INVESTMENT GUIDE
Metric guide
What PER, ROE, debt ratio and other metrics mean, how each is calculated and what range is common. The financial metrics Jupick uses to grade stocks, explained one at a time from the ground up.
Where to start reading
If you already know what you want to find out, starting from this list grouped by situation is quicker.
New to Jupick
Start with what the grade summarizes and what it does not say.
When a grade looks surprising
Cases where one metric is strong but the grade is low, where a whole metric drops out because of the sector, and the score range a single grade letter covers.
When a metric moved but the business did not
Years when one-off gains or losses mixed into net income, when share buybacks or cancellations shrank the share count and equity, and when an unusually low or high prior year inflated or deflated growth.
When a number differs from one seen elsewhere
Why values for the same company differ from source to source, gathered in one place.
Why one company can show several different PERsWhat free data Jupick is built onCompanies with two sets of financial statementsToday's price next to last year's financialsHow old is the data behind a grade badgeWhat a price change is measured againstDividend dates and earnings dates share the same trapWhen the share count changes, past numbers are rewritten
When you meet a blank or a hyphen
What an empty number means, whether the blank is a value that will be filled in later, and how blanks are reflected in the grade.
Not all blanks are the sameWhen a metric goes blank in only one marketWhy banks have no current ratioHow a 30-year dividend streak is countedThe ex-dividend date is there but the payment date is blankIs this earnings date confirmed or estimatedHow many measured values sit under a gradeScoring without the metrics that could not be measured
When looking into dividend numbers
Dividend yield, payout ratio and consecutive dividend years, read apart as the different things they measure.
Three ways to read a dividend stockHave high-dividend companies given up on growthCompanies that paid out more than they earnedWhat going ex-dividend meansHow a 30-year dividend streak is countedThe ex-dividend date is there but the payment date is blankDividend yieldPayout ratioConsecutive dividend years
When comparing price and company size
Where the current price sits relative to earnings and assets, where it sits in the past year's range, how company size is measured and whether size divides grades.
When reading overseas stocks too
Once exchange rates and tickers come in, the same number can mean something different.
When the whole market is the question
What a single index does not show, and terms that come up often on the coin side.
When coin grades are the question
What each of the three axes behind coin grades (size, liquidity and volatility) measures, and what the kimchi premium and stablecoins mean.
When exchange rates are the question
What it means when the KRW/USD rate goes up, what moves exchange rates, and how to read exchange rate statistics and the exchange rates inside financial statements.
Metric glossary
20 metrics · by grade weightProfitability 40%
How well a company turns capital and sales into profit. The largest share of the overall grade, at 40%.
ROEReturn on equity. How much profit the company made in a year with the money shareholders put in (equity).Operating marginThe share of sales left as operating income from the core business, showing how profitable that business is.Revenue growthHow much sales grew compared with a year earlier, the pace of the company's top-line growth.EPSEarnings per share. The company's net income for a year divided by the number of shares, the amount one share earned.ROAReturn on assets. How much profit the company made in a year with all its assets, including those funded by debt.Net marginNet income after all costs and taxes, as a percentage of sales.ROICReturn on invested capital. How much after-tax operating income the company made with the capital actually invested in the business.
Stability 35%
Whether a company can carry the weight of its debt. 35% of the overall grade.
Debt ratioHow large liabilities are relative to equity, a leading measure of financial stability.Current ratioHow large the assets that can be turned into cash within a year are, relative to the debts due within a year.Interest coverageHow many times a year's operating income could pay the interest expense.Reserve ratioHow much profit and capital surplus the company has kept instead of paying out, compared with paid-in capital.
Dividends 25%
Whether returns to shareholders are steady. 25% of the overall grade.
Valuation Not graded
Whether the current price is high or low relative to earnings and assets. Read separately from the grade.
PERPrice to earnings ratio. How many times the company's net income the current price trades at.PBRPrice to book ratio. How many times the company's net assets (book value) the price trades at.BPSBook value per share. The company's net assets (equity) divided by the number of shares, the book value held in one share.PSRPrice to sales ratio. Market cap divided by sales, how much the market puts on each unit of sales.EV/EBITDAEnterprise value (EV) divided by earnings before interest, taxes, depreciation and amortization (EBITDA), a valuation metric that includes debt.
Size Not graded
The size of the whole company as priced by the market.
Read further
59 articles · 7 groups59 articlesReading grades10 articlesWhat a one-letter badge summarizes and what it covers up, down to how far scores spread within the same grade.From financial statements to metrics9 articlesWhich line items of the financial statements each metric comes from, and why a metric with the same name takes different values in different sources.Reading dividends6 articlesDividend yield, payout ratio and consecutive dividend years measure different things. The ex-dividend date and the payment date have their own stories too.When numbers mislead10 articlesYears when only the metric moved while the business stayed the same, and cases where a low-looking value hides the reason it is low.Blanks and time lags: the limits of the data10 articlesWhat Jupick cannot measure, when a value was measured and where it came from. The story behind the screen.Seeing the whole market4 articlesWhat a single index does not show, what it means to say asset classes moved together, and how to read news tone and macro indicators.Coins and exchange rates10 articlesAssets beyond stocks. The three axes behind coin grades, and the ways exchange rates enter company numbers.
Articles by group
For a wider picture of grades and financial statements than a single metric gives. The 59 articles are split into 7 groups by topic, and within each group the articles to read first come first. Articles that are a step in the learning path show their place in it.
Reading grades 10 articles
What a one-letter badge summarizes and what it covers up, down to how far scores spread within the same grade.
Reading Jupick grades: what S, A, B, C and D say and do not sayJupick's overall grade weights profitability 40%, financial stability 35% and dividends 25%, and turns the 0 to 100 score into S to D at cutoffs of 85, 70, 50 and 30. It summarizes disclosed financials and says nothing about future prices.Step 1 of 40 in the learning pathUsing Jupick grades well: does a low grade mean a bad stock?A Jupick grade summarizes financial metrics and is not a verdict. A low grade does not mean a bad stock, a high grade does not mean a stock to acquire now, and the grade works best as a starting point for comparison and exploration.Step 2 of 40 in the learning pathHow wide are Jupick's grade bands: 166 major stocks (29.4%) sat at exactly 50, the floor of grade BGrade B spans 50 to 69 points, a 20-point band. Measuring where major stocks sat inside their bands on September 16, 2026 found 166 of the 358 B-grade stocks at exactly 50.0, and 159 of them had none of the nine metrics measured.Step 4 of 40 in the learning pathThree faces under one badge: counting the area splits the overall grade hides in major stocksThe overall grade is a weighted average of three area grades. Among major stocks on September 22, 2026, only 6 percent had all three areas on the same grade, and 29 percent hid a gap of three grades or more, such as A in profitability and D in dividends.Step 21 of 40 in the learning pathDo larger companies get better grades: splitting major Korean stocks by size, the top 100 and bottom 100 differed by 2.5 pointsRatings research abroad keeps finding that larger companies score higher. Splitting major Korean stocks by market cap on September 20, 2026, the top and bottom 100 differed by only 2.5 points, and grade A or above was more common among the smaller ones.How many measured values sit under a grade: in a count of major stocks, 150 (37.5%) had all nine metrics filledJupick's overall grade scores nine metrics, yet stocks with all nine measured and with none carried the same badge. A September 2, 2026 count of major stocks shows how common each is and why the ones with none landed at C 46.How old is the data behind a grade badge: counting how fresh each of the three areas is in major stocksJupick grades Korean stocks on annual reports and overseas stocks on quarter-end data. Counting the watchlist on September 22, 2026, all three areas rested on the latest data in only 14 percent of Korean stocks and 0 percent of overseas ones.Is a PBR of 1.5 high or low: on three yardsticks, major stocks got split answers in 1,178 of 1,711 casesThe same number lands in a different place depending on the group it is set against. Jupick placed every metric of its major stocks on its three yardsticks at once and counted how often the three answers differ.Jupick has two grades: a stock S and a coin S do not mean the same thingJupick grades both stocks and coins S to D with the same letters and cutoffs, but the two share no inputs. Stock grades read nine financial statement metrics, coin grades read size, liquidity and stability, and commodities and exchange rates get no grade.Scoring without the metrics that could not be measured: 66 grades changed, so the rule was not adoptedA proposal to drop missing metrics instead of filling them with neutral values was tested on Jupick's major stocks on September 1, 2026. It changed 66 grades (16.5%), and lowering one threshold sent the same stocks the opposite way.
From financial statements to metrics 9 articles
Which line items of the financial statements each metric comes from, and why a metric with the same name takes different values in different sources.
Financial statement basics: the balance sheet, income statement and cash flow statement at a glanceFinancial statements are made of three tables: the balance sheet, the income statement and the cash flow statement. Each answers a different question, and a beginner can read them in a simple order.Step 5 of 40 in the learning pathCompanies with two sets of financial statements: consolidated and separateMany listed companies report two sets of financial statements for the same year, consolidated and separate. Sales, profit and the debt ratio differ between them, and which one is read shapes the view of the company.Why one company can show several different PERsDifferent sites often show different PERs for the same company on the same day. Usually neither is wrong: they divide different numbers. The split comes from which profit, which period and which share count is used, and Jupick states its definition.Reading PER and PBR together: priced low, or low for a reasonPER measures the share price against earnings and PBR against net assets. Read together they reveal more than either alone, and a stock that looks priced low often has a reason for it.Choosing a valuation metric: different companies need different yardsticksPER, PBR, PSR and EV/EBITDA each make sense for a different kind of company: steady earners, asset-heavy businesses, growth companies still posting losses, and companies carrying heavy debt.Choosing a profitability metric: ROE, ROA, ROIC and marginsROE, ROA and ROIC measure profit against different kinds of capital, while operating margin and net margin measure it against sales. Read side by side, the gaps between them show where profit comes from.High ROE with a high debt ratio: why the same number has two facesSome stocks rank at the top on profitability and near the bottom on stability. The metrics do not contradict each other; debt is already built into ROE. Splitting ROE into three parts shows the structure.What market cap actually countsMarket cap is share price times share count, and the choice of share count gives one company several market caps. Treasury shares, preferred shares, free float and dual listings split it, and Jupick's Korean and overseas figures use different definitions.Is a PER of 12 high or low: reading a metric by lining up the marketA single metric value cannot show whether it is common or rare. The average fails as a baseline, the median and quartiles show position, and filled-in values have to be left out when counting the distribution.
Reading dividends 6 articles
Dividend yield, payout ratio and consecutive dividend years measure different things. The ex-dividend date and the payment date have their own stories too.
Three ways to read a dividend stock: dividend yield, payout ratio and consecutive dividend yearsA dividend is a share of a company's profit paid out to shareholders. Dividend yield, payout ratio and consecutive dividend years each answer a different question about it, and a high yield reads correctly only next to the other two.How a 30-year dividend streak is counted: the definitions behind one numberConsecutive dividend years look like a simple count, but the answer depends on where counting starts and what counts as a dividend. Korean and overseas stocks use different records, special dividends are a gray zone, and an uncountable case is not a zero.What going ex-dividend means: why the share price drops on the day the dividend right falls awayOn the ex-dividend date the share price is adjusted by about the amount of the dividend. The drop reflects cash about to leave the company, the date differs from the record date, and the drop is not a loss for the shareholder.Companies that paid out more than they earned: what a payout ratio above 100% meansThe payout ratio measures how much of a year's earnings went to dividends, so it rarely tops 100%. On September 22, 2026, 15 of the 348 stocks with a measured payout ratio on Jupick's watchlist were above 100%.Have high-dividend companies given up on growth: dividends and growth across major stocksComparing dividend yield with revenue growth on Jupick's September 22, 2026 watchlist shows a weak tradeoff in direction, but 82% of the high-dividend stocks were still growing and the correlation was close to 0.Step 20 of 40 in the learning pathThe ex-dividend date is there but the payment date is blank: the same gap in all 147 Korean stocksIn the dividend schedule on Jupick's stock pages, Korean stocks show only the ex-dividend date. A count on September 1, 2026 found payment dates for 0 Korean stocks and 171 overseas stocks, because the two dates are produced in different ways.
When numbers mislead 10 articles
Years when only the metric moved while the business stayed the same, and cases where a low-looking value hides the reason it is low.
Is a low-PER stock a good company: PER and grades compared across major stocksOn September 22, 2026, 69% of the lowest-PER quarter of the major stocks Jupick covers had the middle grade B, and top grades were more common at the high-PER end. Financial strength and price level did not move together.Step 19 of 40 in the learning pathEarnings mixed with one-off items: why PER and ROE swing together in a single yearOne-off items such as a gain on a property disposal or a large provision flow into net income, so PER, ROE and EPS move together in that year even when the business is unchanged. Reading them means telling a business change from a one-year event.Step 24 of 40 in the learning pathShare buybacks and cancellation: why EPS, ROE and PBR move while earnings stay flatWhen a company repurchases or cancels its own shares, the share count and equity shrink, so EPS, ROE and PBR move even with flat earnings. Repurchase and cancellation differ, and the shift is separate from any change in the business.Step 25 of 40 in the learning pathWhat is the base effect: when a weak last year makes this year's growth look largeGrowth rate is built from this year and last year, so an unusually weak last year makes this year look better than it is, and a strong last year makes it look weak. Reading revenue growth means checking the denominator too.Step 26 of 40 in the learning pathWhen the share count changes, past numbers are rewrittenA stock split or rights offering rewrites a company's past prices and dividends on today's share count. That is why dividend history tables do not show what was actually paid each year, and why most Jupick metrics stay unaffected.Is high ROE an illusion created by debt: ROE and debt ratio compared across major stocksROE can rise just by adding debt, which raises the worry that a high ROE is a leverage illusion. On Jupick's watchlist on September 22, 2026, ROE and the debt ratio were nearly unrelated, and high-ROE stocks carried lighter debt.What is the holding company discount: why a holding company trades below its subsidiaries combinedA holding company's market cap is often below the combined market value of its subsidiary stakes, a gap called the holding company discount. It has several causes, and PER, PBR and the debt ratio read differently for these companies.Step 27 of 40 in the learning pathWhy banks have no current ratio: when a different sector makes a metric disappear entirelyA blank metric field is not always missing data. In some sectors the number does not exist at all, as with a bank's current ratio, and filling such blanks with zero produces a distorted score.Today's price next to last year's financials: reading the time lag in financial dataA stock price and the financial metrics beside it on one screen are not from the same moment. The lag builds up in four places, published figures can later be revised, and Jupick records the fiscal reference date and the update time separately.Step 8 of 40 in the learning pathHow far below the 52-week high: what the 52-week position says and does not sayMeasured on one day, the median 52-week position was 37.7 percent for 198 Korean stocks and 65.4 percent for major overseas stocks. Real distributions show why a drop from the high says nothing about valuation and how band width differs from position.
Blanks and time lags: the limits of the data 10 articles
What Jupick cannot measure, when a value was measured and where it came from. The story behind the screen.
Not all blanks are the same: five things a hyphen can mean in financial metricsA hyphen on a financial screen can mark a value that was not disclosed, one that failed to arrive, one cut off by the data window, one frozen in place, or one that never applies. Measured counts show which of these Jupick can tell apart.Step 22 of 40 in the learning pathWhen a metric goes blank in only one market: the hole an average hidesOn August 30, 2026, interest coverage was blank for 196 (98%) of Jupick's major Korean stocks, yet the two markets combined were about half filled and set off no alarm. Measured counts show how the sector and collection-rule explanations fell short.What a price change is measured against: different baselines for stocks, crypto and overseas marketsThe same +1.2% is a change from the previous close for Korean stocks, from 24 hours before now for crypto, and from the session that ended overnight for US stocks. These different baselines explain most gaps with figures seen elsewhere.Step 23 of 40 in the learning pathIs this earnings date confirmed or estimated: why some dates are exact and others show only the monthThe next earnings date on a stock page or the calendar can be a date the company has confirmed in a disclosure, or an estimate inferred from past quarterly cycles. Jupick marks the two differently and shows only the month for estimates on the calendar.Dividend dates and earnings dates share the same trap: keeping dates from slipping by a dayDividend schedules and earnings dates on Jupick stock pages share one data source and one pipeline. Dates stamped at UTC midnight that could slip a day when read in Korea, occasional corrupted dates and the different validity windows are handled together.When an index showed an old close as today's value for five days: why Jupick added a second data sourceIn September 2026, Korean index data quietly froze and the screen showed an old close as today's value for five days. The numbers looked valid, so no error check caught it. Jupick now takes each index from two sources and uses the more recent one.A ticker does not identify a companyA stock ticker is not a company's name but a slot an exchange lends out for a while. Jupick queried public registries directly to see where its data stopped when a large bank's ticker moved, and how many different things one ticker points to today.What a six-digit Korean stock code can and cannot tell youThe only thing a Korean stock code reveals with certainty is its last character. A check of all 2,767 listed stocks shows how common and preferred shares split, what codes with letters mean and what happened while Jupick had the rule wrong.How Jupick selects its major stocks: more rules for exclusion than for inclusionJupick redraws its Korean stock list every day from two inclusion rules and a longer list of exclusions. Preferred shares, SPACs and KONEX stocks are left out, updates stop if the list runs short, and being listed is no endorsement.What free data Jupick is built on: a map of data sources and their limitsJupick uses no paid data: Korean prices come from Korea Exchange (KRX), Korean financials from DART, overseas stocks from Yahoo Finance, crypto from CoinGecko, macro indicators from FRED and news from RSS headlines. Each source has its limits.Step 28 of 40 in the learning path
Seeing the whole market 4 articles
What a single index does not show, what it means to say asset classes moved together, and how to read news tone and macro indicators.
When the index rises but fewer stocks go up than downOn the day the Nasdaq rose 1.57%, only 55 (27.5%) of the major overseas stocks Jupick tracked went up. The index weights money by market cap while advance-decline counts give each stock one vote, so the two can point opposite ways.Step 34 of 40 in the learning pathOne number for moving together: how to read the asset class correlation tableJupick's asset class correlation table sums up in one number from -1 to +1 how closely assets such as the KOSPI, Nasdaq, gold and crude oil moved together over the last 90 days. It shows how often they moved the same way, not size, cause or future.Step 32 of 40 in the learning pathHow headline sentiment labels work: the tone of words and what it leaves outThe positive, negative and neutral badges on Jupick news come from counting Korean words in each headline. Phrases are checked before single words and ties are neutral. The label is word tone, never a price direction, and a word list has clear limits.Keeping current inflation from being a three-year-old figure: release cycles and staleness of macro dataMacro indicators such as interest rates, inflation and unemployment are released on different cycles, and some series stop updating. Jupick labels each with its reference date, drops values that are too old, and converts index levels into year-over-year rates.
Coins and exchange rates 10 articles
Assets beyond stocks. The three axes behind coin grades, and the ways exchange rates enter company numbers.
Crypto market cap: circulating supply times price, and the size behind the coin gradeA coin's market cap is its circulating supply times its current price. Circulating, total and max supply differ, and Jupick's coin grade gives market cap the largest weight and scores it on a log scale.Step 35 of 40 in the learning pathCoin liquidity and turnover: why trading volume is divided by market capLiquidity is how readily a coin trades near the intended price. Jupick measures it with turnover, daily trading volume divided by market cap, and high or low turnover describes the current market, not future prices.Step 36 of 40 in the learning pathWhy crypto prices swing so much: what volatility and the stability score sayVolatility is the size of a price's swings regardless of direction. Jupick's coin stability score is higher the smaller the average absolute 7-day and 30-day price change, and a high score does not mean a coin is safe.Step 37 of 40 in the learning pathKimchi premium: why coin prices in Korea run higher than overseasThe kimchi premium is the percentage gap by which a coin's price on Korean exchanges runs above (or below) its overseas price converted to KRW. A positive value is a premium, a negative one a reverse premium, and both describe the current gap, not the future.Stablecoins and Tether (USDT): how a coin pegged to $1 worksA stablecoin is designed to hold a steady price, usually pegged at 1 coin to $1. Tether (USDT) is the crypto market's main unit of account, the USDT premium and dominance describe the market's current state, and the peg is a design goal, not an assurance.What an exchange rate is: a rising KRW/USD rate means a weaker wonAn exchange rate is the ratio at which two currencies trade. A KRW/USD rate of 1,300 means one dollar takes KRW 1,300, so when the rate rises the won has weakened, which pushes up import prices and travel costs.Step 38 of 40 in the learning pathWhat moves exchange rates: interest rates, trade and risk appetiteExchange rates move with demand for and supply of each currency, shaped mainly by interest rate gaps, the trade balance, risk appetite and the dollar's strength. These explain why rates moved; they do not forecast future rates.Step 39 of 40 in the learning pathReading exchange rate statistics: period changes and the 52-week positionOn Jupick's exchange rate screen, a plus change means the currency strengthened against the won and a minus means it weakened. The 52-week position shows where the rate sits in the past year's range, a record of the past and not a forecast.Step 40 of 40 in the learning pathWhen exchange rates move, the company stays the same but its numbers changeConverted into won, an overseas company's market cap shifts daily even when nothing happens at the company, while PER and the debt ratio stay the same in any currency. The dividing line is whether a number is a ratio of two values in one currency.Where exchange rates enter the financial statements twiceExchange rates enter the statements of a company that earns abroad more than once. Revenue and assets are converted at different rates, and the gap goes to equity rather than profit, which is why net income can swing while operating income stays calm.
What this guide is
Brokerage apps show numbers such as PER, ROE and debt ratio, but working out what each number means and what range is common is still left to the user. This guide takes the metrics Jupick uses to grade stocks one at a time and covers the meaning, the formula, how to read the ranges and the common misreadings in a free glossary.
The ranges for each metric match the yardsticks Jupick uses when it marks strengths and weaknesses, so the guide reads straight on into How grades are made.
Try it yourself
If reading is not enough, Grade lab lets you move sliders for nine metrics and watch the values become 0 to 100 points, pass through the three areas and fold into an S to D grade in real time. Once that feels familiar, the grade quiz tests your reading by guessing grades from financials alone. To see where a metric comes from in the financial statements, Statement lab breaks line items down into metrics. All of these are tools for understanding the calculation, not real stocks.
This content is for information only and is not investment advice. Nothing here suggests trading any stock or says anything about future prices. The aim is to build the habit of reading metrics for yourself. Investment decisions and responsibility for them rest with the user.