Why Korean Ratios Go Stale — And How Trailing Twelve Months Fixes It

If you look up a Korean company's P/E in August and the figure was computed from the December annual report, you are looking at a number that is eight months old. In a market where semiconductor earnings can triple or vanish inside a year, that is not a small problem.

1. The calculation

Trailing twelve months (TTM) means the most recent four quarters, whenever you happen to be looking. Korean filings report cumulative year-to-date figures rather than discrete quarters, so in practice it is computed as:

TTM = last full year + this year cumulative − last year same-period cumulative

Subtracting last year's first half from last year's full year leaves last year's second half; adding this year's first half gives the most recent twelve months. Every earnings-based figure on this site — EPS, P/E, ROE, net margin — is built this way.

2. Balance-sheet items are not summed

Flow items accumulate over a period; stock items are a snapshot. Equity, assets and liabilities are taken at the latest quarter-end, not added across four quarters — summing them would quadruple the company's equity.

The consequence is that a single page legitimately mixes two bases: book value per share and P/B are quarter-end figures, while EPS, P/E and ROE are trailing-twelve-month figures. That is correct, and we label which basis each figure uses rather than hiding it.

3. What TTM still does not fix

Seasonality is handled — all four quarters are included, which is why annualising a single quarter (Q1 × 4) is far more dangerous than TTM.

Cyclicality is not. If the last twelve months were the peak of a memory or chemical cycle, TTM earnings are peak earnings and the P/E computed on them looks deceptively cheap. This is a judgement problem, not a calculation problem: for cyclical businesses, compare against a multi-year average of earnings as well.

One-off items are not. A property sale or a legal settlement inside the window flows straight into TTM. Looking at the quarterly progression usually makes the outlier obvious.

4. Why data providers disagree about the same company

The most common reason two sources show different P/E ratios for the same Korean company is not an error — it is a different basis:

• One uses the last annual filing; another uses TTM; another uses forward estimates.
• One uses consolidated net income; another uses the controlling-interest portion.
• One uses total shares; another nets out treasury shares.
• Prices are struck on different dates.

Before concluding a number is wrong, align the basis. We publish ours in full on the methodology page, and round per-share figures so that dividing the displayed price by the displayed EPS reproduces the displayed P/E — you can check the arithmetic yourself.

Summary

• TTM = last annual + current cumulative − prior-year same-period cumulative.
• Flow items use TTM; balance items use the latest quarter-end.
• TTM handles seasonality but not cycles or one-offs.
• Most cross-source disagreements are basis differences, not errors.