Ritvik Dashora, CFA

"India's gold demand falls 29%." That's a kilos number. In rupees, Indians spent 3% more on gold in Jul-Sep than a year ago. (World Gold Council, India Q3 2026) We didn't stop buying gold. We bought smaller gold. It's the petrol pump logic. Petrol gets costly, you still say "₹500 ka daal do bhaiya" and get fewer litres. Nobody says India stopped driving. Same family budget. Thinner bangle. Whenever you read "demand fell", ask one question: in kilos or in rupees? A jeweller's sales, the import bill and your shopping list tell very different stories depending on the answer. The festive numbers come out in February. Did your Dhanteras gold get lighter this year, or did the budget go up?

"What if @zerodha builds this?" Every fintech founder in India gets asked this by VCs. Fair question. The big player has the users, the money and the engineers. But a food court adding biryani to its menu doesn't kill the biryani shop down the road. For the food court it's item no. 47. For the shop it's the only dish. They make it every day, for people who came only for that. Big apps build what helps most users a little. Startups build what helps a few users a lot, and go deeper than a big company can justify. A feature in their app is one more tab. For a startup, it's the whole company. The real risk isn't that they copy you. It's that you stay shallow enough to be copied in a quarter.

We cleaned every corner of the house for Diwali. Nobody cleaned the demat. Most portfolios have an attic. The small position from a 2021 WhatsApp tip. The stock you bought because a colleague did. The one where you open the app and go... why do I own this? None of these are bad by default. Some might be your best holdings. The problem is you don't remember why they're there. A stock you can't explain is clutter with a ticker. Post-Diwali safai, 3 questions per holding: 1. Do I remember why I bought this? 2. Would I buy it today with fresh money? 3. If it vanished tomorrow, would I notice? No answer to all three? That's not a view. That's a leftover. Oldest "why do I own this?" in your demat? No names, just the story.

Hiring on a 90-day notice period is like booking a cab that shows up in 3 months... And the driver can cancel any time. In those 90 days: - the candidate holds your offer - their company makes a counter-offer - 2 more recruiters call The founder waits at the roadside, refreshing the app. Negotiating is fair. Everyone should... But 90 days isn't loyalty. The old company gets 3 months from someone who has mentally left. The new one plans around a date that may never come. The person sits in limbo for a quarter. 30 days plus a buyout option seems fairer to all three. Your opinion... keep 90 or move to 30? What do you think?

Muhurat trading is a shagun, not a forecast. This Sunday (8 Nov), the market opens for about 70 minutes on Diwali. A normal day is 375. Lots of us will buy one share for luck. Nice tradition. Then the headlines start: "Muhurat closed green, the year looks good." Or red, "bad omen." Think about what that hour is: → A short session where prices can move on very little trading → Mostly small token buys made for luck, not analysis → Zero new information the market didn't have on Friday Reading the next 12 months from it is like judging an IPL season by the toss. Do the puja. Buy your shagun share if you like. Just don't change a long-term plan because of one festive hour. What's your ritual: token buy, watch, or skip? Not investment advice.

If you cut a pizza into 8 slices instead of 4, you didn't lose half the pizza. Raw stock data says you did. That's the first trap anyone building AI on Indian market data falls into. A 1:1 bonus or a stock split gives you twice the shares at half the price. Nothing changed. But a model reading raw prices sees a 50% crash overnight, and learns from it. Two more traps like it: → Symbol or name changes: one company, two price histories. The model treats them as strangers. → Price bands: many Indian stocks can only move a set % in a day. A flat chart can be a stock that hit its limit, not a calm one. None of this is hard to fix. The hard part is knowing it exists before your model finds it for you. In finance AI, the model is the easy part. Cleaning what it eats is the job.

For every 100 SIPs started in September, about 79 ended (@amfiindia "stoppage ratio"). Every month, that number gets a scary headline. But it can't tell you who actually quit. "Ended" lumps three people together: → the one who quit after a bad month → the one whose SIP simply finished its full term → the one who switched funds (one stop + one start) It also counts SIPs, not people. One investor with several SIPs shows up several times. Meanwhile, SIP money hit a record ₹32,490 Cr in September. A gym's exit door counts the dropout, the guy with the medal, and the guy who moved to a closer gym the same way. If you've stopped a SIP, which one were you?

Nifty closed October at a 6-month high, but most portfolios didn't feel it. Of 2,274 NSE stocks listed all month, only 406 ended higher (Fewer than 1 in 5) Imagine an index is a class photo. The tall kids stand in front. In October, the tall kids were a few big IT stocks. Smallcaps and midcaps, the back rows, had a losing month. Even inside the Nifty, fewer than half the 50 stocks rose. The index isn't lying. Big companies are supposed to count more. But "the market went up" and "my stocks went up" are two different sentences.

Employer Diwali gifts are now tax-free up to ₹15,000 a year (was ₹5,000). Cross it by a rupee and, as most CAs read it, the whole gift is taxed. A ₹16K hamper can be worth less than a ₹15K one.

Option buying feels like the safe trade. Your loss is capped at the premium you paid. That's true. For one ticket. This week, @Nithin0dha shared a rough Zerodha estimate: about half of futures traders are profitable. Option buyers? About 1 in 10. The cheap ticket is the problem. A small loss feels affordable, so it gets bought again next week. Weekly expiries make it easy. Zero-day options make it daily. It's like a subscription you forgot to cancel. You never feel the charge, only the statement. Not saying option buying is gambling. Saying the cap is on the ticket, not on the habit. Option traders: do you track total premium paid per month, or only P&L per trade?

Most of us do our trading homework at 8 PM in pyjamas. Then show up at 9:15 AM in office clothes. 😂 @tradomateOne usage shows two clear peaks: around 8 PM and 9 AM. Night is for research. Morning is for execution. Screen stocks. Test ideas. Make a plan. Then let the 9 PM you guide the 9:15 AM you. The goal isn't more screen time. It's better decisions. When do you do most of your trading research?

Didn't get Jio shares? That wasn't bad luck. It was a lottery working exactly as designed, and most people misread how it works. When the retail part of an IPO gets more applications than it has shares, SEBI's rules try to give one minimum lot to as many people as possible. A computer draw picks who. So one PAN = one slip in the bowl. One lot or the full ₹2 lakh, same slip. A bigger slip doesn't help in a lucky draw. Second thing: retail was booked about 2.7x. That's not a 1-in-2.7 chance. The 2.7x counts shares, not people. Your odds depend on how many people applied, not how much they asked for. So if you got nothing, it says nothing about you or the stock. Lotteries don't reward conviction. Did you get an allotment, or are you in the "better luck next IPO" club? Not investment advice.

I just crossed 10,000 followers on @LinkedIn, and the lesson is a bit embarrassing. My policy explainers (RBI, Budget, inflation) averaged about 570 views. A post about the VC line "let us know once you have a lead" did 8,000+. The explainers were fine. They were just the 1,001st copy of the same take. On RBI day everyone posts the same thing, and nobody remembers who said it. What travelled were things only I could write: a VC's polite no, a candidate who rejected our rejection. My filter now: if a 1000 people could post it today, find the angle only you have, or skip it. Analysis still matters. It just needs a person in it. What's the post you were sure would flop that did the best?

Jio's IPO is heavily oversubscribed, and everyone is quoting it like a rating. It isn't. It's the length of the queue outside a restaurant. It tells you how many people want a table tonight, not whether the food is good or the bill is fair. A subscription number is demand at a price the company chose. That's all it measures. Two things the queue hides: → Queues grow because they're long. "Everyone's applying" becomes the reason to apply. → It's not one crowd. By day 2, big institutions had bid 9.9x their share. Retail was only just past full. And for a small investor, a longer line doesn't improve the meal. It lowers your odds of getting a table. Oversubscribed retail allotment is a lottery. Read the menu (the prospectus), not the line. I'm not saying apply or skip. That's your call... I'm saying the queue is the easiest number to see... and the least useful one to decide with. Not investment advice.

"Let us know once you have a lead" is the politest no in Indian VC. Fundraising works like a sangeet. The music's on, everyone's tapping their feet, and the floor stays empty until one chacha walks out and dances first. Two minutes later it's packed. Most investors want to be guest number two. The lead does the homework, sets the price, takes the board seat and takes the blame. Followers borrow someone else's conviction for free. So founders... stop counting "interested" investors. Count the ones who'd dance first. What I'd do: treat "we'll follow" as a no until someone leads. Stay warm, but don't count them. Put your time into the one investor who can get conviction on the thing you understand better than anyone. Once the first "chacha" is on the floor, the rest show up fast.

SEBI just shut the front door on gold funds. The side door is still open, and that's where people overpay. Since 16 Oct, most fund houses stopped new money into gold FoFs (funds that buy gold ETFs for you, no demat needed). Running SIPs continue. New ones don't. The reason isn't gold. It's the rupee. Gold is one of India's biggest imports, and every rupee of gold buying needs dollars. Gold ETFs on the exchange are still open. So the crowd moves to the side window. When the queue gets long, the ETF can trade above the value of the gold it holds (its iNAV). It's happened before. Last October, when silver FoFs shut, most silver ETFs traded 5–10% above their iNAV. Two habits if you're buying this week: check price vs iNAV, and use a limit order. And if your gold FoF SIP is already running, don't cancel it to restart it. A new registration won't go through. Smart rupee defence, or just a longer queue at the side door? Not investment advice.

Most startups don't have a growth problem. They have a leaky bucket and a very expensive hose. Founders keep asking how to grow faster. More ads, more influencers, more sales hires. The first question should be: why do people leave? If users try you once and never come back, every new user you buy is water poured into a bucket with a hole in it. You can pour faster. You'll just make a more expensive puddle. Two questions I ask before any growth push: 1. Would our users be upset if we shut down tomorrow? 2. Do they come back without us reminding them? If the answer is no, the fix is the product, not the marketing budget. Plug the hole. Then turn on the hose.

Sensex jumped 1,100+ points today because the Finance Minister said the government is "open to hearing views" on LTCG and STT. Not a cut. Not a date. A "we'll listen." #LTCG is the tax on profit from shares held over a year. #STT is the small tax you pay on every trade, win or lose. Traders have wanted relief on both since they went up in 2024. IT stocks were already up on results, so not all of today was the FM. But a big part of the move came after the headline. It's the kid who hears "we'll see" and starts picking the ice cream flavour. The problem: a rally built on a "maybe" can reverse on a "no." The #Budget2027 is months away. Every comment and leak until then will move prices both ways. What was said and what was decided are two different things. Today only one of them happened. Honest question: if STT were cut tomorrow, would you trade any differently?