Ansari Information

Six per cent. That is roughly what an app's digital gold gives up between buying and selling, before the gold price moves a paisa. A gold ETF charges no toll at the door; it skims a small fee every year instead. So the digital gold vs gold ETF question is one sum: how many years of that fee add up to the toll?

Is digital gold safe after the SEBI warning?

Not in the regulated sense: SEBI said on 8 November 2025 that digital gold is neither a security nor a commodity derivative, so none of the securities market's investor protections apply if the platform fails.

That is the reading of SEBI press release 70/2025, as Moneylife reported its wording, and it steers buyers to gold ETFs. If the app or its vault partner folds, you have no SEBI complaint route and no investor-protection fund. Whether vault and trustee arrangements hold when an app shuts is the open question; our view is that an untested promise is worth less than a regulator's rulebook.

Price is the other half. Bajaj Finserv AMC's explainer, updated March 2026, lists GST at 3% on digital gold purchases, so part of every rupee buys tax, not metal. Selling costs again through the buy-sell spread, which Weekend Investing's Alok Jain told Business Today in October 2025 can reach 5%, enough to erase a good year for gold. The same report cites a FinPrint check where three platforms showed 2% to 3.6% losses on October dates when the IBJA benchmark showed a profit: short holds can lose in a rising month.

A user on a payments app's community forum wrote in April 2024 that the app showed about ₹60,363 invested but offered ₹58,284 on sale. The displayed figure uses the buy price, so it is not what a sale pays.

Zerodha's Nithin Kamath, whose firm sells gold ETFs, told Business Today in November 2025 that a buyer starts ₹500 to ₹600 down on ₹10,000; our four numbers, worked from SBI and Bajaj Finserv AMC documents, test that.

The yearly fee is the one buyers miss, because the fund takes it from its own value and no bill arrives. Strictly, it comes out daily, not yearly, though over a few years the difference is rounding. Small recurring costs also decide the home loan prepayment versus SIP break-even and the post-tax gap between an FD and a debt fund.

Digital gold vs gold ETF: where the 6% goes

On ₹1 lakh held three years with gold flat, digital gold costs about three times what a gold ETF does, because its 3% GST and exit spread arrive at once while the ETF's fee accrues slowly.

Worked example: assume ₹1 lakh, a 3% sell spread, a three-year hold and gold ending where it started, so only costs show. GST takes ₹2,913; selling the rest at 3% returns ₹94,175, so the round trip costs ₹5,825. The same ₹1 lakh in SBI Gold ETF, at the 0.65% expense ratio its factsheet gives as of June 2026, keeps 0.9935 cubed of its value, about ₹98,063: a cost of ₹1,937, brokerage assumed nil. Break-even comes at ln(0.97 ÷ 1.03) ÷ ln(0.9935) years.

Find your own spread in ten seconds: divide the app's sell quote by its displayed value and subtract the result from one. Set that in the calculator. Above 3%, the ETF's lead runs longer than our base case.

Every money row and the regulator row favour the ETF; only tax argues for patience, and it changes timing, not direction.

Over five years, a gold ETF keeps about ₹4,559 more of every ₹1 lakh than digital gold sold at the top reported spread. Derived from 3% GST, Jain's 5% ceiling and SBI's fee, gold held flat.

Switching what you already hold: the tax clock decides when

For digital gold you already own, switching saves no money, because the GST is spent and the spread is due whenever you sell; what you buy by moving is SEBI's protection, and timing it is a tax decision.

Common advice runs ahead of the arithmetic here. "Sell it all now" sounds prudent after a warning, but the ETF's fee makes a switched holding slightly dearer to keep. You move for safety, not savings, much as you might pay a card's annual fee rather than chase a waiver.

How is digital gold taxed in India?

Bajaj Finserv AMC's summary of current rules taxes digital gold held over 24 months at 12.5% without indexation; a gold ETF qualifies after 12 months, and sooner sales pay your slab rate. A 30%-slab reader selling digital gold at month 18 with a ₹20,000 gain pays ₹6,000, against ₹2,500 as long-term: ₹3,500 more, before cess. By our count, a lot bought before 30 October 2024 is already long-term today.

• Each purchase has its own clock, so a monthly saver's newest grams stay short-term longest.
• An ETF needs a demat account, and Bajaj Finserv AMC notes that brokerage and demat charges may apply.

Switch. Put new gold money into a gold ETF from now on: it is cheaper for roughly nine years and sits inside SEBI's rules. Stay only where the tax clock says so, holding a digital gold lot in gain until it passes 24 months, then switch. This week, list every lot's purchase date and set a reminder for the day each turns long-term.

Five years and three months. On the sums below, that is the rooftop solar payback for a 3 kW system on a home using 300 units a month at ₹6 a unit, after the PM Surya Ghar subsidy. At ₹4 a unit the wait stretches by two and a half years. The subsidy and the sun are fixed; your bill is not.

Why does the rooftop solar payback hinge on your tariff?

Because the subsidy and the panels' output are both fixed for a given size, the only large variable left is what each unit you stop buying is worth, and that is the slab tariff on your own bill.

The subsidy is a set rupee sum. The government's PM Surya Ghar page on india.gov.in (checked 29 October 2026) pays ₹30,000 a kilowatt for the first two and caps support at 3 kW, leaving ₹18,000 for the third. Every kilowatt past three costs full price, so the advice to fill the roof while the subsidy lasts is backwards.

Output is fixed too. BSES Delhi's Rooftop Solar FAQ (September 2025) puts each kilowatt at 3 to 4 units a day, and nothing under cloud cover, so 3 kW roughly matches a 300-unit home. It's the units-first logic we used to price an AC by kWh, not its star label.

Two dates sit on top of the sum. The scheme runs till FY 2026-27, so a system commissioned after 31 March 2027 may get nothing; our read is not to plan around an extension nobody has announced. Its collateral-free loan costs repo plus 50 basis points: 5.75% when MNRE minister of state Shripad Naik replied in the Rajya Sabha, as SolarQuarter reported on 29 July 2026. The RBI lifted repo to 5.50% on 7 October 2026, adding about ₹15 a month to a ten-year instalment (we ran the same hike against a home loan prepayment versus a SIP). Four numbers carry the decision.

So borrow the post-subsidy cost at the scheme rate and the instalment is smaller than the electricity you stop buying: the system pays its way from the first bill, once the subsidy lands.

How many years does a 3 kW system take to pay back?

A 3 kW PM Surya Ghar system takes about five years to pay back for a 300-unit home at ₹6 a unit, closer to eight years at ₹4 and about four at ₹8.

Worked example: a 3 kW system on a home using 300 units a month at ₹6 a unit. Assumptions: BSES's estimate of about ₹65,000 a kilowatt, 3.5 units a kilowatt a day (the middle of its range) and surplus valued at an assumed ₹4 a unit. Gross cost is ₹1,95,000; less ₹78,000 of subsidy leaves ₹1,17,000. The panels make 315 units a month: 300 replace units worth ₹21,600 a year, and 15 exported are worth ₹720. That is ₹22,320 a year, and ₹1,17,000 divided by ₹22,320 is 5.2 years.

The grid runs that sum for a 300-unit home. Your one-minute check: take the top-slab rate from your last bill, find your size down the side and read across to years to break even.

The bold cell is the worked example. A 2 kW system pays back sooner but leaves 90 units a month on your bill. The third kilowatt, about ₹47,000 after its subsidy share, earns roughly ₹7,200 a year: about 6.5 years. The fourth gets no subsidy and exports nearly everything, so it needs close to 13 years.

What happens to extra solar units under net metering?

Under Delhi's net-metering rules, extra units carry forward from month to month, and any surplus left at the year's end is paid at the regulator's average power purchase cost (APPC), not at your retail tariff.

BSES says Delhi's regulator sets APPC and may revise it yearly; your state may differ. That is the fourth-kilowatt problem: you sell wholesale and buy retail.

Headlines need the same care. A Rajya Sabha written reply, reported by SolarQuarter, counted 3,972,447 rooftop systems installed by 22 July 2026 and 1.89 million households reporting zero bills, but only in months when generation met most of their needs. Model twelve months of partial savings; the zero-bill figure is the wrong yardstick.

BSES gives panels 25 years and inverters 10 to 12, and subsidised systems typically carry a five-year system warranty, so later repairs are yours. On the worked example's terms, payback reaches ten years only near ₹3 a unit; below that, a new inverter arrives first. State money pulls the other way: Delhi adds ₹10,000 a kilowatt up to ₹30,000, cutting the example to about 3.9 years. If the cash would otherwise sit in a deposit, check the post-tax gap between an FD and a debt fund, as we did for the payback years on a mini split.

• Subsidy timing: carry the full cost until the subsidy lands.
• Oversized quotes: sized to your roof, not your bills.
• Monsoon months: output stops under cloud, so bills won't hit zero.
• Warranty gap: ask who pays for the inverter after the warranty.

Our call: switch if your top slab costs ₹5 a unit or more, and stop at 3 kW unless you use well above 300 units a month. Stay on the grid tariff nearer ₹3, because the inverter will need replacing before the panels pay you back. This is wrong if your state adds a top-up or pays more for exported units, so this week run your last twelve bills through the calculator above and check your DISCOM's subsidy status on the national PM Surya Ghar portal before asking for a quote.

$329. That's what Apple's repair estimator charged for a new front screen on an iPhone 17 Pro, as 9to5Mac reported on 12 September 2025, and the whole question turns on it. So is AppleCare worth it? Add two years of AppleCare+ with Theft and Loss to the fee Apple charges per crack, and set that beside the repair bill. One cracked screen is roughly where the lines cross.

Is AppleCare worth it if you crack one screen?

Yes, but barely: on yearly billing an iPhone 17 Pro owner who cracks one front screen in two years comes out about $20 ahead, and anyone who cracks nothing has paid $279.98 for reassurance.

Apple Legal's AppleCare fees page, as of October 2026, charges $29 for outer display or back glass damage, and a separate fee applies to each repair. Apple's New York price disclosure for AppleCare+ with Theft and Loss, dated September 2025, lists the iPhone 17 Pro at $13.99 a month, $139.99 a year or $269 for a 24-month fixed term. Same phone, same crack, three different answers.

Worked example: assume an iPhone 17 Pro, AppleCare+ with Theft and Loss billed yearly, 24 months of ownership and one front-screen crack. Premiums are 2 × $139.99 = $279.98. Add the crack fee: $308.98. Without the plan you pay the $329 repair, so the plan wins by $20.02. Billed monthly, premiums become 24 × $13.99 = $335.76; with the fee that's $364.76, and the plan now loses by $35.76.

Twenty dollars is a thin margin for peace of mind. We'd judge it like the payback years on swapping a window AC for a mini split: it earns its place only if payback lands inside the time you own it. Monthly billing is the trap of paying a credit card annual fee to chase its waiver: a cost sliced small stops looking like one.

Four numbers frame every case below. Apple's AppleCare page, read in October 2026, says most Apple products get one year of hardware coverage under the limited warranty, and that plans include unlimited incidents of accidental damage, each with its fee. The fourth card is our own arithmetic.

The free first year matters: any defect then is Apple's problem anyway, so year-one premiums buy only accident and theft cover.

That odds figure is one division: $279.98 of premiums over the $300 a crack saves after the fee.

What two years of cover cost, crack by crack

Over 24 months on an iPhone 17 Pro, AppleCare+ only finishes ahead when a front screen breaks, and how you pay moves the result by more than the plan's entire margin on one crack.

Find your row by what breaks in two years, then read across at how you pay. Each cell is what the plan leaves you ahead (+) or behind (−) against Apple's reported repair prices, including $159 for back glass.

The shaded cell, one front-screen crack on yearly billing, is the base case at +$20. One column left, monthly, flips it to −$36; one row up, back glass only, drops it to −$150. Now the one-minute check: divide your plan's two-year cost by $300 and compare it with your honest odds of cracking a front screen. Count the screens you broke on your last two phones first.

Where the one-crack rule bends

The one-crack rule bends when the phone is lost outright, which the plan covers far more generously than a crack, and when several Apple devices could share one AppleCare One bill.

Apple Legal sets the theft and loss deductible at $149. A new iPhone 17 Pro started at $1,099 in the US, per Apple's September 2025 launch release, so one lost phone saves about $950. On our arithmetic, yearly premiums break even on loss alone if your odds of losing it within two years top roughly 29%. That bar sits far below the crack bar, and we think most buyers undervalue it.

We found no current, method-published US figure for how often owners crack a front screen, so your odds are yours to estimate (the honest grey area here). Weigh them like renewing versus porting a health insurance policy: by what you claimed, not what you fear.

Is AppleCare One cheaper than AppleCare+?

For one iPhone alone, no: AppleCare One Individual costs $19.99 a month for up to three products, so it beats a $13.99 iPhone plan only once a second or third device rides on it.

At $239.88 a year it costs $99.89 more than the yearly iPhone plan, so the extra devices must be ones you'd otherwise insure.

• Batteries: Apple's battery service page says the plan replaces it free once capacity drops below 80%. The reported $119 out-of-warranty price counts only if yours dips under that line within two years.
• Monthly billing: 24 payments cost $55.78 more than two yearly ones.
• Pro Max: its reported $379 screen puts yearly AppleCare+ $70.02 ahead after one crack.

Our call: stay uninsured on an iPhone 17 Pro if you can't honestly see a front-screen crack coming within two years, and bank the premium yourself. Switch to AppleCare+, billed yearly, if you can, or if losing a phone is a real risk. This is wrong if Apple has cut its screen price since September 2025, so this week, before you buy, open Apple's repair estimator for your exact model and confirm the screen still costs more than your two-year plan plus the $29 fee.

A window unit cooling one room through a 1,000-hour summer burns about $202 of electricity. A mini split doing the same job burns about $110. That $92 gap is the entire case in the mini split vs window ac argument, and it has to repay an install bill of thousands. Here is the sum, line by line, so you can swap in your own numbers.

Is a mini split worth it over a window unit?

No, not on cooling savings alone: at the average US residential price of 18.31 cents per kilowatt-hour, a mini split's electricity saving is worth under $8 a month averaged over a year.

That price is the July 2026 residential average in the Energy Information Administration's Electric Power Monthly (released 24 September 2026), up 4.9% on a year earlier. Rising prices favor the mini split, but 4.9% of a $92 saving is under $5 a year, so the trend will not rescue the payback.

Three assumptions, all mine: a 12,000 BTU/h (one-ton) unit running 1,000 full-load hours a summer, and a $3,000 install, inside the $2,000 to $6,000 Filterbuy's 2026 cost guide gives for one room (Carrier quotes $2,000 to $10,500 for 1,000 square feet).

• Window unit: the Department of Energy's floor for a louvered unit this size, built since 2014, is 10.9 CEER, a combined efficiency score in BTU of cooling per watt-hour. 12,000 ÷ 10.9 is about 1,101 watts, so 1,101 kWh over the summer, times 18.31 cents, is $202.
• Mini split: assume 20 SEER2, a seasonal score of the same kind. 12,000 ÷ 20 is 600 watts, so 600 kWh, which is $110. Carrier's single-zone line lists top ratings from 22.4 to 35.1 SEER2, so 20 is not a stretch.
• The gap: 501 kWh, or $92 a summer.

One caution: SEER2 averages a mild season while CEER is a single hot-day test, so the comparison tilts a little toward the mini split (I would rather say so than flatter the upgrade). We ran the same kWh-first method in rupees in our piece on pricing an AC by kWh instead of stars.

Four numbers decide the swap, and read together they turn a $92 headline into a decision.

Take the yearly saving a ten-year payback demands. At the 501-watt gap between the two machines, reaching it would take about 3,270 full-load hours in a summer, which is 136 days of nonstop cooling. That is my arithmetic, not a published figure, and no ordinary room comes close.

A certain saving set against an upfront cost is a familiar shape: see a home loan prepayment against a SIP at the break-even rate.

Mini split vs window ac: the decision table

Stay with the window unit if it still works, and switch only when one of four conditions applies; the table prices each side so you can see which way every row pushes.

Read the left column first. A tick favors the swap, a warning sign means it depends on your home, and a cross counts against it.

Only one row clearly favors the mini split, and it is about running it, not buying it.

What breaks the payback sum?

Cheaper power or fewer cooling hours stretch payback toward half a century, while a dead window unit or a winter heating need shrinks the gap the install has to repay.

At 12 cents a kWh, an assumption of mine, the $3,000 takes about 50 years; at 25 cents, about 24. Buying down does not help: the federal floor for heat pumps since 2023 is 14.3 SEER2, per Carrier's summary of the Department of Energy rule. A unit at that floor saves only $48 a summer against the window unit, so payback passes 60 years.

Tax credits used to bend this sum. The IRS page for the Energy Efficient Home Improvement Credit (last reviewed 28 April 2026) allows up to $2,000 a year for heat pumps, but only for property placed in service through 31 December 2025, so a 2026 install gets nothing. Whether your utility or state still pays a rebate is the open question, and my view is that a rebate you cannot see in writing is not a number.

• Get quotes with any rebate printed on them.
• Read the rated SEER2 on the exact model; the 20 here is an assumption.
• Count your real cooling hours: a room cooled only on summer evenings sits far below 1,000.
• If the window unit is dying, price its replacement and subtract it, because only the difference needs repaying, the same sunk-cost move as in paying a card annual fee instead of chasing its waiver.

Stay with the window unit. Unless it is dying, the room needs winter heat, your building bans window units, or you cool around the clock for months, the swap is a 30-year bet at best. This week, pull last summer's electric bills, estimate your cooling hours honestly, and re-run the sum above with your own price per kWh before anyone hands you a quote.

Related: rooftop solar payback after the PM Surya Ghar subsidy, worked at your own tariff