The Concept Map

8 min read

One order. Two markets. The hidden machinery behind every trade.

4 min read

A market price is not discovered by magic. It is negotiated in public.

4 min read

A trade is not done when two people agree on price.

8 min read

Your phone said "Payment Successful." The two banks hadn't settled yet.

6 min read

HR knows you by your legal name. Payroll knows you by an employee code. The badge reader knows you by neither. The moment those three stop agreeing, paychecks and badges both quietly break.

5 min read

Nobody sees the machines that turn a click into a trade.

7 min read

Finance decides if you're allowed to travel. The travel agent decides which flight you get. A trading desk runs the identical split — with a lot more money on the line.

8 min read

Every trade executed notification is one sentence of code deciding whose order went first.

7 min read

Two people can watch the identical match and see two different games — one live, one delayed. Markets run the same trick, legally, all day.

6 min read

A racetrack doesn't just check your license at the gate. A marshal watches every lap, ready to black-flag you the instant your car starts swerving.

7 min read

A shipping label tells everyone what to move and where. It doesn't tell you who unpacks the box, stores it, or handles the recall notice six months later.

7 min read

A landlord raising the rent doesn't need your reply. A landlord offering a discount does — miss that deadline and nothing happens to you at all. You just quietly lose an option you'll never see again.

11 min read

The same event can happen to an order twice in a row — and the second time can mean something completely different.

19 min read

What FX, Equities, Rates, Credit, Commodities and Prime Brokerage actually do.

6 min read

Same stock. Different game.

5 min read

TWAP ignores the market's rhythm. VWAP follows it.

5 min read

VWAP asks if you traded in line with the market. Arrival price asks a sharper question.

5 min read

The gap between an imaginary instant trade and the real one — accounts for every algorithm in this series.

10 min read

VWAP and arrival price grade one trade. TCA is built to grade the broker.

5 min read

A fund needs to buy ten million shares — TWAP refuses to rush.

5 min read

TWAP follows the clock. VWAP follows a forecast. POV follows the market itself, live.

5 min read

An order book shows what's for sale and how much. An iceberg order breaks that last promise on purpose.

5 min read

TWAP, VWAP and POV trade continuously. A sniper does nothing — until one condition appears.

5 min read

The fake order is the bait. The real trade happens elsewhere.

5 min read

A futures gain isn't paper profit. It was already paid to you yesterday.

6 min read

A futures contract with the exchange taken out — same bet, no safety net.

5 min read

One side can walk away from a bad price. The other side can't.

5 min read

Ten crore changes nothing hands. Two and a half lakh does.

5 min read

Nothing happened to your bond. The number that changed was somebody else's.

5 min read

Same rate move. Twelve times the damage.

5 min read

Lending for ten years paid less than lending for two. That's not supposed to happen.

6 min read

Duration predicted a ₹124 loss. The bond lost ₹113.

9 min read

By the time a central bank cuts, the market has usually finished trading the cut months earlier.

6 min read

Same day. Same maturity. A 230 basis point gap that had nothing to do with rates.

8 min read

The bond never changes hands. The default risk does — and so can anyone's opinion of it.

9 min read

Two banks price the same dollar-for-rupee swap off the same numbers — and quote two different rates anyway.

8 min read

A stock has never traded below zero. On 20 April 2020, a barrel of oil did — in the US, then in India, six hours later, for a different reason entirely.

7 min read

Same loans. Same pool. One slice rated AAA, one slice rated junk — paid from exactly the same cash.

8 min read

It looks like borrowing. Legally, ownership flips twice.

8 min read

One order fills. Then someone decides who actually gets it.

4 min read

Two strangers trade. One institution becomes the promise.

6 min read

A clearinghouse turns future fear into collateral today.

7 min read

The hedge and the danger turned out to be the same thing.

7 min read

A flight recorder doesn't fly the plane. It just remembers everything, so that if something goes wrong, someone can find out exactly what happened.

7 min read

The market does not settle every promise. It compresses the mess.

7 min read

At some point, a trade must stop being reversible.

12 min read

A trade can execute perfectly and still not actually happen.

7 min read

Not every share that changes hands was ever bought or sold.

8 min read

A clearing member fails. The waterfall gives the loss a pre-written path.

7 min read

A bank doesn't fail alone. It fails into every balance sheet that was counting on it.

8 min read

Sort a toolbox by what each tool is, or by what you do with it — both are correct. FICC and Equities sort a trading floor the first way. GECD is proof a bank can sort it the second.

13 min read

When Lehman Brothers collapsed in 2008, regulators weren't unwinding one company. They were unwinding thousands of separately incorporated ones, scattered across dozens of countries, with no consistent way to tell which one owed what to whom.

13 min read

Lehman's collapse in 2008 froze hedge fund assets their own prime broker had the legal right to re-lend. The industry's fix — spread your business across several prime brokers — is exactly what let Archegos hide $20 billion of leverage from all of them in 2021.

11 min read

Bill Hwang put down $1 of his own money and controlled roughly $20 of stock — without signing a single loan document. No promissory note, no pledge agreement, nothing that used the word 'loan.' He signed a swap instead.

11 min read

Credit Suisse's own risk committee flagged Archegos as a concern in 2020 — a full year before the $5.5 billion loss. The bank had the function whose entire job was to catch exactly this. It caught it. Nothing changed.

9 min read

Morgan Stanley sold $5 billion of Archegos-linked stock in one quiet Thursday night. Goldman Sachs sold $10.5 billion the next morning, publicly. Credit Suisse, still negotiating the same trade on the same stocks over a week later, lost $5.5 billion.

10 min read

Bill Hwang built a stake in ViacomCBS large enough to move its stock price on exit. He never once had to file the disclosure a direct shareholder that size would have owed the market — because he never technically owned a single share.

7 min read

The VIX has closed above 80 exactly twice in its history. Both times, stocks were crashing. Neither time were they rallying.

11 min read

The fastest way to stop a market from panicking is to stop the market.

7 min read

For thirty days after a company goes public, the bank that sold its stock to the world can be secretly short it — by design, with regulators' blessing, named after a shoe company that has nothing to do with finance.

7 min read

93% of Indians who traded options for three straight years lost money doing it, on average ₹2 lakh each. What they were losing to wasn't a bookmaker. It was time itself.

7 min read

In FY24, India's prop desks and FPIs made ₹61,000 crore in the same market where individual traders lost the most — 96% of it from algorithms doing one thing all day: rebalancing delta.

8 min read

GameStop rose 2,321% in eleven trading days in 2021, blamed on a 'gamma squeeze' the SEC's own investigation couldn't fully confirm — the same expiry-day sensitivity SEBI says Jane Street engineered on purpose in India, for ₹4,843 crore.

8 min read

India's VIX jumped 51% in one session on election counting day 2024. Six years earlier, a $1.9 billion fund betting the opposite way shrank to $63 million in an afternoon. Same mechanic, both times: vega.

7 min read

The Fed's fastest hiking cycle in 40 years and the RBI's own climb to 6.5% both changed what a stock option was worth on paper — with the stock price never entering into it. That's rho, the Greek almost nobody on a weekly-options desk has ever had a reason to check.

6 min read

In January 2021, 140% of GameStop's public float had been sold short. Not a typo — more of the company's stock had been shorted than existed.

6 min read

You paid zero commission on that US trade. Somebody still paid for it — just not you, and not anywhere the confirmation screen shows.

8 min read

Two clients want opposite things. Does their broker send that to a market — or just settle it in-house?

8 min read

A US wholesaler internalizing your order never has to show its price. Past a threshold in Europe, the law makes it.

3 min read

Judging a decision by its outcome teaches you the wrong lesson from both wins and losses.

2 min read

A test that's 99% accurate can still be wrong more often than it's right.

3 min read

The real cost of a choice is the single best thing you didn't do instead.

3 min read

Continuing because of how much you've spent is a math error dressed up as loyalty.

3 min read

Tell someone to do the right thing and they might. Change what's rewarded, and they will.

6 min read

A forgotten $12 charge — businesses that were never trying to convince you twice.

3 min read

From steam engines to AI, every revolution begins by copying the world it eventually replaces.

4 min read

It doesn't see 'strawberry.' It sees two number IDs that used to be a word.

3 min read

It's a token budget — once it's full, the oldest content doesn't fade, it stops existing.

3 min read

Confidence is not the same as truth.

4 min read

The most valuable output of AI isn't what it creates — it's what it teaches your organization.

13 min read

The fraud wasn't hidden from the control. The person committing it was the control.

9 min read

The unfair edge wasn't a hack. It was the wiring.

10 min read

Every other edge in a market is earned. This one is borrowed from someone who owed you nothing.

3 min read

Every unexplained acronym charges an entry fee.

4 min read

KYC, CDD, EDD, SAR — four acronyms that look interchangeable and aren't.

6 min read

Two clean diagrams walked into a steering committee. Neither one said what had to happen to turn the first into the second.

3 min read

A vague request becomes useful only when behavior becomes testable.

3 min read

A good requirement still works after the meeting ends.

3 min read

A ticket isn't a spec. The gap between idea and test is where defects are born.

3 min read

A backlog is not storage. It is a decay chamber.

3 min read

The same change request costs ten times more once code exists to contradict it.

8 min read

Ask a BA, a PM, and a PO the same question in the same meeting, and you'll get three different answers — because they're not answering the same question.

7 min read

Cancel a resting order and it vanishes. Cancel an executed trade and — it doesn't. Not really.

8 min read

A classroom vote only means something if nobody can shout, get counted, and take it back before the teacher finishes tallying. A closing price works the same way — and this is what happens when someone figures out the tally isn't actually watching for that.

8 min read

A genuine correction fixes an error nobody could have profited from knowing about in advance. A laundering correction fixes exactly that.

7 min read

A bank told institutional clients its dark pool would protect them from predatory high-frequency traders. Internally, it was courting those same traders for their fees.

7 min read

Two regulators proposed the identical fix for the same problem. One built it. One wrote a paper about it and moved on.

10 min read

On October 28, 2008, Volkswagen's stock nearly quintupled in four trading days — not because the company changed, but because short sellers had nowhere left to buy.

7 min read

The same stock trades on more than a dozen exchanges at once. US law says your broker isn't even allowed to pick the wrong one.

9 min read

Accenture traded for one cent. Not because the company was worthless — because the only price left in the book was never meant to be traded at.

7 min read

On the morning of August 24, 2015, roughly one in five US-traded ETFs fell 20% or more within minutes — some more than 40% — not because the stocks they held were actually worth that much less, but because the one person whose job was to correct the price had, for about thirty minutes, no way to do it.

10 min read

On April 23, 2020, roughly three lakh Indian investors found they couldn't withdraw a single rupee from six Franklin Templeton debt funds holding nearly ₹25,000 crore — even though every one of those funds had promised, in writing, redemption on any business day.

17 min read

The fee on your contract note and the number your bank credits its own salesperson for that same trade are computed by two completely different formulas — and you only ever see one of them.

11 min read

Nine out of ten individual traders in India's F&O market lost money last year. They didn't all lose it the same way.

8 min read

The person pitching you the stock and the person deciding whether to buy it don't work for the same side.

15 min read

Four counters under one signboard, a trade between two of them that no customer ever sees — and, once, a trade just like it that cost the world's biggest banks over $10 billion.

4 min read

An agent isn't a bigger brain answering harder questions. It's the same model, wrapped in a loop that won't let it stop after one guess.

8 min read

Protect someone from the downside of a risk, and you change how much risk they're willing to take.

4 min read

"We RAG'd our documents into the AI" makes it sound permanent. It isn't — it never touches the model at all.

5 min read

Every test before UAT asks 'does it work.' UAT is the only one that asks 'is this what the business actually needed.'

13 min read

Every other order finds its price by fighting for it in public. A block trade agrees on the price before the market even finds out.

12 min read

A flat, a loan-chit, and a rain bet — sitting under three tabs that look like the same button.

10 min read

A bond is a promise on paper. A share is a promise on paper. A barrel of oil is not a promise at all — it's an actual barrel, and someone, somewhere, has to store it.

9 min read

FICC prices the macroeconomy. Equities prices one company at a time — and that single difference changes almost everything about how the desk actually works.

10 min read

Equities prices one company at a time. FICC prices everything else that moves money.

15 min read

Every entry-level BA posting still asks for experience. Here's how to build proof instead of waiting for permission to start.

8 min read

Every order is answering the same question: how badly do you want this, right now?

4 min read

The job description says 'gather requirements.' It doesn't mention any of the parts that are actually hard.

6 min read

You deposit ₹1,000. Most money in the economy was never printed — it was lent into existence.