NEPSE near record highs: how to judge if the market is expensive

Ask around right now and someone is buying shares, or thinking about it. The market has climbed for two years, dividend and bonus announcements fill the news, and more than seven million demat accounts are open. A cousin who had never touched NEPSE asked me last week whether he had missed the boat or whether it was still a good time to get in.

The honest starting point is that "the market is at record highs" is half true, and the half that is false matters. The index is not at a record. What is near a record is how much you pay for what the market earns, and that is the thing worth understanding before you put money in.

First, the "record" is a valuation record, not an index record

NEPSE closed at its all-time high of 3,198.60 on 18 August 2021. In mid-June 2026 the index was around 2,702, about 15% below that peak, after recovering from roughly 1,807 in the 2022 slump. So on the index alone, there is headroom to the old top, and "record high" is not literally true.

What is at or near records is the price tag. The market-wide P/E ratio, the dividend yield, and the surge of new investors all sit at historic extremes even with the index off its peak. That gap between a not-quite-record index and near-record valuations is exactly why a single number like "2,702" tells you almost nothing about whether shares are cheap or dear. For that you need the three lenses below. Reading them for a single stock is covered in EPS, P/E and book value; this post applies the same lenses to the whole market.

Lens one: P/E against history and peers

The price-to-earnings ratio asks how many rupees you pay for each rupee a company earns. For the whole market in early-to-mid 2026 it sat near 38.

Put that in context two ways. Against NEPSE's own history, the average is about 31 and the prior peak was about 42, so 38 is well above normal and near the top of the range the market has ever reached. Against other countries, the gap is wider:

One 2026 analysis used exactly this comparison to argue NEPSE is "among the world's most expensive stock markets," noting that company profitability (return on equity around 10-12%) doesn't justify paying nearly double what investors pay in India. A rough rule: a P/E above 25 to 30 is rich unless earnings are growing fast enough to grow into it, and Nepal's earnings are not growing at that pace.

Lens two: market-cap-to-GDP

The second lens zooms out from companies to the economy. Market-cap-to-GDP compares the total value of all listed shares to the size of the economy, a gauge Warren Buffett popularised as a quick overvaluation check.

Nepal's total market capitalisation was about Rs 4.6 trillion, putting the ratio near 72.6% in April 2026 after peaking around 83.5% in August 2025. The classic reading treats above 100% as overvalued and 75-90% as fully valued, so Nepal is comfortably in fully-valued territory, not bubble territory, but not cheap.

The caveat matters as much as the number. The ratio climbed largely because share prices rose, not because the economy expanded, with real GDP growth only around 4.6%. When the value of paper rises far faster than the output behind it, the ratio is flashing the same warning the P/E is: prices have outrun fundamentals.

Lens three: dividend yield

The third lens is the simplest. Dividend yield is the cash a share pays out as a percentage of its price. By one 2026 estimate the overall market yield had fallen to just 1-2%.

A falling yield is an expensive-market tell, because yield drops as prices rise faster than payouts. It also frames the choice against safer options plainly: even after FD rates fell to around 4%, a deposit pays you more cash than the average share, and it does so with no capital risk. Shares earn their place by the prospect of the price rising, the capital gain, not the dividend. When the dividend yield is this thin, you are leaning almost entirely on that price-rise bet, from an already-high valuation.

How to read these numbers in Nepal specifically

The lenses work, but Nepal's market has quirks that distort them if you take the headline figure at face value.

• Financials dominate. Banks, finance companies, and insurers make up over half of total market value, around 52%. The market P/E is really a blend: commercial banks traded near 15.8 times earnings, the cheapest and most reasonably valued group, while microfinance ran at 50-100 times, insurance very high, and many hydropower firms had no positive earnings at all. The cheap banks drag the index P/E down; everything pricey pushes it up. So "the market" being at 38 hides a wide spread underneath.
• Bonus shares and book closure distort P/E. Companies issue bonus shares from reserves, which changes share counts and prices around book-closure dates, so the reported P/E can wobble for mechanical reasons that have nothing to do with value. Read valuations away from book-closure season, or look through to underlying earnings.
• Cheap money and leverage are inflating it. Falling deposit rates have pushed savers toward shares, and margin lending lets investors buy with 30% of their own money. Both add buying power that lifts prices independent of earnings, which is exactly the kind of fuel that reverses when rates or sentiment turn.

So is it expensive, and what do you do

By all three lenses, yes, NEPSE is historically expensive and dear relative to peers. But expensive is not a crash signal. A richly valued market can stay rich for years, or grow into its valuation if company earnings rise to meet prices. Plenty of analysts make the bull case alongside the bears.

What a high valuation reliably does is two things: it lowers the return you can expect from here, and it widens the fall if sentiment turns. The sensible responses follow from that, and none of them is a stock tip:

• Temper your return expectations. Buying at a P/E of 38 is buying less future return than buying at 20. Plan accordingly.
• Prefer steady investing to a lump at the top. Averaging in or a mutual-fund SIP spreads your entry price instead of betting it all on today's level.
• Look at what you pay per stock, not the index. The market average hides cheap and dear corners. A reasonably valued bank is a different proposition from a microfinance share at 80 times earnings, even on the same day.

What you actually need to know

1. The index is below its 2021 peak, the valuation is near records. Don't read "NEPSE near highs" as the index; read it as P/E near 38 against a 42 peak, a fully-valued market-cap-to-GDP, and a 1-2% dividend yield. That is the expensive part.
2. Three lenses, with Nepali caveats. P/E against history and peers, market-cap-to-GDP, and dividend yield all point to rich pricing. Adjust for the fact that cheap banks and pricey insurance and hydro are blended into one index number.
3. Expensive lowers returns, it doesn't predict a crash. The response is to temper expectations, average in rather than lump in, and judge each stock's valuation rather than chasing the index from a high base.

Trying to figure out whether to start investing now or wait for a pullback? Email parjanya57@gmail.com with your horizon and how you'd react to a 20% drop, and I'll talk through the trade-off, not a stock pick.

This post is part of the Nepal Money Basics guide — the investing section.