Start free

THE RESEARCH WORKSPACE BUILT FOR INDIAN MARKETS

An idea is a start.
Know what survives.

Turn your trading idea into a strategy. Test it against history.
See what’s left after every charge. Before you risk a rupee.

Start your free trialExplore the story

Paper trading only. Not investment advice.

✳
IT STARTS WITH YOUR WORDS

“Buy when the 20-day average crosses the 50.”

BHAV / STRATEGY LABHISTORICAL SIMULATION
THE IDEA, TESTED

20 / 50 EMA crossover

April 2021 — March 2026
BEFORE CHARGES-3.5%
AFTER CHARGES-6.0%
THE COST OF TRADING₹24,395
After charges Before chargesSample backtest · not a forecast
EVERY RUPEE ACCOUNTED FOR

24 orders. All charges included.

ONE WORKSPACE. MORE PERSPECTIVE.NSEBSECryptoMutual fundsRESEARCH → TEST → UNDERSTAND
A smaller example: can a winning move lose money?
Hypothetical cost check: one NIFTY option lot of 65, bought at ₹120 and sold at ₹121. Gross ₹65.00; charges ₹65.81, 101.2% of the gross; net minus ₹0.81. The premium must rise ₹1.01 to break even.

1 NIFTY option lot, ₹120 pe liya, ₹121 pe becha. Charges ke baad kitna bacha?

Round-trip charges · NIFTY option · 1 lot (65) · NSEHypothetical trade

Nothing. The charges were bigger than the move.

101.2%of the gross went to charges

Gross, 65 × ₹1.00
₹65.00
Charges
minus ₹65.81
Net
minus ₹0.81

Gross ₹65.00 · Charges minus ₹65.81 · Net minus ₹0.81

Points to break-even: 1.01. The premium has to rise ₹1.01 (0.84%) before you keep a paisa.

Rates effective 1 Apr 2026 · ₹20 flat brokerage per order1

See it line by line

THE SPARK

A thought.
A few words.
A testable idea.

You: 20 EMA jab 50 EMA ke upar jaaye toh buy, neeche aaye toh sell. Simple hai na?

English or Hinglish. Start where you are.

11:40 pm · The idea

It always starts with one line crossing another.

The 20-day average climbs over the 50-day. You've seen it on a dozen charts and in one very confident video. By midnight you've half-decided: this is the one.

So you type it out, the way you'd say it to a friend.

What that sentence becomes
You saidBhav tests
'20 EMA jab 50 EMA ke upar jaaye'Buy when the 20-day EMA crosses above the 50-day EMA.
'toh buy'Put 95% of the cash in, whole shares, held as delivery.
'neeche aaye toh sell'Sell the whole position when it crosses back below.
(unsaid)Read the signal at the close; fill at the next session's open, with 2 bps slippage.
(unsaid)Wait 60 sessions for the averages to warm up.
(unsaid)Test it on one large-cap NSE stock, fixed before the run, from 1 Apr 2021 to 31 Mar 2026, with ₹10,00,000.
(unsaid)Charge every order the way a contract note would.

The gaps you didn't fill get defaults that Bhav states before it runs.3

Skip the chart

One idea, five years

Hypothetical backtest

Gross · end of March 2026−3.5% before charges

Net−6.0% after charges

Charges₹24,394.69

Hypothetical backtest of a 20/50 EMA crossover, April 2021 to March 2026. Before charges, equity fell 12.6% by May 2023, rose to 15.1% up on 4 March 2024, and ended 3.5% down. After ₹24,394.69 of charges on 24 orders, it ended 6.0% down. Its largest fall was about 22%, from 4 March 2024 to 7 April 2025.

Where the ₹24,394.69 went

Brokerage
₹0.00
STT
minus ₹21,761.83
Exchange txn charge
minus ₹668.07
GST
minus ₹124.18
SEBI charges + IPFT
minus ₹21.74
Stamp duty
minus ₹1,634.79
DP charges
minus ₹184.08
Total
minus ₹24,394.69
Hypothetical backtest: 20/50 EMA crossover · one large-cap NSE stock, fixed before the run · daily bars · 1 Apr 2021 to 31 Mar 2026 · ₹10,00,000 · delivery, 95% of cash per entry · every order filled at the next day's open with 2 bps slippage · charges at rates effective 1 Apr 2026, applied to past dates · price returns only, no dividends. Historical simulation, not a forecast.34

12:15 am · The backtest you usually see

The best stretch always looks like the whole story.

Bhav fills every order at the next day's open, and for the first 60 sessions it only watches while the averages warm up. The first buy comes on 14 January 2022. Then a long slide, and one strong climb: stop the chart on 4 March 2024 and this idea is up 15.1%, before charges. A quick backtest often stops somewhere like here. It fills at a close you could only have known after the market shut, charges nothing to buy or sell, and measures risk against zero.

12:25 am · The rest of the five years

Then the next two years happened. Before a single rupee of charges, the same idea finished 3.5% below where it started: ₹10,00,000 became ₹9,64,575.70. Nothing changed except where you stopped reading.

12:40 am · Now put the charges back

Every order pays, win or lose.

Bhav charges each of the strategy's 24 orders the way a contract note would. The line grows a shadow: −3.5% before charges becomes −6.0% after them. The 2.44 points in between are ₹24,394.69, about ₹1,016 an order.

12:50 am · Where it went

Mostly one line. On delivery the modelled plan charges no brokerage, but STT is 0.1% of the value on every buy and every sell. Across ₹2.18 crore of turnover that came to ₹21,761.83, or 89% of the total. Then stamp duty ₹1,634.79, exchange transaction charges ₹668.07, DP charges ₹184.08, GST ₹124.18, SEBI charges and IPFT ₹21.74. The charges didn't sink this idea; it was already under water. They made the loss deeper: ₹35,424.30 before charges, ₹59,818.99 after.

Where the ₹24,394.69 went

Brokerage
₹0.00
STT
minus ₹21,761.83
Exchange txn charge
minus ₹668.07
GST
minus ₹124.18
SEBI charges + IPFT
minus ₹21.74
Stamp duty
minus ₹1,634.79
DP charges
minus ₹184.08
Total
minus ₹24,394.69

1:00 am · The whole strategy, net

Twelve trades. Four won. One long fall.

At its worst, on 7 April 2025, the account sat about 22% below its March 2024 high. By the end of March 2026 it still hadn't got back to it. Over five years, after every charge, it lost 1.24% a year.

Hypothetical backtest: 20/50 EMA crossover · one large-cap NSE stock, fixed before the run · daily bars · 1 Apr 2021 to 31 Mar 2026 · ₹10,00,000 · delivery, 95% of cash per entry · every order filled at the next day's open with 2 bps slippage · charges at rates effective 1 Apr 2026, applied to past dates · price returns only, no dividends. Historical simulation, not a forecast.34

Show this chart as a table
Gross and net equity at each financial year's last session
YearSessionGross (before charges)Net (after charges)
Start1 Apr 2021₹10,00,000.00₹10,00,000.00
FY2231 Mar 2022₹9,73,124.60₹9,70,012.72
FY2331 Mar 2023₹8,82,646.50₹8,74,744.73
FY2428 Mar 2024₹11,35,045.60₹11,24,203.31
FY2528 Mar 2025₹9,85,716.10₹9,66,191.77
FY2630 Mar 2026₹9,64,575.70₹9,40,181.01
Gross peak4 Mar 2024₹11,50,690.90 (+15.07%)₹11,39,848.61
Gross low3 May 2023₹8,74,020.58 (−12.60%)—
Largest fall (net)4 Mar 2024 to 7 Apr 2025—−22.15%
Charges by line, across all 24 orders
Brokerage₹0.00
STT₹21,761.83
Exchange txn charge₹668.07
GST₹124.18
SEBI charges + IPFT₹21.74
Stamp duty₹1,634.79
DP charges₹184.08
Total₹24,394.69

7:30 am · Same idea, told the truth

Describe the idea. Bhav does the arithmetic.

Quantwala's engine is called Bhav (भाव): Hindi and Marathi for price, and a nod to the bhavcopy NSE and BSE publish every evening. Tell it your idea the way you'd tell a friend, in English or Hinglish. It states the rules it will test, filling any gaps with sensible defaults it tells you about. Then it writes the strategy, backtests it on NSE and BSE daily price history with every order filled at the next day's open, applies every charge, and explains what happened, with the equity curve and every trade drawn right there in the chat.

What it explains, every time

  • net vs gross return
  • maximum drawdown
  • Sharpe against the 10-year G-Sec
  • trade count and win rate
  • rejected orders
  • a warning when the sample is small or the rules look fitted to the past

You: 20 EMA jab 50 EMA ke upar jaaye toh buy, neeche aaye toh sell. Ek large-cap NSE stock, April 2021 se March 2026, ₹10 lakh capital. Charges ke saath batao.

Bhav: Here's what I'll test.

  1. Buy when the 20-day EMA crosses above the 50-day EMA.
  2. Sell when it crosses back below.
  3. Signals are read at the close; orders fill at the next session's open, with 2 bps slippage.
  4. ₹10,00,000 starting capital; 95% of cash per entry, whole shares, delivery.
  5. The first 60 sessions warm up the averages.
  6. Every charge on an Indian contract note, on every order.

Backtested EMA crossover 20/50. EMA crossover 20/50 on one large-cap NSE stock, daily bars, 1 Apr 2021 to 31 Mar 2026; hypothetical backtest.

Net return
−5.98%
Gross return
−3.54%
Charges
₹24,394.69
Max drawdown
−22.15%
Sharpe vs G-Sec
−0.50
Trades
12 (4 won)

Exits: signal, 12.

Bhav: Before charges this idea lost 3.54%. After every charge it lost 5.98%: the 2.44-point gap is ₹24,394.69 across 24 orders, and ₹21,761.83 of that was STT. Each order filled at the next day's open, because a crossover is only confirmed after the close. Two things before you read much into it. It rests on 12 trades, which is a hint, not proof, and one of them, from November 2023 to May 2024, did most of the lifting. And its Sharpe against the 10-year G-Sec is −0.50: for the risk it took, a government bond did better. Want me to compare it with simply holding the stock over the same five years, or add a 200-day trend filter?

Sample wording, written from the run's figures. Every figure is from a real engine run: commit b888963, generated 29 September 2026.3

What this backtest can't know

  • Daily bars only, so it can't see what happened within a day.
  • Today's charge rates (effective 1 Apr 2026), applied to trades from 2022 to 2026.
  • One stock, chosen before the run: an anecdote, not a sample.
  • A list of today's large caps, which flatters the past (survivorship).
  • Price returns only; dividends are left out.
  • A flat 2 bps of slippage; real fills vary with liquidity.
Test your own idea, free

The same honesty, wherever money leaves quietly

Not just backtests. Everywhere the arithmetic hides.

02 /

Crypto tax, the way Section 115BBH actually works

How it works

Made ₹40,000 on one coin and lost ₹25,000 on another? You'd expect to be taxed on ₹15,000. Section 115BBH taxes the ₹40,000 at 30% plus 4% cess, which is ₹12,480, and the loss can't be set off or carried forward. Of the ₹15,000 you thought you'd made, ₹2,520 is left. Bhav works this out across a whole financial year of closed trades, and counts the 1% TDS already deducted under Section 194S.

Hypothetical example, before any surcharge. Not tax advice; confirm with a chartered accountant.

Check crypto tax
03 /

Mutual funds, with SIP returns as XIRR

How it works

Look up a fund from AMFI NAV data. You get the latest NAV, 3-, 5- and 10-year CAGR, drawdown and volatility, and what a monthly SIP started in the past would be worth today, measured as XIRR: the rate that respects when each instalment actually went in.

Mutual fund investments are subject to market risks. Read all scheme-related documents carefully.

Look up a fund
04 /

Paper trading with ₹10 lakh that isn't real

How it works

Move an idea from backtest to a paper portfolio with ₹10,00,000 of virtual cash. Bhav prepares each paper order as a card, and nothing happens until you press Confirm. Every fill is charged like a real one. No order ever reaches a broker or an exchange. The money is virtual. The lessons aren't.

An illustrative paper order card: buy 50 shares at a ₹2,400.00 limit, order value ₹1,20,000.00, estimated charges ₹142.48, waiting for your confirmation.

Open a paper portfolio

Pricing

Priced like a tool, not a tip.

No tips, no calls, no 'sure-shot' anything. You pay for Bhav's work (reading your idea, writing the strategy, running it and explaining it), measured in tokens.

Loading plans

Strategy counts are estimates; longer conversations use more. Billed monthly through Razorpay. Cancel any time in Settings, and your plan runs to the end of the period you've paid for. Checkout shows the exact amount before you pay. It would be odd for us, of all people, to hide a charge.6

Full fee schedule

Before the open

Ask before the market answers.

NSE and BSE price history, every charge on the contract note, crypto tax the way it actually works, plain English, and nothing real at stake. Test in calm water first.

Start your free trialAlready have an account? Sign in

Gross ₹65.00 · Charges minus ₹65.81 · Net minus ₹0.81

For the curiousOpen the research notebookEvery trade, charge, assumption and source

1:10 am · Twelve trades, one yardstick

One trade did most of the work.

Four of the twelve trades won. One of them, held from 30 November 2023 to 13 May 2024, made ₹1,46,652 before charges. The other eleven lost ₹1,82,076 between them. Every trade paid between ₹1,892 and ₹2,218 in charges, whichever way it went. Twelve trades is a hint, not proof.

Show:
Every trade, before and after charges (hypothetical backtest)
#HeldProfit or lossChargesNet
T114 Jan 2022 – 31 Jan 2022₹2,060.91minus ₹67,954.71
T223 Mar 2022 – 27 Jun 2022₹1,992.12+₹5,341.88
T312 Aug 2022 – 22 Sep 2022₹1,956.99minus ₹38,152.59
T47 Nov 2022 – 5 Jan 2023₹1,891.75minus ₹24,489.85
T53 May 2023 – 21 Sep 2023₹1,908.12+₹41,221.48
T630 Nov 2023 – 13 May 2024+₹1,46,652 before charges₹2,101.86+₹1,44,550.14
T727 May 2024 – 7 Jun 2024₹2,218.29minus ₹37,816.29
T810 Jun 2024 – 9 Aug 2024₹2,167.00minus ₹9,130.00
T92 Sep 2024 – 10 Sep 2024₹2,123.31minus ₹34,610.31
T1026 Mar 2025 – 8 Apr 2025₹1,994.53minus ₹88,070.53
T1123 Apr 2025 – 4 Aug 2025₹1,963.14+₹61,606.86
T1223 Oct 2025 – 16 Jan 2026₹2,016.67minus ₹12,315.07
All twelve₹24,394.69minus ₹59,818.99

Charges: ₹1,892 to ₹2,218 a trade, win or lose.

Sharpe, measured against:

Sharpe against the 10-year G-Sec: minus 0.50

Below 0: it earned less than the yardstick for the risk it took.

Same strategy. Different yardstick.

A quick backtest measures risk-adjusted return against zero. Bhav measures it against a 10-year G-Sec yield, 6.8% a year in this run. A government bond asks nothing of you, and a strategy has to beat it, for the risk it takes, to be worth the trouble. Against zero, this idea scores minus 0.02: roughly nothing. Against the bond, minus 0.50. It doesn't clear the bond.

Gross is a rumour. Net is the bhav.

Same hypothetical backtest. After every charge: −1.24% a year. Sharpe is annualised from daily returns, against zero and against a constant 6.8% a year (the engine's 10-year G-Sec proxy, not a live yield). Entry prices and quantities are withheld so the stock can't be identified.3

Then the rupees start leaving. Quietly, one line at a time.

1:20 am · When you call it right

You called it right. You lost 81 paise.

Put the strategy aside and take one trade: a single NIFTY options lot of 65, bought at a ₹120 premium and sold at ₹121. The market moved your way: ₹65.00, gross. Now watch what leaves before any of it reaches you.

Had the premium gone nowhere, the same lines would have cost ₹65.69. Because brokerage is a flat fee per order, the break-even falls as the premium rises and climbs as it falls.

Every segment charges differently. What never changes: you pay on every order, win or lose.

Worked example from Bhav's own charge table, rates effective 1 April 2026: ₹20 brokerage per order on intraday, futures and options (flat on options; 0.03% capped at ₹20 on the others), none on delivery; NIFTY lot of 65 (NSE circular NSE/FAOP/70616); before income tax. Rates change by circular; your broker's contract note is the final word.12

Cost your own trade
Hypothetical trade

What reaches youNet loss of 81 paise

1 NIFTY options lot of 65, from a ₹120 premium to ₹121.

Contract note for one hypothetical options trade
Gross, 65 × ₹1.00₹65.00
Brokerage₹20 an order, in and outminus ₹40.00
STT0.15% of the sell-side premium of ₹7,865minus ₹11.80
Exchange txn chargeon the premium, both legsminus ₹5.56
GST18% on brokerage, exchange and SEBI chargesminus ₹8.20
Stamp dutybuy side onlyminus ₹0.23
SEBI charges + IPFTminus ₹0.02
DP chargesdelivery sells only₹0.00
Total chargesminus ₹65.81
What reaches youminus ₹0.81

Charges ₹65.81 · 101.2% of the gross · points to break-even 1.01 (0.84%)1

NSE and BSE closed at 3:30 in the afternoon. Crypto didn't.

2:10 am · The market that doesn't close

In crypto, a loss doesn't cancel a gain.

Crypto trades all night, and India taxes it by its own rules. On an Indian exchange, 1% of every sell is withheld as TDS under Section 194S. It comes back as a credit against your tax, but not until you file, so it's missing from your balance the moment you sell. Then, once a year, Section 115BBH taxes your gains at 30% plus 4% cess. A loss on one coin can't be set off against a gain on another, and it can't be carried forward.

One sell

Hypothetical trade

₹1,00,000 of a coin on an INR pair, sold at ₹1,02,000.

One hypothetical crypto sell on CoinDCX
Gross₹2,000.00
Exchange fee, 0.5% each wayminus ₹1,010.00
GST, 18% on the feeminus ₹181.80
Before tax₹808.20
TDS withheld, 1% of ₹1,02,000minus ₹1,020.00
In your balance tonightminus ₹211.80

The ₹1,020 isn't lost: it's credited against your tax when you file. But tonight, a trade that went your way left your balance ₹211.80 lower. A 1.18% move covers the fees; 2.18% covers the TDS too, until you file.

A year under 115BBH

100 trades in one year: 60 win ₹100, 40 lose ₹120

The year's tax arithmetic
Gains, 60 trades₹6,000
Losses, 40 tradesminus ₹4,800
What the trades made together₹1,200
Taxed under 115BBH: gains only₹6,000
Tax, 30% + 4% cessminus ₹1,872
After taxminus ₹672

A year that made ₹1,200 ends ₹672 down.

Tax came to 156% of what the trades made together.

Crypto products and NFTs are unregulated and can be highly risky. There may be no regulatory recourse for any loss from such transactions.

Hypothetical examples from Bhav's crypto cost model, effective 1 April 2026: CoinDCX 0.5% a side plus 18% GST on the fee; Binance 0.1% a side; 194S TDS at 1% on sells where the exchange withholds it (annual thresholds not applied); 115BBH at 30% plus 4% cess on gains, before any surcharge. Not tax advice; confirm with a chartered accountant.5

Work out 115BBH on your own trades

What Quantwala won't do

No tips. No calls. No promises.

Read the Terms of Use · Privacy Policy

  1. Tell you what to buy.

    Quantwala tests your ideas. It doesn't hand you its own, and it won't tell you to buy, sell or hold anything.

  2. Promise a return.

    A backtest describes the past under stated assumptions. It's evidence about an idea, not a forecast.

  3. Touch real money.

    Every Quantwala portfolio is paper and the cash is virtual. No order reaches a broker or an exchange.

  4. Hide the bad parts.

    Drawdowns, losing trades, rejected orders and every charge get the same space as the good parts. You've just watched a strategy lose money on this page.

  5. Pretend to be advice.

    Quantwala is not registered with SEBI as an investment adviser or research analyst. It's a tool for learning and testing. Decisions about real money stay yours, ideally made with a qualified professional.

Notes and sources

How every number on this page was made

Method in full

  1. Worked options example (hero, one order, final line). One NIFTY options lot of 65 (NSE circular NSE/FAOP/70616), bought at a ₹120 premium and sold at ₹121. ₹20 flat brokerage per order; STT 0.15% of the sell-side premium; exchange transaction charge, SEBI fee and IPFT on premium turnover; stamp duty on the buy side; GST 18% on brokerage and fees. Rates effective 1 April 2026. The break-even is the round-trip charge at a flat premium (₹65.69). Before income tax. Your broker's contract note is the final word.

    Methodology ↩ Back to text

  2. Charge table. Eight components (brokerage, STT, exchange transaction charge (NSE's, or BSE's by scrip group), SEBI turnover fee, IPFT, stamp duty, GST, DP charge) across delivery, intraday, futures and options, effective 1 April 2026, after the STT revision to 0.05% of sell-side notional on futures and 0.15% of sell-side premium on options. GST applies to brokerage, exchange, SEBI and IPFT, never to STT or stamp duty. DP charge ₹15.34 per scrip on delivery sells.

    Methodology ↩ Back to text

  3. Sample run. A 20/50 EMA crossover on one large-cap NSE stock, fixed before the first run and not named, so the example can't read as a recommendation. Daily bars: price returns only, dividends excluded. 1 April 2021 to 31 March 2026 (last bar 30 March 2026; 1,235 sessions; first trade 14 January 2022). ₹10,00,000, delivery, 95% of cash per entry, whole shares, 60-session warm-up. Sharpe against zero and against a constant 6.8% a year. Engine b888963, generated 29 September 2026.

    Methodology ↩ Back to text

  4. Fills and limits. A signal is read at the close; the order fills at the next session's open with 2 bps adverse slippage. Daily bars only; today's charge rates applied to past dates; today's large-cap list (survivorship); dividends excluded.

    Methodology ↩ Back to text

  5. Crypto. Bhav's crypto model, effective 1 April 2026: CoinDCX 0.5% a side plus 18% GST on the fee, with 1% TDS withheld on sells (Section 194S); Binance 0.1% a side, no withholding at source, tax self-reported; Section 115BBH 30% plus 4% cess on gains, no set-off, no carry-forward. Hypothetical, before surcharge; annual TDS thresholds not applied. Not tax advice.

    Methodology ↩ Back to text

  6. Plans and prices are read live from the server when this page loads.

    ↩ Back to text

  7. SEBI, study of individual traders in the equity F&O segment (FY22 to FY24), September 2024. Counsel to confirm the exact study and wording before publishing.

    ↩ Back to text

quantwala

Read this before you trade anything

Quantwala is an educational research and paper-trading tool. Nothing on this site or in the app is investment, trading, tax or legal advice, a research report, or a recommendation to buy, sell or hold any security, derivative or crypto asset. Quantwala does not manage money and never places orders with any broker or on any exchange. Quantwala is not registered with SEBI as an investment adviser, research analyst, stock broker or portfolio manager.

Backtests are simulations on historical daily prices. They fill orders at the next day's open and apply charges from the app's rate table, at today's rates, to past dates. They can't capture every real-world cost, liquidity condition or execution detail. Past performance, simulated or real, does not indicate future results. Every example on this page is a hypothetical backtest or a hypothetical calculation, not a prediction and not a recommendation.

Trading in securities, and especially in derivatives, carries a high risk of loss. SEBI's study of individual traders in the equity futures and options segment found that 9 out of 10 incurred net losses.7

Worked charge examples use rates effective 1 April 2026. Charges change, and your broker's contract note is the final word. Tax figures, including the Section 194S and 115BBH examples, are estimates for learning; consult a chartered accountant. Mutual fund investments are subject to market risks; read all scheme-related documents carefully. Crypto products and NFTs are unregulated and can be highly risky. There may be no regulatory recourse for any loss from such transactions.

AI answers can be wrong. Check anything that matters before you act on it. Paper trading uses virtual money only. You must be 18 or older to use Quantwala.

Market data. NSE, NIFTY, BSE and SENSEX are trademarks of their respective owners. Quantwala is not affiliated with, endorsed by or sponsored by NSE, NSE Indices, BSE, Asia Index, SEBI, CoinDCX, Binance or any broker. Figures on this page come from engine run b888963, generated 29 September 2026.

Quantwala (n.): Indian English for 'the quant person'.

© 2026 Quantwala · quantwala.in