Open now SEBI's comment window on the Closing Auction Session paper closes 3 October. SEBI's CAS comments close 3 Oct. Comment in two minutes →
QuantNifty Talk to us
QuantLiveSharpe AIGuidesBlogAboutFAQContact Talk to us
Algorithmic trading systems · built to spec

Your capital works once.
We make it work twice.

A long-term portfolio is productive and idle at the same time. The stock compounds. The margin locked behind it just sits there. We build the second engine that runs on that margin.

Alpha · cumulative return on ₹1 Cr · backtest, then live
Apr 2024 to 24 Sep 2026 365 sessions shown vs NIFTY 50 TRI
50%100%150%200% 20252026 LIVE
Alpha backtest, +193% Since go-live, +12.6% NIFTY 50 TRI, +6%
Apr 2024Jan 2025Jan 2026Sep 2026
0.00Peak Sharpe ratio
0.0%Avg. ROI per year
0.00%Least drawdown we achieved
0%Months in profit
0Indices · NIFTY & SENSEX
The idea in one picture

Nobody buys a flat
and leaves it empty.

Buy an investment property and you do two things with it. You let it appreciate, and you rent it out. One asset, two income streams. Everybody understands this.

Property · ₹1 crore

Appreciates and earns rent

Capital appreciation
yes
Rental income
yes
Portfolio · ₹1 crore

Appreciates. Earns nothing else.

Capital appreciation
yes
Income from the interest-free margin behind it
nothing

Pledging your holdings is the equivalent of renting out the flat. You still own it. It still appreciates. It now also produces income.

How it works

The Double Engine

Your shares earn the market's return. The margin locked behind those shares earns nothing until you put a second engine on it. Same capital, nothing sold, two returns instead of one.

01 You hold

The stocks work

₹1 crore in NIFTYBEES earns whatever the index earns. Historically that has averaged about 12% a year.

02 You pledge

The margin unlocks

Your broker releases trading margin against those holdings. You have not sold anything and you have not borrowed, so there is no interest to pay.

03 We deploy

The second engine runs

That idle margin funds systematic option strategies on NIFTY and SENSEX. Its return stacks on top of the first.

Capital deployed₹1,00,00,000
Engine one: NIFTYBEES long-term equity 12.0%₹12,00,000
Whale on pledged margin Interest free 36.3%₹36,30,000
Combined return on the same ₹1 crore You did not sell anything or borrow to get here.
32.6% ₹32,56,000 a year
The equity leg assumes a 12% long-run NIFTY CAGR. That is indicative rather than guaranteed, and index returns vary a lot year to year. The engine leg is the backtested figure on ₹1 crore of deployed margin. Pledging applies a haircut, so ₹1 crore of holdings usually releases somewhere around ₹75 to 85 lakh of usable margin. Size the engine to the margin actually released, not to the portfolio value.
Built in-house · running on our own capital

Our proprietary algorithms

Three algos, written from scratch and traded with our own money. Each runs a basket of independent strategy sleeves rather than one setup. The sleeves disagree with each other by design, and that disagreement is what flattens the curve.

Longest record
Apex
NIFTY · 0–1 DTE
0.0%per year
NIFTY 50 TRI
Jan 202020222024Sep 2026
Average month
+3.37%
Monthly income
₹3.37 L
Annual income
₹40.40 L
Maximum drawdown
−5.82%
Sharpe ratio
4.86
Return / drawdown
46.6
Months in profit
86%
700 sessions, Jan 2020 to Sep 2026.
Shaded: the COVID crash. NIFTY fell 37%. Apex took the worst drawdown of its record there, −5.82%, and still ended the period +14.8%.
Highest return
Alpha
NIFTY + SENSEX · 0–1 DTE
0.0%per year
NIFTY 50 TRI
Apr 202420252026Jul 2026
Average month
+6.96%
Monthly income
₹6.96 L
Annual income
₹84.17 L
Maximum drawdown
−4.04%
Sharpe ratio
6.58
Return / drawdown
48.2
Months in profit
100%
347 sessions, Apr 2024 to Jul 2026
Two indices with different expiry days. Every position carries a protective wing, which is what keeps the margin efficient.
Risk first
Whale
NIFTY + SENSEX · drawdown-capped
0.0%per year
NIFTY 50 TRI
Apr 202420252026Aug 2026
Average month
+4.26%
Monthly income
₹4.26 L
Annual income
₹51.22 L
Maximum drawdown
−1.49%
Sharpe ratio
8.00
Worst month
+0.96%
Months in profit
100%
370 sessions, Apr 2024 to Aug 2026. No standing hedge, broker-level stops, drawdown capped at 1.5%. Charges no hedge cost: hedges bought on the day come off these figures.

* Figures are backtested on ₹1 crore of deployed margin, net of 0.2% slippage per fill, the cost of the protective hedge, and all statutory charges (Whale on-demand carries no standing hedge and so no hedge cost). A backtest is not a live track record, and it is not a promise of future returns. These systems are built and run for our own capital. Full disclosures are below.

The grey dashed line on each card is the NIFTY 50 with dividends reinvested, over that algo's own window. Each card is scaled to its own return range, so the benchmark is directly comparable to the algo beside it but not across cards.

Forward test & live · Aug to Sep 2026

The backtest said one thing.
Here is what the market said.

August is a full month of exchange data the strategy was never fitted on, replayed session by session. September is live trading on a real account. Every day it traded is below, wins and losses both. 23 sessions is far too short to prove anything, and we would rather show it than wait until the number flatters us.

Net return+12.56%Aug +7.06% · Sep +5.50%
Winning sessions16 of 2370% of days
Best day+3.82%11 Aug
Worst day−1.07%22 Sep
Max drawdown−1.24%peak to trough
Every session, coloured by returnGrey squares are days the book does not trade. NIFTY runs 0–1 DTE, SENSEX 0 DTE, so the two never share a session. Outlined squares are live.
AugSepMTWTF3 Aug 2026: +0.406% (NIFTY, forward test, 4 positions)4 Aug 2026: −0.198% (NIFTY, forward test, 7 positions)6 Aug 2026: +2.239% (SENSEX, forward test, 3 positions)10 Aug 2026: +1.280% (NIFTY, forward test, 3 positions)11 Aug 2026: +3.816% (NIFTY, forward test, 2 positions)13 Aug 2026: −0.060% (SENSEX, forward test, 10 positions)17 Aug 2026: −0.243% (NIFTY, forward test, 7 positions)18 Aug 2026: −0.008% (NIFTY, forward test, 6 positions)20 Aug 2026: +1.047% (SENSEX, forward test, 4 positions)24 Aug 2026: +0.021% (NIFTY, forward test, 4 positions)25 Aug 2026: −0.772% (NIFTY, forward test, 10 positions)27 Aug 2026: +0.093% (SENSEX, forward test, 11 positions)31 Aug 2026: −0.563% (NIFTY, forward test, 8 positions)1 Sep 2026: +1.000% (NIFTY, live)3 Sep 2026: +0.650% (SENSEX, live)7 Sep 2026: +0.700% (NIFTY, live)8 Sep 2026: +0.270% (NIFTY, live)10 Sep 2026: +0.850% (SENSEX, live)15 Sep 2026: +1.320% (NIFTY, live)17 Sep 2026: +0.650% (SENSEX, live)21 Sep 2026: +0.040% (NIFTY, live)22 Sep 2026: −1.070% (NIFTY, live)24 Sep 2026: +1.090% (SENSEX, live)
Loss Profit
Cumulative returnPercent of the capital the two books share, plotted per session traded. August is replayed net of slippage, hedge cost and statutory charges; September is the live account’s actual result.
SEPTEMBER · LIVEAUGUST · FORWARD TEST 3%6%9%13% 3 Aug: +0.406% on the day, +0.41% cumulative (forward test)4 Aug: −0.198% on the day, +0.21% cumulative (forward test)6 Aug: +2.239% on the day, +2.45% cumulative (forward test)10 Aug: +1.280% on the day, +3.73% cumulative (forward test)11 Aug: +3.816% on the day, +7.54% cumulative (forward test)13 Aug: −0.060% on the day, +7.48% cumulative (forward test)17 Aug: −0.243% on the day, +7.24% cumulative (forward test)18 Aug: −0.008% on the day, +7.23% cumulative (forward test)20 Aug: +1.047% on the day, +8.28% cumulative (forward test)24 Aug: +0.021% on the day, +8.30% cumulative (forward test)25 Aug: −0.772% on the day, +7.53% cumulative (forward test)27 Aug: +0.093% on the day, +7.62% cumulative (forward test)31 Aug: −0.563% on the day, +7.06% cumulative (forward test)1 Sep: +1.000% on the day, +8.06% cumulative (live)3 Sep: +0.650% on the day, +8.71% cumulative (live)7 Sep: +0.700% on the day, +9.41% cumulative (live)8 Sep: +0.270% on the day, +9.68% cumulative (live)10 Sep: +0.850% on the day, +10.53% cumulative (live)15 Sep: +1.320% on the day, +11.85% cumulative (live)17 Sep: +0.650% on the day, +12.50% cumulative (live)21 Sep: +0.040% on the day, +12.54% cumulative (live)22 Sep: −1.070% on the day, +11.47% cumulative (live)24 Sep: +1.090% on the day, +12.56% cumulative (live)
3 Aug24 Sep 2026
Session by session, all 23 of them
DateBookSourcePositionsReturnCumulative
3 Aug 2026NIFTYForward test4+0.406%+0.41%
4 Aug 2026NIFTYForward test7−0.198%+0.21%
6 Aug 2026SENSEXForward test3+2.239%+2.45%
10 Aug 2026NIFTYForward test3+1.280%+3.73%
11 Aug 2026NIFTYForward test2+3.816%+7.54%
13 Aug 2026SENSEXForward test10−0.060%+7.48%
17 Aug 2026NIFTYForward test7−0.243%+7.24%
18 Aug 2026NIFTYForward test6−0.008%+7.23%
20 Aug 2026SENSEXForward test4+1.047%+8.28%
24 Aug 2026NIFTYForward test4+0.021%+8.30%
25 Aug 2026NIFTYForward test10−0.772%+7.53%
27 Aug 2026SENSEXForward test11+0.093%+7.62%
31 Aug 2026NIFTYForward test8−0.563%+7.06%
1 Sep 2026NIFTYLive·+1.00%+8.06%
3 Sep 2026SENSEXLive·+0.65%+8.71%
7 Sep 2026NIFTYLive·+0.70%+9.41%
8 Sep 2026NIFTYLive·+0.27%+9.68%
10 Sep 2026SENSEXLive·+0.85%+10.53%
15 Sep 2026NIFTYLive·+1.32%+11.85%
17 Sep 2026SENSEXLive·+0.65%+12.50%
21 Sep 2026NIFTYLive·+0.04%+12.54%
22 Sep 2026NIFTYLive·−1.07%+11.47%
24 Sep 2026SENSEXLive·+1.09%+12.56%

Open any August day and read the whole thing → The forward-test page carries the mark to market minute by minute for each August session, the drawdown it sat through, and the positions behind it. September is shown here as the live account’s daily result.

The number that matters

Return is loud.
Sharpe is honest.

Anyone can post a return. Take enough risk and a good month is inevitable. Sharpe asks the harder question: how much return did each unit of risk actually buy?

A Sharpe above 2 is considered excellent by institutional standards. Above 3 is rare. Apex sits at 4.86. Whale sits at 6.96 and Alpha at 6.58: the same signals, with Whale spread across more strategies at smaller size.

Sharpe is computed on daily returns, annualised ×√252, over the backtest window stated on each algo.

Bank fixed deposit
~0.3
NIFTY 50, buy & hold
~0.6
Good equity mutual fund
~1.0
Strong hedge fund
~2.0
Apex
4.86
Alpha
6.58
Whale
6.96
Cumulative return · normalised to each algo's own capital
COVID-19 CRASH 300%200%100% 202120222023202420252026 ALPHA & WHALE START Jan 2020Sep 2026
Apex, +271% over 6.7 yrs Alpha, +195% over 2.3 yrs Whale, +84% over 2.3 yrs at capped risk NIFTY 50 TRI, +107% over 6.7 yrs COVID-19 crash, 20 Feb to 18 May 2020

One time axis, one percentage scale. Each curve is a return on that algo's own capital, so heights are directly comparable. Apex runs from Jan 2020, straight through the COVID crash (shaded). Alpha and Whale begin in Apr 2024, with NIFTY and SENSEX both trading from the first month. Whale is deliberately the flattest line on the chart. That is the point of it.

The grey dashed line is the NIFTY 50 with dividends reinvested, over the same window. The three algo curves are backtested; the benchmark is actual index data. Total return is approximated by compounding a 1.3% annual dividend yield onto NSE closing levels. Since Jan 2020 the index returned +107%, including a 38% peak-to-trough fall in the COVID crash. Whale returned +84% from Apr 2024 with its drawdown held under 1.5%.

Whale

You set the drawdown.
The algo sizes itself.

Most systems hand you a return and leave the risk wherever it lands. Whale works the other way round. You name the deepest drawdown you are willing to sit through, and the position sizing is solved backwards from that.

At tighter ceilings Whale spreads the same signals across more strategies with smaller lots. At 4.0% it runs as Alpha. Risk-adjusted return holds up along the way: Sharpe is 6.96 at the 1.5% ceiling and 6.58 uncapped.

1.50% Drawdown ceiling ₹1,50,000 worst case
33.8% Return per year ₹33.79 L · ₹2.82 L/mo
1.0%2.5%3.5%4.0% uncapped

Hedged the same way at every setting, so the margin per lot does not change. At 4.0% the algo runs unrestricted, with no size reduction at all.

AI and machine learning

AI does the searching.
Rules do the trading.

Every algo on this page survived a machine that spent weeks trying to break it. Millions of versions get built, tested against years of real market data, and thrown away. What we actually trade is the small handful that held up.

The search behind one algo SENSEX sleeve selection · 2026
3,200,000configurations generated
20,000still worked on data they had never seen

Search at scale

A person tests twenty ideas in a week. A machine tests three million, against years of real prices and real costs.

IS → OOS testing

Test three million settings and one looks brilliant by chance. So nothing ships until it still works on data it has never seen.

AI systems we build

Models that learn a signal, systems that read market conditions, and the research tooling around both. We build those too.

What QuantNifty is

An algo development and consulting firm.

We are engineers, not advisors. We write trading software, test it honestly, and hand it over. The algos above are systems we built for our own capital, and the same work can be done for yours. How an engagement runs.

Strategy development

From an idea, or a rule you already trade by hand, to production code with stop-losses, re-entry logic and execution safeguards built in.

Backtesting that tells the truth

Slippage on both sides, statutory charges, real expiry calendars, walk-forward validation. We kill more ideas than we ship. That is most of the job.

Live execution systems

Broker-side stop-losses, fill confirmation, spike-trigger detection, entry repricing and a kill switch. These are the parts that decide whether a good backtest survives a live market.

Who builds this

The engineer and the trader
are the same person.

Anuj Baid, founder of QuantNifty, algorithmic trader and software engineer
Anuj Baid Founder · QuantNifty
Connect on LinkedIn

A seasoned algorithmic trader with hands-on experience in Indian index options and US SPX 0-DTE options and MNQ futures. Also a software engineer, which is why the systems get written here instead of outsourced.

India NIFTY & SENSEX options

Short-volatility and intraday expiry structures on NSE and BSE index options. Multi-year tick-level datasets, with our own execution stack running on live broker APIs.

United States SPX 0-DTE & MNQ futures

0-DTE option structures on the SPX, and MNQ futures on the Nasdaq-100. Two different indices in a session that runs while India sleeps.

The markets teach different lessons. Indian index options are dense with expiry-day flow and structural quirks; SPX is deeper and far less forgiving of sloppy execution. A strategy that only works in one session's microstructure usually isn't a strategy.

On the engineering side: data pipelines, broker API integrations, and live order-management systems with native stop-losses and failure handling. Every algo here was built in-house, along with the backtester, the execution layer and this website.

Most trading-software shops can code but have never had money at risk. Most traders have conviction but no way to ship it. QuantNifty sits in the overlap, which is a narrower place than it sounds.

Algorithmic trading Options & volatility Software engineering Quantitative backtesting Broker API automation Data engineering
Common questions

What people ask us first.

Straight answers about pledged margin, what we build, and what we are not.

How can I earn from my stock portfolio without selling it?

By pledging it. When you pledge shares or ETFs with your broker, the broker releases trading margin against their value while you keep ownership, dividends and long-term capital-gains treatment. That released margin can fund a separate options strategy, so the same capital produces two returns instead of one. Nothing is sold and nothing is borrowed.

What is pledged margin, and does it cost interest?

Pledged margin is collateral value your broker recognises against securities you already own. It is not borrowed money, so no interest is charged. Brokers apply a haircut, so ₹1 crore of holdings usually releases about ₹75 to 85 lakh of usable margin, depending on the instrument.

Is QuantNifty SEBI registered?

No. QuantNifty is not a SEBI-registered Research Analyst, Investment Adviser, Portfolio Manager or Broker. We are an algorithm development and consulting firm, so what we do is write and test trading software. We do not manage money, sell tips or signals, or give investment advice. For what registration actually attaches to, see SEBI algo trading rules.

What is a good Sharpe ratio?

The Sharpe ratio measures return per unit of risk. Roughly speaking, a bank deposit sits near 0.3, buy-and-hold NIFTY near 0.6, a good equity mutual fund near 1.0, and a strong hedge fund near 2.0. Anything above 3 is rare. QuantNifty's backtested algos sit between 4.86 and 6.96 over their stated windows.

What does 0–1 DTE mean?

DTE means days to expiry. A 0-DTE trade is opened and closed on the contract's expiry day, and 1 DTE is the session before it. Both are intraday, so nothing is carried overnight and there is no gap risk between sessions.

Are the returns on this page live or backtested?

Backtested. Every figure is simulated on historical 1-minute data over the window stated on each algo, net of 0.2% slippage per fill, the cost of the protective hedge, and all statutory charges. A backtest is not a live track record. Live results will differ because of execution, liquidity, latency and changing market regimes.

Do you use AI or machine learning in your trading systems?

Yes, but in the research, not in the trading. Millions of versions of a strategy are generated and tested automatically against years of real market data, then filtered down to the few that still work on data they have never seen. The algos that go live are fixed rules: every entry, stop and exit is written in plain code you can read, so no model decides on its own when money moves. We also build learning systems for clients, including models that predict a signal and systems that detect which market conditions you are in.

How do you stop an AI search from overfitting?

Test three million settings and one will look brilliant purely by chance, the same way someone in a big enough crowd flips ten heads in a row. So we check whether the winners on old data are still winners on data they have never seen. On our SENSEX search across 3.2 million versions, they were not: finishing first on the old data told us nothing about what happened next. So the basket was not picked by taking the best score. It was picked from the setups that kept appearing across thousands of good results, because a pattern that repeats is much harder to fake than one lucky number. Every test also charges realistic slippage on both sides of a trade plus full statutory costs, so nothing looks good only because the costs were left out.

What does an algo development engagement actually include?

Turning a trading idea into production software: entry and exit logic, stop-losses, re-entry rules, position sizing, and honest backtesting with slippage, statutory charges and real expiry calendars. Where it is needed, live execution as well, which means broker-side stop-losses, fill confirmation, order repricing, spike-trigger detection and a kill switch.

What capital do these strategies need?

The figures on this page use ₹1 crore of deployed margin because it makes the arithmetic readable. The strategies themselves are sized in lots, so they scale down. A single NIFTY sleeve needs roughly ₹1.6 lakh of hedged margin. What matters more than size is that the drawdown is set to something you can actually sit through.

Which markets and instruments do you work with?

Indian index derivatives: NIFTY 50 options on the NSE and SENSEX options on the BSE, traded intraday on 0 and 1 DTE contracts with a protective hedge on every position. We do not publish the entry and exit logic of our own systems. On client work the rules come from you and stay yours.

Get in touch

Tell us what you want built.

We will tell you honestly whether the idea is worth building before quoting anything.

This opens WhatsApp with your details already filled in, so nothing sits waiting in an inbox.

What we do not do

We do not manage anyone's money, sell tips or signals, guarantee returns, or give investment advice. We are not SEBI registered. We write trading software to your specification, and you own the strategy decision along with the risk that comes with it.