages, social spending). The $S_k$ term in the $V_e$ identity falls, reducing equity returns.

The lesson: high growth need not translate one-for-one into higher equity returns unless it can be expected to continue forever AND has not yet been priced in AND maintains current profit shares.

The CFA-exam pattern

Module 1.04 questions typically ask you to:

1. Decompose historical equity returns into the five Bjornsdottir components
2. Project baseline returns from forecasts of $g_Y$, $\pi$, $D/P$, and changes in $S_k$ and P/E
3. Adjust for expected mean reversion of P/E or profit margins
4. Anchor bond yields to trend real growth + trend inflation
5. Reconcile high-growth countries with their actual equity returns (already-priced argument)

For the foundational growth-accounting framework that produces $g_Y$, see our TFP article. Practice CME applications in our CFA Level III question bank.