Part 5 · Comparing companies with financial ratios · Lesson 27
Look at 3–5 year trends, not a single year
One year of numbers is a photo; five years of numbers is a video. Only by stacking several years can you tell a company that keeps improving from one that shone for a single year.
QToday's question
One company doubled its operating profit last year. Another grew a little every year for 5 years. Whose numbers can you trust more?
A big number in a single year stands out, but you can't tell from that year alone whether it's skill or luck. In this lesson we build up, step by step, how to stack several years side by side, how to reduce growth speed to a single number, and how to pull several years of data out of a real annual report.
MAPLearning map for this lesson
We build it up in three steps. At the end of each step, a 'Quick check' lets you test your understanding before moving on.
In Lesson 12 we put year 2 and year 3 profits side by side, and in Lesson 13 we saw a real company's 3-year gross margin swing widely. In Lesson 22 we stacked three years of net income and operating cash flow to judge earnings quality. If the last lesson, Lesson 26, compared companies sideways, this time we stack one company vertically. Also recall the price cycle from Economics Lesson 3.
- STEP 1 · Basic conceptsOne photo versus a 5-year video
Explain why looking at only one year's numbers makes them easy to misread, and what you can see by looking at the .
- STEP 2 · Understanding the principleCAGR: measuring growth speed as one number
Calculate the and use a 5-year dashboard to tell steady companies from erratic ones.
- STEP 3 · Real-world applicationReading several years from summary financial information
You can pull 3 years of data from the annual report's summary financial information, calculate the flow of margins and ROE, and read cyclical industries.
STEP 1One photo versus a 5-year video
No matter how good one year's numbers are, you should hold off on judging if you don't know what came before and after. First, let's see why.
So far we have studied mainly Robin Bakery's year-3 numbers: sales of KRW 12 billion, operating profit of KRW 1.5 billion, and ROE of 20%. But from year 3 alone you cannot tell whether these numbers are the result of steady accumulation or whether this year was just unusually good. Stacking several years of numbers side by side to see direction and volatility is called .
An easy analogy · One report card
A student scored 95 on this exam. If the last five were all in the 90s, that's real ability; but if they were in the 60s and only this one was 95, the exam may have been easy or the student lucky. Company results likewise reveal real ability only when you overlay several report cards.
Looking at just one year creates three main illusions.
1 Base effect. If last year was unusually bad, even an ordinary year this year shows a large growth rate. It is the same principle as the inflation base effect in Economics Lesson 8.
2 One-off items. Non-recurring numbers, like the land-sale gain we saw in Lesson 15, inflate a single year.
3 Position in the cycle. In industries whose prices rise and fall, like semiconductors, a year at the cycle's peak and a year at its trough look like completely different companies (Lesson 26).
So look at at least 3 years, and 5 if possible. Stack not only revenue but also operating profit, ROE and operating cash flow. If revenue grows but margins keep falling, it may be 'expensive growth,' and if profit grows but operating cash flow doesn't follow, the warning light from Lesson 22 is on.
Quick checkA company's operating profit this year is +200% from last year. Last year's operating profit was −70% from the year before. How should you read this year's number?
STEP 2CAGR: measuring growth speed as one number
When the growth rate differs each year, you want to boil it down to a single figure for 'how many percent it grew per year on average.' That number is the CAGR.
Let's extend Robin Bakery's sales to five years. Adding two hypothetical years before year 3, sales are KRW 8.5 billion → KRW 9.2 billion → KRW 10 billion → KRW 11 billion → KRW 12 billion. Annual growth rates differ slightly from year to year: about 8.2%, 8.7%, 10.0%, and 9.1%. Boiling these down into one number gives the .
Be careful not to confuse this with a simple average. If revenue goes 100 → 150 → 100, the yearly changes are +50% and −33%, so the simple average is +8.5%, yet in reality nothing has changed. CAGR correctly tells you it is exactly flat at 0%. It is the same principle as compound interest in Economics. You can also calculate it with the site's Return and CAGR Calculator, or check it right below.
Compound annual growth rate
The beginning and ending values only need to be in the same unit (KRW 100 million, KRW trillion or KRW all work). Robin Bakery's numbers are fictional.
But CAGR has a weakness too. It looks at only two points, the start and the end, so it doesn't show how much the figures swung in between. That's why we also look at the year-by-year numbers as a chart. In the dashboard below, compare Robin Bakery with a fictional company, Pado Electronics. Pado Electronics is set up as a components company that rides the price cycle heavily.
Both companies are fictional. Robin Bakery's years 1–3 are the course's shared numbers, and years −1 and 0 are fictional numbers added in this lesson. ROE is based on ending equity. 'Looking at one year only?' keeps only the best-looking single year for each company.
Wave Electronics' sales CAGR is only about 3.0% (KRW 8 billion → KRW 9 billion), but looking at year 2 alone, with operating profit of KRW 3 billion and ROE of 30%, it looks much better than Robin Bakery. If you had first looked at this company that year, its PER would also have come out low. Earnings at the top of a cycle are called , and valuing a company on these earnings makes disappointment easy the very next year.
Common misconceptions
"If last year's growth rate was high, it's a growth stock?" No. A single year can be high because of the base effect or one-off items. Look at the 3–5 year CAGR and the year-by-year flow together.
"If CAGR is high, it's a steady company?" No. If the first year is a trough and the last year is a peak, CAGR is inflated. Check which years you chose as the start and end.
"You should avoid cyclical industries?" No. Just know that results swing, and think in terms of average earnings across the whole cycle. The problem is not knowing about the swings and trusting a single year's numbers.
Quick checkRevenue went from KRW 20 billion to KRW 25 billion 3 years later. Roughly what is the CAGR?
STEP 3Reading several years from summary financial information
A real company's numbers for several years are gathered in one place in the annual report. Let's pull them out and calculate directly.
The first item of 'III. Financial Matters' in the annual report you saw in Lesson 5 is . Key numbers from the balance sheet and income statement are collected in one table for the latest three years, which makes it the most convenient starting point for looking at trends. Below are numbers taken from the summary consolidated financial information in Samsung Electronics' March 11, 2025 annual report (56th fiscal year), whose screens you saw in Lessons 6 and 12. This is an example to explain company structure and is not an investment judgment.
| Item (consolidated) | 2022 (54th term) | 2023 (55th term) | 2024 (56th term) |
|---|---|---|---|
| Revenue | KRW 302.2 trillion | KRW 258.9 trillion | KRW 300.9 trillion |
| Operating profit | KRW 43.4 trillion | KRW 6.6 trillion | KRW 32.7 trillion |
| Operating margin (calculated) | 14.4% | 2.5% | 10.9% |
| Net income attributable to owners of the parent | KRW 54.7 trillion | KRW 14.5 trillion | KRW 33.6 trillion |
| Equity attributable to owners of the parent | KRW 345.2 trillion | KRW 353.2 trillion | KRW 391.7 trillion |
| ROE (calculated, ending equity basis) | 15.9% | 4.1% | 8.6% |
Source: DART, Samsung Electronics annual report (filed 2025.03.11, rcpNo 20250311001085), III-1 Summary financial information, 'A. Summary consolidated financial information'; units of KRW million rounded to KRW trillion. Operating margin and ROE calculated by Robin Invest (verified 2026-10-08).
Laying the numbers out vertically reveals a story. Revenue is nearly flat over 3 years at KRW 302 trillion → 259 trillion → 301 trillion, yet the operating margin fell from 14.4% to 2.5% and came back to 10.9%. It is the same pattern as the swings in gross margin we saw in Lesson 13. If you had judged it "a company with a 2.5% margin" from 2023 alone, or conversely "a company with 15.9% ROE" from 2022 alone, you would have seen only one scene of the cycle. The memory price cycle we saw in Economics Lesson 3 is stamped right into the actual financial statements.
Summary financial information has 3 years of data, so if you want to look at 5 years, just open one more summary from the annual report of two years earlier. Also get into the habit of checking whether the numbers for the year where the two tables overlap (2022 in this case) are the same. Past numbers are sometimes rewritten because of accounting standard changes or error corrections.
Try it yourself: stitching 5 years together from two reports
Let's do it with Robin Bakery (fictional). The summary financial information in the year-1 annual report contains years −1, 0 and 1, and the year-3 annual report contains years 1, 2 and 3. Placing the two tables side by side, year 1 overlaps.
| Revenue (KRW 100 million) | Year −1 | Year 0 | Year 1 | Year 2 | Year 3 |
|---|---|---|---|---|---|
| Year 1 report | 85 | 92 | 101 | — | — |
| Year 3 report | — | — | 100 | 110 | 120 |
| Stitched 5 years | 85 | 92 | 100 | 110 | 120 |
This is a fictional example. Year 1 revenue changing from KRW 10.1 billion to KRW 10 billion assumes a case where the accounting treatment was changed to 'deduct sales incentives from revenue' and past figures were rewritten (restated).
Year 1 revenue, which overlaps, differs between the two reports. In such cases, use the numbers from the later report as the standard. The latest report restates the past under the changed accounting standard, so only by connecting those numbers do you compare on the same yardstick. The reason for a restatement is usually written in a note below the table or in the 'Changes in accounting policies' note (Lesson 29).
Recalculating CAGR with the stitched 5 years gives about 9.0% over 4 years, from KRW 8.5 billion to KRW 12 billion. The growth rate over the 2 years from year 1 to year 3 is about 9.5% on a KRW 10 billion base, but linking from the old year-1 figure of KRW 10.1 billion gives a lower figure of about 9.0%. The difference may look small, but for companies with large restatements (mergers and acquisitions, business disposals, accounting standard changes), the trend can look completely different. When numbers don't match, the first step in reading a trend is to find out what happened at that spot.
Finally, when connecting five years of numbers, connect like with like. Mixing consolidated for one year and separate for another makes the company's size itself look different (Lesson 11), and mixing net income attributable to owners of the parent with total net income makes ROE jumpy (Lesson 16). The habit of first checking the 'consolidated/separate' label and the unit at the top of each table prevents most number mistakes.
Find it yourself on DART
1 As you learned in Lesson 3, open the company's latest annual report. 2 In the left-hand table of contents, click 'III. Financial Matters' → '1. Summary financial information'. 3 Of 'A. Summary consolidated financial information' and 'B. Summary separate financial information', look at the consolidated one first (Lesson 11). 4 Check the unit at the top (KRW million, etc.) and copy down revenue, operating profit, net income attributable to owners of the parent, and equity. 5 Open the same table in the annual report from two years earlier and stitch on the earlier two years, and you have 5 years.
How to use it in investing
| If you see this pattern | Ask first | Where to look together |
|---|---|---|
| Revenue and profit grew a little each year for 5 years | Did operating cash flow grow too, and are margins holding? | Lesson 22 Earnings quality |
| Profit jumped only this year | Is it the base effect, a one-off, or the top of the cycle? | Lesson 15 One-offs, Lesson 17 Quarters |
| Revenue grows but margin falls every year | Can't it absorb price competition or rising costs? | Lesson 13 Costs, Lesson 14 Operating leverage |
| A cyclical industry's PER is very low | Are earnings at a peak, and what multiple is it on several years' average earnings? | Lesson 16 PER, Lesson 26 |
Not always. There is no guarantee that past trends will continue, and for companies whose business structure has changed (acquisitions, spin-offs, new businesses), numbers from several years ago can't be compared as the same company. A trend is not a tool for predicting the future but a tool for gauging how special today's numbers are.
Quick checkIn the table above, revenue in 2022 and 2024 is similar, but the operating margins differ at 14.4% and 10.9%. What can't you tell from revenue alone?
✓What we learned today
- One year of numbers is a single photo. Only by stacking revenue, operating profit, ROE and operating cash flow over 3–5 years can you tell skill from luck.
- CAGR = (ending value ÷ beginning value)^(1 ÷ number of years) − 1. Robin Bakery's revenue going from KRW 8.5 billion to KRW 12 billion (4 years) is about 9.0% a year on average.
- Even with the same CAGR, the path can differ. Look at the year-by-year numbers together to check whether growth was steady or swung widely.
- The annual report's III-1 Summary financial information holds 3 years in one table. The real example company's operating margin swung widely, 14.4% → 2.5% → 10.9%.
- In cyclical industries, PER looks low in the year when earnings are at the top and high in the year when they are at the bottom. Think of the average over the whole cycle rather than a single year's number.
In Part 5 you learned how to compare numbers sideways and vertically. In Part 6 we go beyond the numbers. In the first lesson, Lesson 28, you will find for yourself how the company makes money in the annual report's 'Business overview'.
Quiz5-question quiz
Pick an answer and it's graded right away with an explanation.
Q1. Revenue went from KRW 10 billion to KRW 12.1 billion 2 years later. What is the compound annual growth rate (CAGR)?
121 ÷ 100 = 1.21, and the square root of 1.21 is 1.1. So CAGR is 10%. 21 ÷ 2 = 10.5% is a simple average that ignores compounding.
Q2. If revenue moves 100 → 150 → 100, what is the CAGR?
The start and end are the same, so CAGR is 0%. The simple average of the yearly growth rates (+50%, −33%), +8.5%, hides the fact that nothing actually changed.
Q3. Where in the annual report can you see the last 3 years of revenue, operating profit and equity in one table?
III-1 Summary financial information lays out the key consolidated and separate numbers for 3 years. If you need 5 years, open one more annual report from two years earlier.
Q4. A cyclical company's PER came out at 5x in the year its earnings peaked. What is the most appropriate interpretation?
Dividing by peak earnings gives a low PER. When earnings return to an average level, PER becomes much higher at the same stock price. Recalculate using several years' average earnings.
Q5. When looking at a company whose operating profit this year is +150% from last year, what should you check first?
A big one-year growth rate can come from the base effect, one-offs or the top of the cycle. Stack 3–5 years and check whether operating cash flow grew too.
NextTerms · Next lesson · Related tools
A method of stacking numbers such as revenue, operating profit, ROE and operating cash flow side by side over 3–5 years to see direction and volatility. It reduces the illusions that come from looking at a single year because of the base effect, one-offs and cycles.
Compound annual growth rate (CAGR)Compound annual growth rateA value that reduces growth over several years to 'what percent would it be if it grew at the same rate every year.' It is calculated as (ending value ÷ beginning value)^(1 ÷ number of years) − 1 and does not show swings in between.
Peak earningsPeak earningsEarnings in the year when profit is at the top in an industry that rides a cycle. Calculating PER on these earnings tends to make it look low, so look at it together with several years' average earnings.
Summary financial informationSummary financial informationA table at the very front of Chapter III of the annual report that shows assets, liabilities, equity, revenue and profit for the last several fiscal years at a glance.
ROE (Return on Equity)Return on EquityNet income divided by net assets (equity), showing how much shareholders' money earns in a year. The relationship PBR = PER × ROE links the two multiples.
Operating marginOperating marginOperating profit divided by revenue. It shows how many won are left from the core business per 100 won of revenue. Normal ranges differ greatly by industry.
Cash flow from operating activitiesCash flow from operating activitiesThe cash that actually came in and went out from the core business, such as selling products and paying for materials, wages and taxes. Whether it is consistently positive is the basic check on a company's strength.
Cyclical stockCyclical stockAn industry or stock whose earnings swing widely with the business cycle. Semiconductors, chemicals, steel, autos and shipbuilding are typical examples.
Price cyclePrice cycleA pattern in which supply reacts slowly to changes in demand, so prices rise too far and then fall too far, repeatedly. It is especially pronounced in memory semiconductors, chemicals and shipping.
Base effectBase effectA phenomenon in which the number at the comparison base date is unusually high or low, so the current rate of change looks lower or higher than the actual trend. It often appears when reading year-over-year indicators.
Put what you learned to use right away
CAGR Calculator→Financial Statements Lesson 22 · Checking earnings quality→Economics Lesson 3 · Prices are set by supply and demand→This is educational material and not investment solicitation. There is no guarantee that past relationships will repeat in the future.