Every November, the same message makes the rounds in NEPSE trading groups. A company’s first-quarter report is out, the “annualized EPS” has jumped, the PE ratio has dropped, and someone declares the stock undervalued. Three months later the second-quarter report lands, the annualized EPS slips, and the same stock is suddenly “overpriced.” Nothing about the business changed. Only the arithmetic did.
EPS on NEPSE is the most quoted earnings number in the market, and the most often misread. EPS, or earnings per share, is net profit divided by the number of shares. That part is easy. The problem is the word usually attached to it in Nepal: annualized. Most EPS figures on portals, in quarterly reports and in tip sheets are not what the company earned in a year. They are a projection, built by stretching a few months of profit into twelve.
Here is this article’s position. An annualized quarterly EPS is not a measure of earnings. It is a guess with a formula attached, and for the sectors where seasonality is strongest, hydropower above all, it is a guess that is wrong in a predictable direction. If you are going to use EPS at all, you need to know which number you are looking at, how it was built, and when to throw it out.
What EPS on NEPSE actually measures
Start with the clean version. Basic EPS is profit attributable to ordinary shareholders, divided by the weighted average number of ordinary shares outstanding during the period. In Nepal, listed companies follow Nepal Financial Reporting Standards, and the specific rule for this calculation is NAS 33, the Nepal Accounting Standard on earnings per share issued by the Accounting Standards Board Nepal. Its stated purpose is to make EPS comparable between companies and across periods for the same company.
Two words in that definition matter more than they look.
“Attributable” means profit that belongs to shareholders, after tax and after anything owed to other claimants. For most NEPSE companies this is close to the bottom line of the income statement.
“Weighted average” means the share count is adjusted for when shares were issued during the year. And when a company issues bonus shares, NAS 33 treats the new shares as if they had always existed, so prior-period EPS figures get restated downward to match. Keep that in mind. It becomes important later.
An EPS figure on its own tells you very little. NPR 30 of EPS is excellent for a stock trading at NPR 300 and poor for one trading at NPR 1,500. Its real job is as the denominator in the price-to-earnings ratio, which we cover in depth in our guide to how to use the PE ratio on NEPSE. If the EPS is wrong, every valuation built on top of it is wrong too.
How “annualized EPS” is built in Nepal
Nepal’s fiscal year runs from Shrawan to Ashadh, roughly mid-July to mid-July in the AD calendar. Listed companies report in four quarters: the first ends with Ashwin (mid-October), the second with Poush (mid-January), the third with Chaitra (mid-April) and the fourth with Ashadh (mid-July).
Under the Securities Registration and Issue Regulations, listed companies have to prepare and submit their quarterly reports within a month of each quarter’s close, as ShareSansar has reported in its coverage of SEBON’s disclosure requirements. Those reports are unaudited. The audited annual accounts come months later, often well after the fourth-quarter report has already moved prices.
Here is the key detail most new investors miss. Nepali quarterly reports are cumulative. A second-quarter report shows profit for the first six months of the fiscal year, not for the second quarter alone. A third-quarter report shows nine months.
To turn that partial-year profit into a full-year EPS, the convention is to scale it up:
- First quarter: three months of profit, multiplied by 4
- Second quarter: six months of profit, multiplied by 2
- Third quarter: nine months of profit, multiplied by 4/3
- Fourth quarter: twelve months, no adjustment
ShareSansar has described the logic plainly: the quarter’s EPS is treated as the minimum expected return for each remaining quarter, so a first-quarter EPS of NPR 5 becomes an annualized EPS of NPR 20.
The hidden assumption is enormous. Annualizing assumes the months you have not seen yet will look exactly like the months you have. For a business whose earnings arrive evenly through the year, that is a reasonable shortcut. For a business whose earnings depend on rainfall, harvests, festival spending or year-end accounting adjustments, it is fiction.
The hydropower trap, in real numbers
Hydropower is where annualization does the most damage, because river flows, and therefore generation, swing hard between the monsoon and the dry season. The first quarter of Nepal’s fiscal year, Shrawan to Ashwin, sits squarely in the monsoon. The third quarter, Poush to Chaitra, sits in the driest, lowest-flow stretch of the year. We explain the physics and the tariff structure behind this in our piece on hydropower seasonality and earnings.
Take Chilime Hydropower Company (CHCL), one of the most widely held hydropower stocks on NEPSE, and walk through its fiscal year 2082/83 reports as published on ShareSansar.
- First quarter (published November 16, 2025): cumulative net profit of NPR 205.84 million.
- Second quarter (published February 12, 2026): cumulative net profit of NPR 398.05 million.
- Third quarter (published May 3, 2026): cumulative net profit of NPR 456.99 million.
- Fourth quarter (published August 11, 2026): full-year net profit of NPR 720.06 million, with EPS of NPR 7.59 and paid-up capital of about NPR 9.48 arba, according to ShareSansar’s summary of the company’s report.
Now apply the standard annualization to each report. The figures below are Biznessvue calculations, using the year-end share count of roughly 9.48 crore shares (NPR 9.48 arba of paid-up capital at NPR 100 per share) so that the bonus issue during the year does not muddy the comparison.
After the first quarter, annualized profit comes to about NPR 823 million, an implied EPS of roughly NPR 8.68. After the second quarter, about NPR 796 million, or NPR 8.39. After the third quarter, about NPR 609 million, or NPR 6.42. The actual full-year figure turned out to be NPR 720 million, or NPR 7.59.
Read that sequence again. The first-quarter projection overstated the final result by about 14 percent. The third-quarter projection understated it by about 15 percent. The “earnings” of the same company, in the same year, swung by more than NPR 2 per share depending purely on which report you happened to read.
Look at what each quarter contributed on its own (again, our subtraction from the cumulative figures). The monsoon quarter added about NPR 206 million. The second quarter added about NPR 192 million. The dry third quarter added only about NPR 59 million. The fourth quarter, as flows picked up again, added about NPR 263 million. Annualizing any one of those slices tells you about that season, not about the year.
Now translate that into the number traders react to. ShareSansar reported Chilime’s quarter-end market price at NPR 487 and its PE ratio at 64.15 times on the fourth-quarter EPS. Hold that NPR 487 price constant purely as an illustration. On the first-quarter annualized EPS, the PE would have read about 56 times. On the third-quarter annualized EPS, about 76 times. Same company, same price, a PE range of 20 points produced by the calendar. A trader who bought in November because “the PE dropped” and sold in May because “the PE jumped” would have been trading the weather.
There is one more twist, and it is the one that kills the idea of a simple fix. In the previous fiscal year, 2081/82, Chilime reported nine-month profit of NPR 507.89 million and full-year unaudited profit of NPR 679.59 million, according to ShareSansar. Annualize that nine-month figure and you get about NPR 677 million, almost exactly right. The same method, applied to the same company one year later, missed by 15 percent. So you cannot just apply a fixed “hydropower discount” to annualized EPS. The error depends on rainfall, plant outages, flood damage and when the company books its other income. It moves every year.
It is not only hydropower
Banks are often treated as the safe case, because their income (interest on loans) does not depend on rainfall. ShareSansar tested this directly in a December 2019 analysis. It took each commercial bank’s first-quarter profit for fiscal year 2075/76, annualized it, and compared the result with the actual year-end EPS. Even in this “low seasonality” sector, the gaps were large. Only 8 of 27 listed banks ended the year below their first-quarter projection. Agricultural Development Bank’s actual EPS came in NPR 26.31 above the projection, while Nepal Bank posted year-end EPS of NPR 26.62 against a projected NPR 37.26. On average, the projection was NPR 23.36 and the actual result NPR 24.99.
Notice the direction. For most banks in that year, annualizing the first quarter understated earnings, because a growing loan book earns more in the fourth quarter than in the first. For Chilime in 2082/83, annualizing the first quarter overstated earnings. The bias flips by sector, which is exactly why a single rule of thumb fails.
Banks carry a second hazard. Their fourth-quarter numbers are unaudited, and the final audited accounts, which for banks also pass through Nepal Rastra Bank’s review before dividends are approved, can include additional loan-loss provisions or other adjustments. A single bad-debt write-off or a large recovery in one quarter can swing the annualized figure sharply. If you follow banks, our walkthrough of how to read a bank’s quarterly report shows where provisions and one-off items sit in the statements.
Insurers, with lumpy claims, and consumer businesses that peak around Dashain and Tihar have their own versions of the problem. Almost no sector on NEPSE earns money in four identical slices.
The pre-revenue absurdity
Annualization produces its most ridiculous results for companies that are not yet earning from their core business at all. Many newly listed hydropower companies trade for months, sometimes years, before their plants generate commercially. Any “EPS” they report in that window usually comes from interest earned on the cash they raised in their IPO.
ShareSansar’s coverage of Himalaya Urja Bikas Company shows how far this can go. For the second quarter of fiscal year 2076/77, the company reported net profit of NPR 21.60 lakh, an annualized EPS of NPR 0.44 and a PE ratio of 494.97 times. ShareSansar noted that the company had not yet started selling electricity; its income came from other sources. A PE of nearly 500 is not a valuation signal. It is a sign that the denominator has nothing to do with the business you are buying.
For a company in that position, ignore EPS entirely. What you are really buying is a construction project, and the questions that matter are about the project’s cost, timeline, power purchase agreement and financing.
Three other ways EPS misleads you
Bonus shares change the denominator
Nepali companies issue bonus shares often, and every bonus issue increases the share count without adding a rupee of profit. Picture a hypothetical company earning NPR 10 crore with 1 crore shares: EPS of NPR 10. It issues a 20 percent bonus. Now there are 1.2 crore shares, and the same NPR 10 crore of profit produces EPS of about NPR 8.33. The business did not shrink by 17 percent. The pie was cut into more slices.
NAS 33 requires companies to restate prior-period EPS for bonus issues so the comparison stays fair, but portals, tip sheets and old screenshots do not always do this consistently. Chilime itself shows the effect in miniature: ShareSansar reported its fiscal 2082/83 net profit rising 6.44 percent while EPS fell from NPR 7.70 to NPR 7.59, because paid-up capital grew 8 percent over the same period. Whenever you compare EPS across years, check whether the share count is the same.
Net profit is not distributable profit
For banks and financial institutions especially, the EPS printed on the report is not the same as the profit that can be paid out. Regulatory reserves, fair-value gains that have not been realized, and other adjustments sit between reported net profit and distributable profit. A bank with strong headline EPS can have much less room to pay dividends than the number suggests. We unpack the gap in distributable profit versus net profit.
One-off gains inflate the number
A land sale, a large recovery of a written-off loan, a dividend from a subsidiary, or a gain on selling an investment can all lift profit in a single quarter. Annualize that quarter and the one-off is effectively counted four times. Before you trust any EPS figure, check the income statement for “other income” and non-operating gains, and ask whether they will repeat.
How to read EPS on NEPSE properly
None of this means EPS is useless. It means the version most people use is the wrong version. Here is a practical way to use the number without being fooled by it.
Know which EPS you are looking at. Every time you see an EPS figure, ask three questions. Is it annualized or actual? Which quarter is it from? Is it based on audited or unaudited accounts? If a portal or a tip does not tell you, assume it is an annualized, unaudited projection, and treat it with the skepticism that deserves.
Prefer trailing twelve months over annualized. A trailing twelve-month (TTM) EPS adds up the most recent four quarters of actual profit. Because Nepali reports are cumulative, you can build it yourself from two reports: take the latest cumulative profit, add last year’s full-year profit, and subtract last year’s cumulative profit for the same period. Divide by the current share count. That gives you twelve months of real earnings, covering every season once, rather than one season multiplied. Most free portals do not show TTM figures prominently, which is part of why this mistake is so common.
Compare like with like. For seasonal sectors, compare a quarter against the same quarter last year, not against the previous quarter. Chilime’s first-quarter profit against last year’s first-quarter profit tells you something real. Its first quarter against its third quarter tells you the monsoon is wetter than the dry season.
Use several years, not one. A single year’s EPS for a hydropower company tells you about one year’s rainfall. Three to five years of audited EPS, adjusted for bonus shares, tells you what the plant typically earns.
The verdict
The annualized EPS figure is the most widely circulated earnings number on NEPSE, and for a large part of the market it is the least reliable. It is not a lie. It is a mechanical projection that assumes every quarter looks the same, in a market where very few businesses work that way. The error is predictable in shape: hydropower looks cheapest after the monsoon quarter and most expensive after the dry season, while growing banks often look more expensive early in the year than they turn out to be. What the error is not is predictable in size, which is why you cannot simply correct for it with a fixed haircut.
The fix is not a better formula. It is a habit. Treat any annualized EPS as a rough projection, never as a fact. Build or find trailing twelve-month earnings. Compare seasonal businesses against their own same-quarter history. Read the audited annual figure when it arrives, and check how far it landed from the quarterly projections that moved the price months earlier. Investors who do this will notice something useful: a lot of the PE-driven buying and selling in NEPSE trading groups is a reaction to the calendar, not to the companies. That is not an edge for the people reacting. It is an edge for the people who know the difference.
This is analysis, not financial advice.
Frequently Asked Questions (FAQs)
1. What is annualized EPS on NEPSE?
Annualized EPS is an estimate of a company’s full-year EPS based on its earnings from the available quarters. It assumes the company will continue earning at a similar rate for the rest of the year.
2. Why can annualized EPS be misleading for hydropower companies?
Hydropower earnings change with water flow and electricity generation. Earnings are usually stronger during the monsoon and weaker during dry months, so multiplying one quarter’s EPS by four can give a misleading yearly estimate.
3. What is better than annualized EPS for analyzing a NEPSE stock?
TTM (Trailing Twelve Months) EPS is generally more useful because it uses actual earnings from the latest four quarters instead of assuming future earnings will be the same as one quarter.
4. Does a higher EPS always mean a stock is better or cheaper?
No. EPS must be considered together with the stock price, P/E ratio, share count and the quality of earnings. A high EPS does not automatically mean a stock is undervalued.
5. How should investors compare EPS for seasonal companies?
Compare the company’s same quarter with the same quarter from previous years rather than comparing different quarters. For a stronger picture, also look at 3–5 years of audited EPS and TTM earnings.
