Total Shareholder Return Decoder

I used to read a lower P/E as a sign that a stock had become cheaper. If I could buy the same business for fewer times its earnings, surely I was getting a better deal.

McDonald’s made me question that instinct. There were years when earnings fell, P/E increased, and the share price barely reflected the deterioration. When earnings recovered, P/E fell again—even while shareholders made money.

The interesting part was the way those movements offset each other. Was the market changing its view of the business every year? Or was it looking beyond a volatile earnings number toward something more stable?

That is where total shareholder return becomes useful.

What the shareholder received—and what produced it

Total shareholder return, or TSR, combines the change in the share price with dividends received. Buy a share for $100, receive $3 in dividends, and finish with a share worth $108: your return is 11%.

The outcome alone does not explain the investment. A return can come from a better business, from a higher valuation, or from changes in financing. Those sources tell different stories about what might be repeated.

The TSR Decoder separates the story into its parts. Revenue growth tells us whether the business became larger. Reinvestment shows the cash absorbed in getting there: growth that consumes most of the additional earnings is different from growth that leaves substantial cash for owners. Margin change asks whether each dollar of sales earned more after tax.

Earnings yield connects operating earnings with the price paid for the business. It is not an extra dividend. It helps explain the earnings capacity behind the return, before considering the cash that must remain invested.

Multiple re-rating captures the effect of paying a different valuation for operating earnings. Financing and other effects complete the bridge from the business to the shareholder, including interest, debt, cash and share-count changes.

For 2025, McDonald’s generated a 7.9% TSR. Roughly 3.9 percentage points came from revenue growth and margin improvement after reinvestment; about 4.0 points came from earnings yield. Multiple re-rating was small. The return was largely supported by the earnings base and business performance, rather than a large increase in the valuation investors assigned to it.

One year explains an outcome. The historical pattern raises the more interesting question.

Why earnings and P/E moved in opposite directions

In 2020, earnings per year-end share fell by about 21.4%, while the share price rose 8.6%. P/E increased to approximately 33.8 times.

In 2021, earnings per year-end share recovered by about 59.6%. The share price rose 24.9%, and P/E fell to approximately 26.5 times.

The same pattern appeared again: earnings weakened in 2022 and recovered sharply in 2023, while the price moved much less and the multiple moved against the earnings fluctuation.

The arithmetic is straightforward: Price = Earnings per share × P/E. When earnings move much more than the price, the multiple necessarily moves in the opposite direction. That counter-movement is partly mechanical.

The economic question is why the price moves less.

One explanation is that investors expect some of the earnings change to reverse. A temporary setback can reduce current earnings without reducing the business’s long-term earning capacity by the same amount. Conversely, an exceptional recovery does not mean every part of the increase will repeat.

That interpretation is consistent with McDonald’s price looking through part of the earnings fluctuation. It is not proof that the market was slow to react. Prices may have anticipated the recovery before it appeared in the accounts. Changes in risk, interest rates and growth prospects can also affect the multiple.

This changed how I read the decomposition: a negative multiple contribution does not automatically mean investors became pessimistic. In 2023, the share price rose even though P/E fell. Earnings simply rose faster. Calling that movement “worsening expectations” would hide the most useful part of the story.

The first matrix: read the movement

The price and P/E matrix makes four situations visible.

Price up and P/E down means earnings rose faster than the price. Price down and P/E up means earnings fell faster than the price. Both can be consistent with the market cushioning an earnings fluctuation.

When price and P/E rise together, the price has strengthened relative to current earnings. When both fall, the price has weakened relative to earnings. We still need to investigate growth, risk and the sustainability of profits before deciding whether either movement creates value for a buyer.

This matrix helps explain what changed. It does not establish that the stock is cheap or expensive.

The second matrix: was the market right?

The investment question comes next: is the earnings change actually temporary, or is the business facing a lasting change?

If the setback is temporary and the market looks through it, a high reported P/E may simply reflect depressed current earnings.

If the setback is temporary but the price discounts lasting damage, there may be an opportunity. The work is to establish that the recovery is credible.

If the damage is structural but investors keep waiting for a recovery, the apparently stable price can be misleading. The expected normal earnings may never return.

If both the business assessment and the market point to lasting damage, the repricing may be appropriate. A lower price or multiple is not sufficient evidence of a bargain.

This is where the analysis becomes an investment judgment: what earning capacity will remain after the fluctuation, and what price am I paying for it?

What happened after McDonald’s P/E rose?

We also compared the following year’s TSR after years when McDonald’s P/E increased with the following year’s TSR after years when it decreased.

After the increases in 2019, 2020 and 2022, the next-year returns averaged 17.8%. After the decreases in 2018, 2021, 2023 and 2024, they averaged 5.6%.

The result belongs in the discussion because it runs against the instinct that a falling P/E must make the next purchase more attractive. In this small McDonald’s sample, the better following-year outcomes came after P/E had risen.

An earnings recovery offers a possible explanation for part of that result. Depressed earnings made the multiple look higher; subsequent improvement supported the shareholder’s return. Once earnings had recovered, the lower multiple could reflect a stronger denominator rather than a suddenly cheaper share.

The result does not tell us to buy every high-P/E stock. There are only three observations in one group and four in the other, with the pandemic strongly affecting the sample. It does tell us to ask why the multiple is high or low before treating it as a valuation conclusion.

For me, that is the useful lesson from McDonald’s: P/E can fall because the earnings number improved, without the price becoming cheaper. P/E can rise because current earnings weakened, without long-term earning capacity deteriorating by the same amount.

The TSR Decoder makes those relationships visible. The next step is to judge the earnings that will endure—and the expectations already included in the price.

Try the TSR Decoder: https://eoninvestment.com/tools/tsr-calculator/

Sources and calculation notes

The FY2025 decomposition is from EON Investor Tools v0.3.8 and uses reported after-tax operating profit and EV/NOPAT. Historical P/E is an equity measure and is kept separate from that operating-business multiple.

Historical net income, issued shares and treasury shares: SEC Company Facts and McDonald’s annual filings. Historical P/E = year-end share price × year-end shares outstanding ÷ full-year net income. These historical per-share figures differ from reported diluted EPS, which uses weighted-average shares. The historical study is not a point-in-time backtest.

Returns: Yahoo Finance daily closes and cash dividends, selected by ex-dividend date, before tax and without dividend reinvestment. Observation years are 2018–2024; following-year returns end in 2025. The data cutoff is 31 December 2025. The corrected arithmetic averages are 17.8% and 5.6%; the earlier discussion quoted 17.6% and 5.4%.

The cash-based reinvestment measure in the decoder includes working capital, acquisitions and other investing flows; it is not a pure estimate of growth capex. The earnings-yield contribution uses ending-period NOPAT divided by starting enterprise value. The interpretation matrices are analytical frameworks, not proof of causation.

SEC Company Facts: https://data.sec.gov/api/xbrl/companyfacts/CIK0000063908.json

McDonald’s annual reports: https://corporate.mcdonalds.com/corpmcd/investors/financial-information.html

Market data: https://finance.yahoo.com/quote/MCD/history/

Methodological reference: Koller, Goedhart and Wessels, Valuation, 7th edition, chapter 5; https://www.mckinsey.com/capabilities/strategy-and-corporate-finance/our-insights/a-better-way-to-understand-trs