Understanding ROE (Part 8)

Today concludes the mini-series on return on equity (ROE) that has also discussed ROA, ROIC, and some case studies.

Probably the most important thing I have learned from writing this is how to deal with unconventional cases like BKNG, SBUX, and DPZ. For starters, look at shareholder equity (total assets – total liabilities)—also known as book value. If negative, then expect quirky ROE numbers like negative numbers or “too-good-to-be-true” triple digits because ROE becomes a [broken] ghost metric. As I have discussed, these are cases where ROA and/or ROIC might be better to study.

With such heavy mention of “shareholder equity” throughout this mini-series, I think it’s worthwhile to spend some time discussing that more common synonym often mentioned in financial writings.

For many different reasons, “book value” is often in quotation marks:

• To distinguish it as a specific accounting term rather than a definitive measure of a company’s true current worth.
• As accounting convention based on historical records that may not align with current market prices [capitalization].
• Because balance sheet reflects costs minus depreciation, which may be much lower than current fair market value.
• To indicate the term is being used in its technical sense as shareholder equity rather than a general estimation of value.
• To signal skepticism or irony regarding the accuracy of the figure (i.e. “scare quotes”).
• To suggest the [cooked] “book value” is inflated or inaccurate if a company’s financial statements are under scrutiny.

Shareholder equity is actual accounting book value, but the quotation marks are a warning label that the official number may hide the economic reality. It’s often just a starting point that may need heavy adjusting because it fails to reflect what a company is truly worth due to standard accounting rules. For example, massive value drivers like brand reputation, proprietary software, and trained workforces are almost always ignored in accounting book value. In practice, book value is often viewed as the bare minimum value [rather than its value as a growing business] if the company were to be shut down immediately.

As a result, analysts often move from the “actual” accounting number to an adjusted book value by revaluing [marking to market] tangible assets (e.g. land) to their current sale price, by cleaning up liabilities by adding off-balance sheet items or potential legal costs that haven’t yet been recognized, and by calculating book value/share subtracting preferred stock to reflect what common shareholders would receive.

Understanding ROE (Part 7)

In Part 6, I discussed the misleading return on equity (ROE) for SBUX. Today I want to look at Domino’s Pizza (DPZ).

According to Google AI, ROE for DPZ is as follows according to six different sources:

• 15.4% TTM as of Apr 2026 per FinanceCharts.
• -15.4% TTM based on -$3.9B shareholder equity per Wisesheets.
• 6.2% per Danelfin.
• Distorted by low equity with return on assets (ROA) at 35.1% per ChartMill.
• -10.7% in ’26 per MarketScreener.
• Distorted by negative [shareholder] equity and therefore N/A per PEvaluation.

While negative ROE usually means a company is losing money, DPZ has negative shareholder equity rather than negative profit. The company has aggressively used debt to fund massive share buybacks and dividends. The “asset-light” model results in liabilities exceeding assets with a shareholder equity of -$3.9B. Net income (’25) is +$602M, though. Simply Wall Street notes that DPZ maintains a healthy interest coverage ratio of 4.9x to 5.3x suggesting it easily pays the interest on its debt.

Because DPZ ROE is distorted by negative shareholder equity like SBUX, we may look to ROA or return on invested capital (ROIC). ROA > 35% outperforms nearly all industry peers.

ROA seems “too good to be true” (see Part 4) and may also be distorted. ROA is high primarily because of an asset-light franchise model, not just raw profitability. Franchising means the massive real estate and equipment costs of 20,000+ locations are not on the balance sheet (total assets only $1.7B). Despite owning few stores, DPZ collects high-margin royalties and supply chain fees creating a relatively large net income divided by a very small asset base.

Articles at Business Model Mastery and Restaurant Business compare DPZ with McDonald’s (MCD): two companies with heavy debt but different asset structures leading to vastly different ROA. MCD ROA is 14.4% due to ownership of massive real estate. Owning land and buildings for most locations results in a $60B total assets leading some to describe MCD as a REIT disguised as a burger chain. Both have negative ROE (-478% for MCD) due to debt and share buybacks.

Where negative shareholder equity breaks ROE and an ultra-low asset base inflates ROA, DPZ ROIC remains stable. ROA uses net income in its numerator: after interest expenses are paid. DPZ has massive debt and huge interest payments that make the business look less efficient. ROIC ignores interest by using net operating profit after tax. This isolates the pizza business performance from the “noise” of how it is funded. ROA also falls prey to the “lazy cash” problem because total assets include things that don’t actually help make pizza (e.g. excess cash in a bank account, old unused equipment) that inflate the denominator and lower the score. ROIC ignores non-operating items thereby giving a tighter view of management’s skill.

Where ROA is better for asset-heavy firms like banks and factories, ROIC ~50-60% shows DPZ earning far more than its weighted average cost of capital (~9%) and therefore a significant value generator despite an unusual balance sheet.

Understanding ROE (Part 6)

In Part 5, I discussed the misleading return on equity (ROE) for BKNG. Today I want to look at Starbucks (SBUX).

Recall with BKNG that ROE can be confusing because:

• Standard financial formulas crash when shareholder equity [the denominator; see (1) in Part 1] is negative. Some platforms will show “N/A,” while others might show a massive positive or negative percentage based on automated math calculation failing to account for the negative equity context.
• Negative equity suggests a company is near bankruptcy. BKNG, however, is highly profitable with ~$8B in annual free cash flow. The negative equity is a management decision rather than a sign of failure.
• BKNG has spent over $52B on share buybacks. Buybacks reduce shareholder equity artificially inflating (or breaking) ROE calculation making it impossible to use for comparison with companies employing different capital structures.

Google AI reports SBUX 2026 ROE of -17%. ROE is +16% for ’25—a sharp decline from the 12-month average of +32%. Before shifting to negative shareholder equity (’19), its high positive ROE peaks > 386% in ’18 due to high leverage.

To illuminate and clarify, FinanceCharts.com (cited by Google AI) literally says:

      > The return on equity (roe) for Starbucks (SBUX) stock is 16.32% as of
      > Wednesday, April 22 2026. It’s worsened by -49.51% from its 12-month
      > average of 32.33%. The 5 year average return on equity (roe) is 43.04%.

Worsening by a negative value is the same as improving by the positive. 16.32% worsening by negative 49.51% means it has increased from a 12-month average of 16.32% – 49.51% = -33.19%: inconsistent with the above-stated +32.33%. I consider this a prime example of how broken some of these math algorithms can be when shareholder equity is negative.

Like BKNG, persistent negative equity is a result of deliberate capital allocation strategies. First, SBUX historically returns significant capital to shareholders via share buybacks often funded using debt. Second, SBUX maintains a streak of over 60 consecutive quarters of dividend payouts that further reduce shareholder equity. Finally, long-term debt has climbed to over $15B as the company invests in store expansions and its “Back to Starbucks” turnaround strategy.

Given the distortion in ROE, a stock analyst is advised to use other efficiency metrics. ROA (see Parts 3 and 4) is currently +4.3%. This measures profit against the total asset base rather than just shareholder equity. Return on Invested Capital (ROIC; see Part 2) is currently +6.8%. This is another more reliable metric for SBUX that accounts for both debt and equity.

For more details on financial health, Google AI suggests tracking SBUX ROE trends on Macrotrends or viewing quarterly growth performance on FinanceCharts.com. You can also read the latest valuation reports on Yahoo Finance or explore long-term return trends on Stock Analysis on Net (https://www.stock-analysis-on.net/).

I hope to conclude next time with one more case study.

Understanding ROE (Part 5)

I want to finish the mini-series by presenting some cases where return on equity (ROE) seems nonsensical and may be misinterpreted.

The first case is Booking Holdings (BKNG). Per Google AI, TTM ROE is 97% as of Apr 2026.

In my opinion, 97% ROE falls under the “too good to be true” category, but this is not unusual for BKNG. Pre-Pandemic (2019), ROE is a more “normal” but still strong 82%. In ’22, ROE is 110%, a sharp increase from previous years driven by recovery in travel demand and increased financial leverage. ROE fluctuates wildly from ’23–’25, often appearing highly positive (e.g. 156% in ’23).

ROE is a complex and potentially misleading metric for BKNG because it operates with negative shareholder equity. The impressive percentage turned up by some traditional screens is a mathematical quirk of having a negative denominator [recall (1) here]. BKNG reports shareholder equity of -$5.6B in ’25 largely driven by aggressive share buybacks and debt-financed capital allocation. The negative shareholder equity is not due to operational losses, though. In fact, BKNG remains highly profitable with $5.4B TTM net income.

Despite negative shareholder equity, some sites will report positive ROE due to a programming shortfall. It should be reported as a negative figure or “N/A.” Instead, sites like FinanceCharts.com use absolute values or specific internal logic to show a performance percentage when there really isn’t one. The positive percentage is a ghost metric.

Platforms such as Wisesheets.io report the same data as negative ROE or “N/A” following standard application of the formula with shareholder equity of -$5.6B (i.e. a shareholder deficit) at the end of 2025 divided into positive net income.

The reality is a highly profitable BKNG [$5.4B net income for ’25, as mentioned above] that has wiped out its equity book value through massive share buybacks rather than business losses. Negative equity is a deliberate management decision rather than a sign of operational failure.

For stock analysis when shareholder equity is negative, it can help to:

• Use Return on Invested Capital (ROIC) to see how well the company uses all its capital (equity and debt). BKNG recently reported an ROIC of 43.7%, which remains a valid and strongly positive number.
• Focus on cash flow metrics like Free Cash Flow (FCF) and FCF Yield to study how the company generates cash regardless of its accounting book value. BKNG grew FCF 15.1% in 2025.
• Evaluate solvency by interest coverage rather than Debt-to-Equity. 13.6 interest coverage indicates well-managed debt.

I will continue next time.

Understanding ROE (Part 4)

I left off Part 3 discussing some reasons it may be useful to study return on assets (ROA) as an alternative to return on equity (ROE). Today I will conclude by comparing the two metrics.

With regard to the math, ROA is:

                                Net Income * 100    
                              --------------------  
                              Average Total Assets             

Average Total Assets is the arithmetic mean between beginning and end of the year (or period). Note that ROE uses only shareholder equity in the denominator of its calculation.

ROA can mislead for companies with few physical assets (e.g. software or services) where high ROA may not fully reflect the company’s operational strength if driven by accounting artifacts or strategic underinvestment rather than true efficiency.

Accounting artifacts may include:

• Fully-depreciated [older] assets having a lower book value on the balance sheet thereby shrinking the denominator and artificially inflating ROA despite equipment being potentially inefficient or nearing its end.
• Ignoring off-balance-sheet assets like brand value, patents, or intellectual property that would otherwise lower ROA.
• A temporary spike in net income from non-recurring items (e.g. selling a subsidiary or a large asset write-down) that may make a company appear more efficient in a single period than it actually is.

Strategic underinvestment could be high ROA achieved by delaying necessary capital expenditures (e.g. not upgrading machinery or software) thereby creating a questionable short-term boost while hurting long-term competitiveness. Similarly, aggressive inventory reduction can temporarily lower total assets and raise ROA at the risk of facing stockouts and lost future sales—both signaling hidden operational fragility.

According to some, another ROA flaw is division of net income (truly belongs only to equity holders) by total assets (funded by both debt and equity). A possible fix would be to use ROIC (see Part 2) or to add back interest expense in the numerator.

While 15–20% ROE is often considered strong, a “good” ROA is typically 5–10% with < 1% indicating financial weakness.

ROA is preferred by operational managers, lenders and regulators, and acquisition analysts because it evaluates how effectively internal resources are deployed, because it assesses fundamental operational soundness and asset quality, and because it provides an apples-to-apples comparison across companies with different levels of debt, respectively.

For equity investors, ROE is preferred because it focuses directly on the return generated for shareholders despite being easily inflated by increased debt (including share buybacks).

At the risk of sounding redundant, ROA provides a grounded view of how well a company uses its entire resource base–—both debt and equity—–to generate profit.

ROA is a good metric to determine whether managers are making good use of the company’s infrastructure and technology; consistent, high ROA indicates sustainable competitive advantage.

In summary, ROE describes how much the shareholders are making whereas ROA describes how well the business is running.