ETF ki Dukan Strategy Explained: Mahesh Kaushik's Method for Trading ETFs Like a Shopkeeper

What You'll Learn

By the end of this guide, you'll understand:

• What the "ETF ki Dukan" (ETF Shop) idea actually is, and where it came from
• How Mahesh Chander Kaushik filters roughly 320 listed ETFs down to a working list of 75
• Why the method deliberately keeps only equity ETFs and throws out gold, silver, and debt ETFs
• How the RSI (Relative Strength Index) is used to decide which ETF to buy each day
• A cell-by-cell walkthrough of the actual Class 2 Google Sheet, using a real snapshot of live data, so you can read it even if you have never used a spreadsheet formula
• The buying rule, the averaging-down rule, and the profit-booking target, worked through with real numbers
• How the method evolved across four versions, and what each fix teaches you about trading discipline
• An honest look at where a careful investor should slow down and think before copying it

Reading Time: 24 minutes Difficulty Level: Beginner-friendly Prerequisites: A basic idea of what an ETF is helps. If you are new, read Understanding the NIFTY 50 first.

Please read this first. This article is educational documentation of somebody else's publicly taught method. It is not investment advice, not a recommendation, and not a suggestion that you buy, sell, or hold any ETF named on this page. Every ETF ticker you will see below appears only because it happens to sit in the source spreadsheet, not because we think it is a good or bad investment. The author of this site holds no view on any of them. See the full disclaimer at the bottom.

---

Where This Idea Comes From

This article documents and explains a strategy created and taught by Mahesh Chander Kaushik, a SEBI-registered Research Analyst (registration number INH100000908). He has taught it for free on his YouTube channel over several years, and in 2026 he released an updated version as a short course.

The two videos that this guide is based on are his own work, and they are the original source of everything explained below:

• Class 1, selecting the 75 best ETFs: ETF ki Dukan Updated Method, Class 1 (YouTube)
• Class 2, RSI-based swing trading: ETF ki Dukan Updated Method, Class 2 (YouTube)

He also publishes a free Google Sheet alongside Class 2, which does the daily calculation for you. We were given a copy of that sheet ("ETF Shop Updated Method 2026"), and a large part of this article is a careful walkthrough of what is actually inside it. Every formula, every column, and every number quoted in the walkthrough section comes from his sheet, not from us.

Why we wrote this. The original teaching is entirely in Hindi and spread across several long videos. That is a genuine barrier for the many Indian investors who do not speak Hindi comfortably, and for anyone outside India curious about how retail investors here actually operate. This article exists to make his teaching legible to those readers. It is a translation and explanation of his work, not a replacement for it, and certainly not a rival product. If you find this useful, please watch his original videos and subscribe to his channel. He earns nothing from this page, and the credit for every idea below belongs entirely to him.

What this article is not. It is not investment advice, not a recommendation to use this method, and not an endorsement of it. We explain the honest limitations near the end, and we mean them. If you want to actually deploy money this way, talk to a SEBI-registered investment adviser first.

One more thing worth saying upfront, in Kaushik's own spirit: he repeatedly warns that he never charges fees, never sells a course, and never sends friend requests or messages asking for money. Anyone using his name or photo to ask you for money is a scammer. Keep that in mind if you go searching for his content.

His own disclosure, in his words

The sheet itself carries Kaushik's SEBI-mandated disclosure, and it is worth reproducing because it models the kind of honesty you should expect from anyone teaching a method publicly. He states that the ETFs surfaced by the sheet "are not considered as my recommendations or research reports, they are selected automatically by formulas," that he and his relatives also use the sheet for their own trading, so his personal interest is included in every name it selects, that he holds no ownership of one percent or more in any security it names, that he has received no compensation from any of them, and that he has never served as an officer, director, employee, or market maker for any of them.

He adds a data-accuracy warning too: the sheet is fed automatically by Google Finance, the data "may be delayed or wrong," and you should re-check everything on the NSE or BSE website before acting. That warning applies to every number reproduced in this article as well.

---

The Big Idea, a Shop Full of Index Baskets

The name "ETF ki Dukan" translates to "the ETF Shop." The whole strategy is built on one simple mental picture: you are a shopkeeper.

A shopkeeper does not buy a year of stock in a single morning. She buys a little inventory each day, whatever is cheap and in good supply, stacks it on the shelf, and sells items one by one as customers come in at a small markup. She does this every working day. Some days she buys more than she sells, some days the reverse. Over time, the steady churn of buying low and selling a little higher is the business.

Kaushik applies exactly that behaviour to Exchange Traded Funds. Instead of vegetables or groceries on the shelf, the inventory is a set of ETFs. You buy the cheapest, most beaten-down ETF each day, hold it, and sell it once it has risen by a small target. The shop, not any single trade, is the point.

This works with ETFs for one specific reason that will matter throughout this article. An ETF is not a single company's share. It is a whole basket, like a mutual fund that trades on the exchange. NIFTYBEES, for example, holds all 50 NIFTY 50 companies in the same proportion as the index itself. If you want a refresher on how those baskets are built, our portfolio diversification guide covers why a basket behaves so differently from a single stock.

💡 Why This Matters: A single stock can fall and never recover, because one company can genuinely go bad. A broad index ETF is much harder to kill, because it is the average of dozens of companies. That difference is the entire reason this "buy the falling one" method can work for ETFs when the same behaviour would be dangerous with individual shares.

---

Class 1, Stocking the Shelves With the Right 75 ETFs

Before you can run the shop, you need to decide what goes on the shelves. India has roughly 320 ETFs listed, and new ones keep arriving. Most of them trade in tiny volumes and are useless for daily buying and selling. Class 1 is entirely about narrowing that crowd down to a clean, liquid working list of 75.

Here is the process, step by step.

Step 1, Download the full ETF list from the NSE

After the market closes for the day, go to the National Stock Exchange website (nseindia.com), open the Equity section, and find "Exchange Traded Funds." That page lists every ETF with its symbol, its underlying asset, and its traded volume. There is a download button that gives you the whole thing as an Excel file.

Doing this after close matters. During live trading, the day's real volume is not final yet, so you cannot fairly judge which ETFs are actually liquid. Wait for the closing figures.

Step 2, Sort by volume and keep only the liquid names

Open the Excel file. You can delete the columns you do not need (open, high, low, previous close, 52-week high, 52-week low) so the sheet is easier to read. Keep the symbol, the underlying asset, and the volume.

Now sort by volume, largest to smallest. Kaushik sets a rough cut-off of 100,000 shares traded per day. Anything trading more than that stays; anything below it is removed. After this single filter, the list drops from around 320 ETFs to roughly 186.

He is careful to say there is no magic in the exact number. An ETF might trade 105,000 shares one day and 90,000 the next. The point of the volume floor is not precision. It is liquidity. High-volume ETFs are easy to buy and sell without getting stuck, so you never face a situation where you want out and there are no buyers.

Step 3, Remove gold, silver, and other commodity ETFs

This is the first big judgement call, and it is new in the 2026 version. Kaushik removes every gold and silver ETF from the list. Names like the gold BeES and silver BeES funds are gone.

Why throw out gold, when it has been one of the best performers recently? Because a commodity ETF is not a diversified basket. A gold ETF tracks one thing, the price of gold. A silver ETF tracks one thing, silver. There is no averaging benefit from a spread of companies, and no built-in tendency to drift upward the way a broad equity index does.

His worry is specific. Gold and silver had run up sharply, and a commodity that has tripled can also enter a long, grinding downtrend. If you keep buying and averaging a falling commodity ETF, your money can get stuck for a very long time, because there is no basket of businesses underneath it pulling it back up. If you want to understand how gold, the dollar, and equities actually move against each other, we cover that in the gold, dollar, rupee, and stock market connection.

Step 4, Remove debt, liquid, and gilt ETFs

Next, out go the debt-style ETFs. If the name contains words like "liquid," "one-day rate," "gilt," or "government bond," it is removed. These are bond and cash-equivalent products, not equity baskets, and the shop is meant to hold only pure stock-market baskets. If the difference between bonds and equities is fuzzy for you, bond market vs stock market is a good primer.

What survives is only equity, index-based ETFs, the ones that hold a basket of many shares.

Step 5, Keep one ETF per sector, the highest-volume one

This is the most hands-on part. Among the surviving equity ETFs, many track the same underlying index. There might be five or seven different ETFs all tracking the NIFTY 50. Kaushik keeps only one ETF per sector or index, the one with the highest volume, and deletes the rest.

So for the IT sector he keeps the single most-traded IT ETF (ITBEES, built on the Nifty IT index) and drops the others. For the NIFTY 50 he keeps the most-traded one (NIFTYBEES). For metals, pharma, midcaps, private banks, and every other theme, one ETF each, always the most liquid.

There is one deliberate exception. Within the NIFTY 50 family, an equal-weight ETF and a "value 20" ETF are kept alongside the plain NIFTY 50 ETF. Even though all three are based on the NIFTY 50, they behave differently. An equal-weight version gives every one of the 50 stocks the same importance, so it moves on a different rhythm than the standard market-cap-weighted index. The value 20 version holds only about 20 of the cheaper stocks. Different theme, different behaviour, so they earn their place.

Why he does this by hand, not with AI or filters

Kaushik makes a point of doing this selection manually, going down the list line by line. He compares it to an old grandmother picking tiny stones out of a bowl of lentils one at a time, rather than shaking them through a sieve and hoping. A sieve, or an automatic filter, or an AI, tends to leave a few stones behind, because the same words repeat confusingly ("Nifty Metal Index" and "Nifty Metal Total Return Index" look almost identical). Doing it by eye, once a year, is how he makes sure no junk sneaks onto the shelf.

The finished product is a list of 75 ETFs, each the highest-volume representative of its sector, each trading over 100,000 shares a day. He shares the actual list as a downloadable sheet, and he suggests refreshing it just once a year, because the same large ETFs tend to stay liquid year after year.

Here is the actual top of that list, taken from the 2026 sheet. The order is by trading volume, highest first, which is why these particular names sit at the top.

Listed for illustration of the selection logic only. This is not a recommendation to buy any of them.

The remaining 67 names run all the way down to row 77, covering banks (BANKBEES), PSUs (CPSEETF, PSUBNKBEES), FMCG (FMCGIETF), autos (AUTOIETF), energy (ENERGY, OILIETF), infrastructure (INFRAIETF), consumption (CONSUMBEES), and factor themes like momentum (MOMENTUM50), quality (QUAL30IETF), low volatility (LOWVOLIETF), and value (VAL30IETF). A handful of international equity baskets sit in there as well, such as MON100 (Nasdaq 100), MAFANG (US mega-cap tech), and HNGSNGBEES (Hang Seng). Note that these are still equity baskets, which is why they pass the filter, even though they are not Indian.

One quiet inconsistency worth noticing. The rule says one ETF per sector, but the finished list clearly holds several ETFs that overlap heavily. MOMENTUM50, MOMENTUM30, MOM30IETF, MOM100, and MOMIDMTM are all momentum products. ALPHA and ALPHAETF both chase the same factor. This is not a criticism so much as a caution: when you copy someone's finished list, you inherit their judgement calls, including the loose ones. Doing the Class 1 exercise yourself, even once, teaches you more than downloading the answer.

💡 Why This Matters: Notice that the entire Class 1 is about removing things, not adding them. Liquidity floor, no commodities, no debt, one per sector. What you are left with is a curated shelf of broad, liquid, resilient baskets. A large part of good investing is simply refusing to hold the wrong things, a theme we return to in Valuation 101, when to buy.

---

Class 2, Running the Shop With RSI

Now you have 75 ETFs on the shelf. Class 2 answers the daily question: which one do I buy today, and when do I sell?

The tool for that decision is the RSI.

RSI in one minute

RSI stands for Relative Strength Index. It is a number between 0 and 100 that measures whether something has been bought or sold too aggressively lately. The standard version looks at the last 14 days and compares average gains to average losses.

Here is the whole idea without any mathematics. Take the last 14 trading days. On some days the price went up, on some it went down. Add up all the up-moves and average them. Add up all the down-moves and average them. Now ask: which side was bigger?

• If up-moves dominated, RSI climbs toward 100.
• If down-moves dominated, RSI sinks toward 0.
• If the two are perfectly balanced, RSI sits at exactly 50.

That is genuinely all it is. RSI is a tug-of-war score between buyers and sellers over the last fortnight.

A worked example. Suppose over 14 days an ETF had 9 up-days averaging 0.60% each and 5 down-days averaging 0.40% each.

• Average gain across all 14 days: (9 × 0.60) ÷ 14 = 0.386
• Average loss across all 14 days: (5 × 0.40) ÷ 14 = 0.143
• RS, the ratio of the two: 0.386 ÷ 0.143 = 2.70
• RSI = 100 minus (100 ÷ (1 + 2.70)) = 100 minus 27.0 = 73

An RSI of 73. Buyers clearly won the fortnight. Now flip it, 5 up-days and 9 down-days of the same sizes, and RS becomes 0.37, giving an RSI of about 27. Sellers won. You never have to do this by hand, but seeing it once removes the mystery.

The two rules of thumb everyone learns are simple. An RSI above 70 means "overbought," a lot of people have been buying and the price may be stretched. An RSI below 30 means "oversold," a lot of people have been selling, perhaps more than was justified, and the price may be near fair value or cheap.

Hold on to a caution here, because it matters later: those 70 and 30 lines are conventions, not laws. As you will see when we open the real sheet, on an ordinary day not a single one of the 75 ETFs may be below 30. The method does not actually wait for the textbook oversold line. It buys the lowest RSI available, whatever that number happens to be.

The rule that gets flipped for ETFs

Here is the clever twist, and it is the heart of the whole method.

For an individual stock, Kaushik would never tell you to buy something falling with a low RSI, because a single company can keep falling for good reasons. For stocks, you buy strength, not weakness.

For an ETF, he flips that rule on purpose. Because an ETF is a whole portfolio, a falling ETF with a low RSI is usually just the market having a bad few days, not a business dying. So you deliberately accumulate the cheapest, most oversold ETF, the one with the lowest RSI, and wait for the basket to bounce back. When it does, you sell into the recovery. Buying the falling basket is the feature, not the bug.

So the daily buy decision becomes almost boringly clear: buy the ETF with the lowest RSI that you do not already own.

---

Inside the Class 2 Sheet, Cell by Cell

This is the part that is hardest to follow in a Hindi video if you do not speak Hindi, so we are going to go slowly. Everything below describes the actual "ETF Shop Updated Method 2026" Google Sheet that Kaushik distributes free with Class 2.

The remarkable thing about it is how small it is. One tab. Three working columns. A five-row summary box. That is the entire machine.

The layout

Open the sheet and you are looking at this:

Three columns and a small box. Nothing else.

Column A, the tickers

Each cell holds a ticker with an exchange prefix, like NSE:PHARMABEES. The NSE: prefix is not decoration. It tells Google Finance which exchange to look on, because the same short code can mean different things on different exchanges. Without the prefix you get errors or, worse, the wrong instrument.

This column is the output of Class 1. It is the shelf. You rebuild it roughly once a year and otherwise leave it alone.

Column B, the live price

Every cell in column B contains a variation of this:

=IFERROR(GOOGLEFINANCE(A3, "PRICE"), "#N/A")

Read it in two halves.

GOOGLEFINANCE(A3, "PRICE") is a built-in Google Sheets function. It says: go and fetch the current price of whatever ticker is sitting in cell A3. Because A3 holds NSE:ITBEES, this returns the live price of that ETF. You never type a price yourself.

IFERROR(..., "#N/A") is the safety net. Google Finance does not have clean coverage of every Indian ETF, especially newer or thinly-followed ones. When it cannot find a price, it returns an ugly error that would break other formulas on the sheet. IFERROR catches that and quietly writes #N/A instead. In the snapshot we examined, 51 of the 75 ETFs showed #N/A in the price column while only 24 returned a real price. The sheet keeps working anyway, because the buy decision runs off RSI, not off CMP. But it tells you something important about relying on free data, which we come back to in the bear case.

Column C, the RSI

This is the clever bit. Every cell in column C contains just:

=GET_RSI(A3)

Six characters and a cell reference, and out comes a live 14-day RSI. There is no GET_RSI function in Google Sheets. Kaushik built it himself using a feature called a named function, which lets you wrap a long, ugly formula in a short, friendly name and reuse it everywhere.

This is why people who try to rebuild the sheet from scratch get an error. If you open a blank Google Sheet and type =GET_RSI(A3), nothing happens, because the named function does not exist in your file. It lives in his file, under Data, then Named functions. When you copy his sheet, the function travels with the copy. That is the whole reason he tells you to copy rather than recreate.

What is actually hiding inside GET_RSI

You do not need this to use the sheet. But if you have ever wondered what a "named function" really conceals, here is the genuine definition, unpacked into plain English. It takes one input, a ticker, and then:

1. Sets the period to 14. The standard RSI lookback.
2. Fetches 150 calendar days of daily closing prices for that ticker from Google Finance. Why 150 days for a 14-day indicator? Because Wilder's smoothing (step 5) is recursive, so it needs a long run-up of history before the number settles down. Roughly 100 trading days of warm-up is plenty.
3. Filters out anything that is not a number, throwing away holidays and gaps in the data.
4. Calculates day-to-day differences, then splits them into two lists: gains (down-days recorded as zero) and losses (up-days recorded as zero, and the losses stored as positive numbers).
5. Smooths both lists using Wilder's method, where each new average is the previous average times 13, plus today's value, all divided by 14. This is the recursive smoothing that makes RSI stable rather than jumpy.
6. Takes the final smoothed gain divided by the final smoothed loss to get RS.
7. Returns 100 minus (100 divided by (1 plus RS)), the standard RSI formula.

Kaushik mentions in the video that his RSI uses exponential-style smoothing, so his numbers can differ by a point or two from what you see on Moneycontrol or TradingView. Step 5 is exactly why. Wilder's smoothing weights recent days more heavily than a plain average would, and different sites make slightly different choices about the warm-up period. A gap of one or two points between sources is completely normal and is not a bug.

💡 Why This Matters: Look at what this design achieves. All the complexity is hidden in one place, written once, and the daily-use surface is a single readable formula repeated 75 times. If you ever build a tracker of your own, copy this pattern rather than pasting a 15-line formula into 75 cells. When the logic needs fixing, you fix it in one place instead of 75.

The "Today Buy" box, where the decision actually happens

Column C gives you 75 numbers. Nobody wants to eyeball 75 numbers every morning to find the smallest. So the sheet does it for you, in a small box beginning at E3 with the label "Today Buy."

The headers in row 4 are Rank, RSI, ETF Code. Then five rows, ranked #1 to #5. The formulas are these:

F5:  =MIN(IFNA(C3:C77,""))
F6:  =SMALL(IFNA(C3:C77,""),2)
F7:  =SMALL(IFNA(C3:C77,""),3)
F8:  =SMALL(IFNA(C3:C77,""),4)
F9:  =SMALL(IFNA(C3:C77,""),5)

G5:  =XLOOKUP(F5,C3:C77,A3:A77)

Three functions to understand, and each is genuinely simple.

MIN finds the smallest number in a range. MIN(C3:C77) returns the lowest RSI among all 75 ETFs. That is rank #1, the most beaten-down basket on the shelf.

SMALL(range, n) is MIN's more flexible cousin. It returns the n-th smallest value. SMALL(C3:C77, 2) gives the second-lowest RSI, SMALL(C3:C77, 3) the third, and so on down to fifth. This is how the box fills ranks #2 through #5. Five ranks, one per trading day of the week.

IFNA(C3:C77,"") wraps the range and converts any #N/A into blank text before MIN or SMALL sees it. Without this, one broken RSI would poison the whole calculation and the box would show an error instead of a number.

XLOOKUP(F5, C3:C77, A3:A77) is the translator. Rank #1 gives you a number, say 37.56, but a number is not tradeable. XLOOKUP says: find the value in F5 somewhere in the RSI column, and when you find it, return whatever sits in the same row of the ticker column. So the raw RSI figure becomes a name you can actually act on.

A real snapshot, read together

Here is the actual "Today Buy" box from the sheet, timestamped 3 August 2026, 3:29 pm.

Reproduced to explain how the sheet works, on one specific historical day. These are not recommendations, and by the time you read this every one of these numbers will have changed.

Now read that box the way the method intends. On this particular day, the shopkeeper's shelf-restocking candidate is CPSEETF, at RSI 37.56, unless she already owns it, in which case she moves down to MOCAPITAL, and so on.

And here is the single most important observation in this entire article, which you can only make by looking at real data rather than the theory:

Not one of the ETFs was below the textbook oversold line of 30. The lowest RSI on the entire shelf was 37.56. For context, here is the spread across the 74 ETFs that returned a valid RSI reading that day:

The market was broadly strong that day. The median ETF sat at RSI 59, the highest reading was 75.17, and more ETFs were technically overbought (7) than were anywhere near oversold. In that environment, "buy the most oversold ETF" really means "buy the least-strong ETF," which is a meaningfully different statement. Rank #1 at 37.56 is not a bargain in the textbook sense. It is simply the weakest name in a strong crowd.

💡 Why This Matters: This is the difference between an absolute rule and a relative rule, and it is the sort of thing that gets lost in translation from a video to a summary. An absolute rule would say "buy only when RSI drops below 30," and on 3 August 2026 it would have told you to buy nothing at all. The sheet uses a relative rule: it always produces five names, every single day, no matter what the market is doing. That guarantees you always have something to do, which is good for the discipline of a daily routine, but it also means the sheet will never tell you to sit out. In a market that is expensive across the board, it will still hand you five tickers each morning. Knowing which kind of rule you are following is the reader's job, not the spreadsheet's.

The one-page summary of the sheet

That is the entire Class 2 engine. Six ideas.

The daily buying routine

Put together, the shop runs like this. You buy one ETF per trading day. There are five trading days in a week, which is why the method surfaces exactly five lowest-RSI ETFs, one candidate per day.

Working from the top of the "Today Buy" box, you buy the cheapest ETF you do not already hold. Suppose on day one the lowest RSI belongs to ETF A. You buy A. Day two, A is still ranked #1 but you already own it, so you move down to rank #2, and buy that. Day three, rank #3, and so on. You are steadily stocking the shelf, one basket a day, always adding the weakest name you are missing.

A worked week. Using the 3 August snapshot as a starting point, and assuming for simplicity that the rankings hold steady (in reality they shuffle daily), a beginner's first week would look like this:

An illustration of the mechanics using one day's historical data. Not a suggested trade plan.

After one week you hold five different ETFs, each bought on a separate day, each the weakest available name at the moment you bought it. Notice that you have not made a single judgement call. You have not formed an opinion about power stocks or capital-goods stocks. You just read the top row of a box and acted.

Averaging down, but only with a rule

Now comes Monday of week two. All five names in the box are ones you already own. What then?

This is where the averaging rule kicks in. You check whether any ETF you hold has fallen more than 3% below the price at which you last bought it. If yes, you buy that one again, adding to your position at a lower price. If nothing has fallen 3% since your last buy, you simply do not buy that day, and that is fine. As Kaushik puts it, tomorrow the sun will rise again and some other fruit will be ripe.

The 3% is measured from your last actual buy price, not from your running average. There is a neat mental shortcut: to find the price 3% below, multiply your last buy price by 0.97.

A worked averaging ladder. Say you bought an ETF at ₹94.

Every purchase resets the trigger lower. This spacing is the whole point. Without it, a jittery price that wobbles between ₹93 and ₹94 for a fortnight would tempt you into five buys within a rupee of each other, leaving you with a large position and no cushion. With it, five buys can only happen across a genuine 12% decline.

Note what this does to your average cost. After those four buys, one unit each, your average is (94 + 91 + 88 + 85) ÷ 4 = ₹89.50. The ETF only needs to climb back to about ₹93.60 for the whole position to hit the 4.57% profit target, even though your first purchase was at ₹94. That is the arithmetic that makes averaging down work on a diversified basket, and the same arithmetic that makes it dangerous on a single stock that may never recover.

💡 Why This Matters: The whole buying system is designed to force patience. You cannot buy a big position in one day. You cannot average down just because a price wobbled. Every rule adds a gap between purchases. For most beginners, the hardest part of investing is not knowing what to buy, it is not buying too much too fast. Mechanical rules like these exist to save you from your own excitement.

---

Selling, the Profit-Booking Target

A shopkeeper has to sell, not just stock up. Class 2 sets a small, fixed profit target and a one-sale-a-day rhythm.

Kaushik walks through the maths of picking a target. A very tight target of 3.14% (yes, he uses the value of pi, half in fun and half from a chapter he wrote on the "physics" of markets) triggers too many trades. Brokerage costs and the 20% short-term capital gains tax eat most of a 3% gain. A wide target of 6.28% (double pi) has the opposite problem: your money sits waiting too long, and capital rotates slowly. If short-term versus long-term capital gains tax is new to you, the same tax mechanics come up in our derivatives explained for investors guide.

The 2026 version splits the difference and sets the target at 4.57%. If any ETF you hold is showing more than about 4.5% profit, you can book it.

Here is the target logic at a glance.

Working out your sell price

To find your target price, multiply your average buy price by 1.0457. That is the mirror image of the 0.97 trick used for averaging down.

Carrying forward the averaging example, where four purchases at ₹94, ₹91, ₹88, and ₹85 gave an average cost of ₹89.50:

• Target price: 89.50 × 1.0457 = ₹93.59
• Gross profit on 4 units: (93.59 − 89.50) × 4 = ₹16.36

Now subtract the frictions that beginners forget, using round numbers for a discount broker:

Illustrative arithmetic, not a projection. Your actual costs depend entirely on your broker and tax situation.

This is why position size matters so much in this method. On a four-unit position worth around ₹358, a 4.57% gain is almost entirely eaten by fixed costs. On a position of ₹50,000, the same 4.57% is ₹2,285, and fixed brokerage becomes trivial. The strategy's arithmetic only works at a size where fixed per-trade costs are small relative to the target. Kaushik's own discussion of why 3.14% was too tight is exactly this point, and it applies with even more force to small accounts.

A quick note on that tax, because people get it wrong constantly: 20% short-term capital gains tax means 20% of your profit, not 20% of your capital. On a ₹300 gain, the tax is ₹60, not ₹2,000.

The selling rhythm

The selling rhythm mirrors the buying rhythm: sell one ETF per day. If several holdings are above target, you sell the single most-profitable one and let the rest keep running. Kaushik frames it as plucking the one ripest fruit from the tree each day, leaving the others to ripen further. In a rising market this means you never panic-sell your whole shelf at once, you just take one steady profit a day and let winners keep winning.

Put the two rhythms side by side and the shop becomes very easy to describe:

Two decisions a day, both mechanical, both taking about five minutes. That is the entire operating manual.

---

How the Method Evolved, and What Each Fix Teaches

One of the most useful things about this strategy is that Kaushik shows his working. He has publicly revised it four times, and every revision fixes a real mistake he lived through. You learn more from the evolution than from the final rules, because you see how a method gets stress-tested.

Version 1 (about 2.5 years ago, over 1 million views). The shelf was ranked using a 20-day moving average. The ETF that had fallen furthest below its 20-day average got rank one, and you bought the top-ranked one each day, averaging down on a 3% fall. The weakness: it only ranked three ETFs. On many days none of the three had changed and none had fallen 3%, so you had nothing to do. Empty days, wasted opportunity.

Version 2. A refinement of the formulas, fixing smaller errors from version 1.

The 2025 version. He swapped the 20-day average for proximity to the 52-week low, on the logic that an ETF near its yearly low is a better bargain than one merely below a short-term average. He expanded the ranked list from three to ten, and added an SIP-style automatic buying option. The weakness: rank ten reached too far up the list and swept in weak names alongside strong ones. He gives a painful real example. A real-estate ETF (he calls it MO Realty) reached rank ten, an automatic SIP started buying it, and then a sudden geopolitical shock (a US-Iran flare-up in his timeline) knocked it down 20%. Because the SIP had triggered its buys clustered near 100 with no 3% gap between them, there was no cushion, and the position showed a 20% loss.

The 2026 version (the one this article documents). Two fixes. First, RSI replaces the moving average, because ranking by oversold RSI is more scientific and needs less guesswork. Second, the list is trimmed from ten ranks back to five, matching the five days of a trading week, so weak names near rank ten stop sneaking in. The averaging rule (3% below last buy) is restored as the disciplined way to add, instead of a gap-less SIP.

💡 Why This Matters: Kaushik keeps quoting Steve Jobs, that without constant improvement we would all still be using old Nokia phones. The investing lesson underneath the phone metaphor is more important than the strategy itself. A method is never finished. You keep a record, you find where it hurt you, and you fix that one thing. The MO Realty story is worth more than any winning trade, because it shows exactly how an untested rule (rank ten, gap-less SIP) turned into a 20% loss, and how tightening the rule removed the risk.

---

The Honest Bear Case, Where a Rational Investor Should Pause

We teach methods here, we do not sell them, so it would be dishonest to present this as a money machine. Kaushik himself is blunt that markets are always risky, that no system is 100% accurate, and that you can lose money. Here is where a careful investor should slow down and think.

There is no independent, audited backtest. The evidence for the method is the creator's own experience and a three-month real-money diary he once kept on his blog. That is honest and useful, but it is not a rigorous, out-of-sample backtest across many market cycles. You are trusting a narrative, not a dataset.

The "oversold" rule is relative, not absolute, and the sheet never tells you to stop. This is the limitation we found by opening the actual spreadsheet rather than listening to the description. On 3 August 2026, not one of the 75 ETFs had an RSI below 30, and the median sat at 59. The sheet still confidently produced five names to buy. It always will. There is no market condition, however expensive, in which the "Today Buy" box comes back empty. A method that generates a buy signal every single day regardless of conditions is not identifying bargains, it is enforcing a savings habit with extra steps. That can be a perfectly reasonable thing to do, but it is worth being clear-eyed that it is what you are doing.

The free data underneath it is patchy. In the snapshot we examined, 51 of the 75 ETFs returned #N/A instead of a price, because Google Finance simply does not cover many Indian ETFs reliably. Kaushik flags this himself in the sheet's data-accuracy warning and tells you to verify on the NSE or BSE site. The RSI column held up better than the price column, but both are fed by the same free source. If a wrong or stale RSI pushes a name to rank #1 on a given day, the sheet has no way to know, and neither do you unless you check independently. Building a daily money routine on a free data feed you cannot audit is a genuine operational risk, not a theoretical one.

It has only been stress-tested in an up-and-to-the-right market. "Buy the falling basket and wait for the bounce" is a wonderful rule while the broad market keeps recovering to new highs. In a long, deep bear market, "the bounce" can take years, and you would be buying more all the way down. The method quietly assumes the Indian market keeps trending up over time. That is a reasonable long-run bet, but it is a bet.

Frequent trading has real costs. Buying and selling almost daily generates brokerage, and short-term gains are taxed at 20%. Even Kaushik admits the tight 3.14% target loses most of its edge to costs and tax. Churn is the enemy of net returns, which is one reason long-term, low-turnover investing usually wins for ordinary people.

It demands daily discipline. The whole thing depends on opening the sheet every trading day and following the rules without emotion, for years. Most people cannot do this consistently. A method you will not actually follow is worse than a simpler one you will.

It is a single-analyst method. Smart, transparent, and free, but still one person's system. Genuine diversification of ideas matters as much as diversification of holdings.

💡 Why This Matters: None of this means the idea is bad. The ETF-selection discipline in Class 1 (liquid, broad, index-only baskets) is genuinely sound and useful even if you never touch the daily trading part. The point of a bear case is not to scare you off, it is to make sure you copy the good parts with your eyes open, and size any experiment small enough that being wrong is survivable.

---

Who This Is Really For

If you are a brand-new investor looking to get rich quickly, this is not it, and Kaushik would be the first to say so. If you are someone who finds the ETF-shelf idea appealing, the more valuable half of this strategy is Class 1, the selection filter. Learning to keep only liquid, broad, index-based baskets, and to reject commodities and debt products from a trading shelf, is a transferable skill.

The daily RSI trading in Class 2 is an active, hands-on hobby for someone who enjoys the routine, has the discipline to follow rules mechanically, and treats it as a small, risk-controlled experiment rather than their core wealth plan. For most people, the core wealth plan should still be boring: broad index funds, held for years, as we argue throughout our portfolio diversification guide.

---

Frequently Asked Questions

What does "ETF ki Dukan" mean?

It translates to "the ETF Shop." It is Mahesh Kaushik's mental model for trading ETFs the way a shopkeeper runs inventory: buy a little of the cheapest stock each day, sell items one at a time at a small markup, and repeat daily. The shop, not any single trade, is the business.

How many ETFs does the strategy use, and why 75?

Seventy-five. India has around 320 listed ETFs, but most trade in tiny volumes. After filtering for daily volume over 100,000 shares, removing gold, silver, and debt ETFs, and keeping only one high-volume ETF per sector, about 75 clean, liquid equity ETFs remain.

Why does the method remove gold and silver ETFs?

Because they are commodity ETFs that track a single thing, with no basket of companies underneath them. A broad equity index tends to drift upward over the long run, so buying its dips is reasonable. A commodity can enter a multi-year downtrend, and averaging down into it can trap your money with nothing pulling it back up.

Which RSI setting does it use?

The standard 14-day RSI. Kaushik's Google Sheet pulls 150 calendar days of price history and applies Wilder's smoothing, so his numbers can differ by a point or two from Moneycontrol or TradingView. That difference is expected, not an error. The rule is to buy the ETF with the lowest RSI that you do not already own.

Why does the formula =GET_RSI(A3) not work in my own Google Sheet?

Because GET_RSI is not a built-in Google Sheets function. Kaushik created it as a named function, a custom shortcut stored inside his specific file under Data, then Named functions. It exists only in that file and in copies of it. If you type the formula into a blank sheet, Google does not recognise the name and returns an error. Make a copy of his sheet (File, then Make a copy) and the function comes along with it.

Does the sheet ever tell you to buy nothing?

No, and this is important to understand. The "Today Buy" box uses MIN and SMALL to find the five lowest RSI values on the list, which means it always returns five names no matter how expensive the market is. On 3 August 2026, the lowest RSI across all 75 ETFs was 37.56, comfortably above the textbook oversold line of 30, and the box still displayed five buy candidates. The decision to skip a day has to come from you, not the spreadsheet.

Why do so many cells show #N/A?

Google Finance does not have reliable coverage of every Indian ETF, particularly newer or thinly-traded ones. The formula wraps each lookup in IFERROR so that a missing price shows as #N/A rather than breaking the rest of the sheet. In the copy we examined, 51 of the 75 price cells showed #N/A. The buy ranking still worked because it runs off the RSI column rather than the price column, but it is a good reminder to verify any number on the NSE or BSE website before acting on it.

How much money do you need for this to make sense?

The article does not recommend any amount, but the arithmetic is worth understanding. The profit target is 4.57% per trade, and every trade carries brokerage, exchange fees, GST, stamp duty, and 20% short-term capital gains tax. On a position of a few hundred rupees, fixed costs can consume most of the gain. On a larger position, the same percentage target produces a profit that dwarfs the fixed costs. Any method built on small percentage targets is highly sensitive to per-trade costs, which is exactly why Kaushik rejected the tighter 3.14% target.

Is the ETF ki Dukan strategy safe?

No strategy is guaranteed safe. It is lower risk than trading single stocks because ETFs are diversified baskets, but it still carries market risk, trading costs, taxes on frequent trades, and the risk of a prolonged market downturn. It is educational content, not investment advice. Consult a SEBI-registered investment adviser before acting on it.

Where can I find the original videos and sheets?

Everything is free on Mahesh Kaushik's YouTube channel. Class 1 (selecting the 75 ETFs) is here and Class 2 (RSI trading with the Google sheet) is here. He also links the 75-ETF list and the RSI Google Finance sheet in those video descriptions, at no cost.

---

Key Takeaways

• ETF ki Dukan means running an "ETF shop." You buy a little of the cheapest ETF each day and sell winners one at a time at a small target, like a shopkeeper managing inventory.
• Class 1 is a subtraction exercise. From around 320 ETFs, keep only those trading over 100,000 shares a day, drop gold, silver, and debt ETFs, and keep one high-volume equity ETF per sector, leaving about 75.
• The buy rule flips normal logic for ETFs. Because an ETF is a diversified basket, you buy the most oversold, lowest-RSI ETF you do not already own, expecting the basket to recover.
• The Class 2 sheet is only six ideas. GOOGLEFINANCE for prices, a custom GET_RSI named function for the indicator, IFERROR and IFNA to survive missing data, MIN and SMALL to rank the five weakest names, and XLOOKUP to turn a number back into a ticker. Understand those six and you understand the whole machine.
• Discipline is built into the rules. One buy a day, one sale a day, average down only on a 3% drop from your last buy price (multiply by 0.97), and book profit at a 4.57% gain (multiply by 1.0457).
• The "oversold" signal is relative, not absolute. On the real day we examined, zero ETFs were below RSI 30 and the median was 59, yet the sheet still produced five buy candidates. It will never tell you to sit out. That judgement remains yours.
• Costs decide whether the arithmetic works. A 4.57% target is small enough that brokerage, exchange fees, and 20% short-term capital gains tax matter enormously, especially on small positions.
• The method improved by fixing real mistakes. Four versions, from 20-day averages and three ranks to RSI and five ranks, each fixing a flaw the creator lived through.
• Copy the good parts with open eyes. The ETF-selection discipline is genuinely useful. The daily trading is an active hobby that assumes an ever-rising market, costs money in fees and tax, and demands relentless discipline.
• This is education, not advice. Full credit and thanks to Mahesh Chander Kaushik, who created and teaches all of this for free. Watch his original videos, and talk to a registered adviser before risking real money.

---

Learn More

The original teaching, in Mahesh Kaushik's own words:

• ETF ki Dukan Updated Method, Class 1, selecting the 75 ETFs (YouTube)
• ETF ki Dukan Updated Method, Class 2, RSI-based swing trading (YouTube)

Related reading on The Rational Investor:

• Understanding the NIFTY 50, how India's main index and its ETFs are built
• Portfolio Diversification Guide, why baskets behave differently from single stocks
• The Gold, Dollar, Rupee, and Stock Market Connection, why commodity ETFs move on their own logic
• Bond Market vs Stock Market, the difference between the debt ETFs this method drops and the equity ETFs it keeps

---

Credit and Disclaimer

Credit. The "ETF ki Dukan" strategy, the 75-ETF selection method, the RSI ranking system, the GET_RSI named function, and the Google Sheet described in this article are entirely the original work of Mahesh Chander Kaushik (SEBI-registered Research Analyst, registration number INH100000908). He created it, refined it across four versions, and teaches it publicly for free on YouTube. This article is an English-language explanation of his Hindi teaching, written so that non-Hindi speakers can learn the concept. We claim no originality for any idea on this page. Please watch his original videos, linked above, and support his channel.

Disclaimer. This article is for educational purposes only. It is not investment advice, not a recommendation, and not a solicitation to buy or sell any security. Nothing here should be read as an endorsement of the ETF ki Dukan strategy or as a suggestion that you should follow it.

Every ETF ticker mentioned in this article appears solely to illustrate how the source spreadsheet works. No ETF named on this page is being recommended. The RSI values, prices, and rankings reproduced here are a snapshot from a single day, 3 August 2026, shown to explain the mechanics of the sheet. They are historical, they are already out of date, and they carry no predictive value whatsoever.

The data in the source sheet is pulled automatically from Google Finance and, as its own creator warns, may be delayed or incorrect. A majority of price cells in the copy we examined returned no data at all. Verify everything independently on the NSE or BSE website.

Investing and trading in the stock market carries risk, including the permanent loss of your capital. Frequent trading additionally incurs brokerage, exchange charges, GST, stamp duty, and short-term capital gains tax, all of which reduce returns. Past performance does not predict future results, and no strategy, however disciplined, can eliminate market risk.

Always do your own research and consult a SEBI-registered investment adviser, licensed in your jurisdiction, before making any investment decision.

The Rational Investor has no affiliation with, and receives no compensation from, Mahesh Chander Kaushik or any fund house whose ETFs are named in this article.