Penny stock addition under section 68: no role in the price rigging

An investor buys shares of a small listed company through a registered broker, pays by cheque, takes delivery in his demat account, holds them for two or three years, sells them on the exchange, pays securities transaction tax and claims the gain as exempt. Three years later an Assessing Officer, working from a report of the Investigation Wing about the scrip, treats the entire sale consideration as unexplained cash credit under section 68, adds a notional commission under section 69C, applies the rate in section 115BBE and initiates penalty. Nowhere in the assessment order does the investor’s name appear in any statement, in any SEBI order, or in any list of persons found to have manipulated anything.

That is the standard penny stock assessment, and the question this note is about is a narrow one: what happens to it when the Department cannot connect the assessee to the rigging.

The answer, on the weight of authority, is that the addition does not survive. But the field is genuinely split, the split is territorial, and the split has not been resolved by the Supreme Court. An article that recites only the favourable cases is of no use across the table from a Departmental Representative who has read Swati Bajaj. What follows sets out the contentions as the courts have actually dealt with them, maps the position court by court including the decisions that went the other way, and then takes the section 69C commission separately, because it is the part of these assessments that is least defensible and most often conceded by default.

Key points

• The initial onus under section 68 is discharged by the documentary chain: contract notes, banking trail, demat holding, sale on the exchange, STT paid. Once it is on record and not shown to be false, the onus shifts.
• What the Department must then produce is positive material connecting this assessee to the manipulation. A report about the scrip, a price chart and an operator’s statement that does not name him do not supply it.
• The Calcutta High Court in Swati Bajaj reverses that burden and treats the gain as tainted by the rigging at the other end unless the assessee disproves it. It is the leading and most fully reasoned decision on the Revenue’s side, and a line of Calcutta decisions has followed it.
• Outside Calcutta, benches follow their own jurisdictional High Court, and the Gujarat High Court has now held in terms that Swati Bajaj has no application on different facts.
• There is no Supreme Court judgment on the question. Every order is at the special leave stage, and on Kunhayammed none of them declares law.
• SEBI’s own record is the most underused evidence available. In Kailash Auto Finance, 246 entities were restrained, 244 were cleared eighteen months later, and of those 246 just two were ultimately sanctioned.
• The section 69C commission is the weakest link. Section 69C requires expenditure incurred, and the Comptroller and Auditor General has recorded that the percentage applied varies from 0.5 to 5 per cent with no consistency.

How is a penny stock assessment actually built?

The template is uniform enough to be worth setting out, because each layer has its own answer and the answers are different.

The starting point is an Investigation Wing report. The most frequently cited is the report of the Directorate of Investigation, Kolkata dated 27 April 2015, which identified 84 penny stock companies listed on the Bombay Stock Exchange, some 64,811 beneficiaries and a cash trail of nearly Rs 1,575 crore, against a headline figure of bogus long-term capital gain of about Rs 38,000 crore. Its conclusions are about scrips and about operators. It is not, and does not purport to be, an adjudication about any individual investor.

Next comes a SEBI order, almost always an ex parte ad interim order restraining a long list of entities from the securities market. These orders are made on a prima facie view, at speed, and the later history of several of them is dealt with below.

Third comes a statement, usually of an alleged entry operator, recorded under section 131 or during a survey, in proceedings to which the assessee is not a party. These statements describe the modus operandi. They frequently do not name the assessee at all, and where they have been retracted the retraction is seldom addressed.

On that material the Assessing Officer makes four moves. He denies the exemption under section 10(38). He assesses the entire sale consideration, not merely the gain, as unexplained cash credit under section 68. He adds a percentage of that consideration as unexplained expenditure under section 69C, on the footing that a commission must have been paid to arrange the entry. And he applies the rate in section 115BBE, which denies any deduction and, for years from 2017-18, any set-off of loss.

Two features of that structure are worth noticing at once, because they run through everything that follows. The first is that the assessment taxes the gross receipt while simultaneously allowing the cost of acquisition, which is internally inconsistent. The second is that it characterises the same money twice over, as a receipt the assessee could not explain and as the subject of an expenditure he must have made, which is more inconsistent still.

What does section 68 actually require?

Section 68 provides that where any sum is found credited in the books of an assessee maintained for any previous year, and the assessee offers no explanation about the nature and source of it, or the explanation offered is not, in the opinion of the Assessing Officer, satisfactory, the sum so credited may be charged to income-tax as income of that previous year.

Three elements deserve attention.

A sum found credited in books. The section operates on a credit entry in books of account maintained by the assessee. Where the assessee is an individual who keeps no books, and is not required to keep any, there is a respectable argument that section 68 has no application at all, because a bank statement or passbook is not a book of account maintained by the assessee. The Agra bench accepted exactly that in Ruby Jain v. ITO, deleting the section 68 addition and, with it, the dependent section 69C commission. The point is worth taking as a threshold objection where the facts allow, though it must be taken alongside the substantive grounds rather than instead of them, because the Department will simply shift to section 69 or section 69A.

The explanation about nature and source. In a sale of listed shares the “sum credited” is the sale consideration, and its immediate source is the clearing mechanism of a recognised stock exchange. That is a materially different enquiry from the one section 68 was framed for, which is a credit in the name of a person whose identity, creditworthiness and genuineness can be tested. The Supreme Court’s decision in PCIT v. NRA Iron and Steel Pvt Ltd (2019) 412 ITR 161, which the Department cites constantly, is a bogus share capital case, and its three-limb test of identity, creditworthiness and genuineness has no natural application where the counterparty is the exchange settlement system rather than a named creditor.

The opinion of the Assessing Officer. It is not a subjective licence. The Supreme Court said so in CIT v. P. Mohanakala (2007) 291 ITR 278, a decision that otherwise went against the assessee: the officer’s opinion that the explanation is unsatisfactory “is required to be based on proper appreciation of material and other attending circumstances available on the record”. A generic report that names a scrip but not the assessee is not proper appreciation of material about that assessee.

Who has to prove what, and when does the onus shift?

This is the hinge of the whole dispute, and the two lines of authority answer it differently.

The orthodox position is that the initial onus lies on the assessee, and is discharged by the documentary chain. Once discharged, it shifts. The Supreme Court in CIT v. Orissa Corporation (P) Ltd (1986) 159 ITR 78 put the shift plainly in a cash credit context: the assessee had given names and addresses, the Revenue knew the creditors were assessees, and

The Revenue, apart from issuing notices under section 131 at the instance of the assessee, did not pursue the matter further. The Revenue did not examine the source of income of the said alleged creditors to find out whether they were credit-worthy. There was no effort made to pursue the so called alleged creditors. In those circumstances, the assessee could not do any further.

Transposed, that is the incomplete-enquiry argument, and it succeeds regularly. The Madras High Court in Sohanraj Uttamchand recorded that the Assessing Officer, despite issuing notices under section 131 to the several purchasers of the shares, “did not carry the investigation to its logical conclusion where such notices returned unserved”. The Delhi High Court in Karuna Garg recorded that notices to the company and to the payment intermediary produced nothing and the Assessing Officer “did not take the matter any further”. The Delhi High Court in Reeshu Goel recorded that “no enquiry had been conducted and the assessee’s broker had not even been examined”.

The Calcutta position is different in kind. Swati Bajaj holds that the onus does not shift on the documents, because the documents prove process and not economic reality, and that what the assessee must establish is the genuineness of the price rise itself:

It is incorrect to argue that the assessees have been called upon to prove the negative in fact, it is the assessees duty to establish that the rise of the price of shares within a short period of time was a genuine move.

That is a materially heavier burden, and it is the reason a Calcutta assessee is in a worse position than a Delhi, Orissa, Allahabad, Punjab and Haryana or Gauhati assessee on identical facts.

Can suspicion take the place of evidence linking this assessee?

This is the central contention and the one on which most deletions rest.

The Supreme Court authority is old and it is strong. In Omar Salay Mohamed Sait v. CIT (1959) 37 ITR 151 the Court held:

On no account whatever should the Tribunal base its findings on suspicions, conjectures or surmises nor should it act on no evidence at all or on improper rejection of material and relevant evidence or partly on evidence and partly on suspicions, conjectures or surmises and if it does anything of the sort, its findings, even though on questions of fact, will be liable to be set aside by this court.

In Umacharan Shaw and Bros v. CIT (1959) 37 ITR 271 the Court held that the conclusion there was “the result of suspicion which cannot take the place of proof in these matters”.

The High Courts have applied that to these facts in terms that are worth quoting because they are what wins appeals.

The Delhi High Court in PCIT v. Krishna Devi, ITA 125/2020 decided on 15 January 2021 and reported at (2021) 431 ITR 361, recorded a price jump of 4,849 per cent and that the trade pattern did not move with the index, and still held:

The theory of human behavior and preponderance of probabilities cannot be cited as a basis to turn a blind eye to the evidence produced by the Respondent.

and

the Court has to decide an issue on the basis of evidence and proof, and not on suspicion alone.

and, decisively, that counsel for the Revenue “has not been able to point out any evidence whatsoever to allege that money changed hands between the Respondent and the broker”.

The Madras High Court in CIT v. Sohanraj Uttamchand, T.C.A. Nos. 714 and 721 of 2018 decided on 28 August 2026, gave the fullest modern statement:

suspicion, however grave, cannot be equated with proof, and cannot by itself displace the documentary evidence, which clearly establishes that: (i) the shares were purchased and sold through a recognised and regulated stock exchange; (ii) Securities Transaction Tax was paid on the transactions; (iii) the sale consideration was received through regular banking channels

The Gujarat High Court in PCIT v. Divyaben Prafulchandra Parmar, decided on 2 January 2024, put it shortly: “presumption, however strong, cannot be a substitute, nor can it take place of evidence.” In PCIT v. Vinod Premjibhai Gangani the same court held that an addition made “merely on the basis of suspicion, presumptions and probability of preponderance without any direct evidence to prove the transactions as non genuine or sham or demonstrating assessee’s involvement in any kind of manipulation, cannot be made”. In PCIT v. Ambalal Chimanlal Patel it recorded that there was no evidence of “collusion or connivance between the broker and the assessee”.

The most useful phrase in the whole corpus is from the Gujarat High Court in PCIT v. Sandipkumar Parsottambhai Patel, where the assessee was found to have “no nexus with the company, its directors or operators” and to have obtained only the incidental benefit of price rise. That is precisely the position of the investor this note is about.

At Tribunal level the Ahmedabad bench put the principle as compactly as anyone, in ITO v. Sahil Vinod Agarwal (HUF) decided on 23 September 2026: “unless any link is established, the assessee could not be held to be part of the group indulging into rigging.”

How much does the documentary trail actually prove?

The recurring chain has six elements: contract notes from a registered broker, purchase consideration paid by cheque with a matching bank debit, dematerialisation and holding in the demat account for more than twelve months, sale on the floor of a recognised exchange, securities transaction tax paid, and sale proceeds received through banking channels.

The Orissa High Court in PCIT v. Dipansu Mohapatra stated it completely:

The CIT(A) was satisfied that the purchase of liquid shares have been made through Account Payee Cheques and the shares themselves were held in Demat Account for more than 12 months and then sold through the recognized stock exchange after payment of security transaction tax.

The Punjab and Haryana High Court in PCIT v. Prem Pal Gandhi, ITA-95-2017 decided on 18 January 2018, dealt with a rise from Rs 11 to Rs 400 and held that the shares “were traded on the National Stock Exchange and the payments and receipts were routed through the bank. There was no evidence to indicate for instance that this was a closely held company and that the trading on the National Stock Exchange was manipulated in any manner.”

The Bombay High Court in PCIT v. Indravadan Jain (HUF), ITA No. 454 of 2018 decided on 12 July 2023, is the single most useful decision of its kind, because SEBI had actually found manipulation and it still did not travel to the investor. The shares of Ramkrishna Fincap Ltd had been bought at Rs 3.12 and sold at Rs 155.04. SEBI had investigated and had found the broker, M/s Basant Periwal and Co, engaged in price manipulation by synchronised trades. The Court upheld the deletion because

These shares were purchased by respondent on the floor of Stock Exchange and not from the said broker, deliveries were taken, contract notes were issued and shares were also sold on the floor of Stock Exchange.

That is the cleanest available illustration of the distinction this note is about: manipulation at one end of the market does not make every trade in the scrip a sham.

The counter-argument has to be stated. Swati Bajaj holds the chain insufficient by itself, describing the documents as self-certifying recitals that prove process rather than economic reality. The honest answer to that is not to pretend it was not said, but to observe what it leaves standing: on any view the documents are evidence, and the question becomes what the Department has put on the other side of the scale. Where the answer is a report about the scrip and nothing about the assessee, the Calcutta approach converts the assessee’s onus into a requirement to prove a negative about somebody else’s conduct, which is what the Madhya Pradesh, Gujarat, Madras and Delhi courts have declined to do.

Can a generic Investigation Wing report sustain the addition?

Not by itself, on the great weight of Tribunal and High Court authority.

The Madhya Pradesh High Court, in the batch led by PCIT v. Aniruddha Nyati decided at Indore on 30 April 2024 across thirty-three appeals, held the report unusable for three reasons: it was never furnished to the assessee, it lacked corroborative evidence, and findings about the company could not bind an individual shareholder. The same batch held that a “mere sudden increase in share prices on the stock exchange, does not ipso facto determine that shares are bogus”. The same court in PCIT v. Aditya Agrawal, decided at Jabalpur on 1 April 2026, recorded that the Assessing Officer’s findings “were based solely on the record of the Investigation Wing” with “a lack of independent inquiry”.

The Delhi High Court in PCIT v. Karuna Garg, ITA 477/2022 decided on 23 November 2022, put it in one sentence:

reliance placed on the report, without further corroboration on the basis of cogent material, does not justify his conclusion that the transaction is bogus.

The Delhi bench of the Tribunal had said the same in Meenu Goel as early as 2018, that the lower authorities had rejected the claims by relying on the report of the Investigation Wing without considering the documents. Benches at Mumbai, Nagpur, Hyderabad, Indore, Raipur, Ranchi and Visakhapatnam have repeated it since, the Nagpur bench observing in Indrakumar Ghisulal Agrawal that the Kolkata and Mumbai reports were “general in nature”.

There is a related point that is easy to miss and has succeeded on its own. The Gujarat High Court in PCIT v. Genuine Finance Pvt Ltd recorded that the investigation reports post-dated the transactions. An investor cannot be fixed with knowledge, at the time he traded, of a report written years later.

Again the counter-position must be stated plainly. Swati Bajaj holds that supply of the report is not mandatory, that the test is prejudice, that “non-furnishing of the report has in no manner prejudiced the rights of the assessees to discharge the onus cast upon them in terms of Section 68”, and, memorably, that if the report was in the public domain “nothing prevented the assessees who are ably defended by the Chartered Accountants and Advocates to download such reports and examine the same”. Where an assessee has in fact been supplied with the material particulars and has not asked for more, the Calcutta High Court has applied that against him, as it did in Zulu Merchandise. The practical lesson is to ask for the report in writing, early, and to record the refusal.

Is denial of cross-examination fatal?

This is the sharpest split in the field.

The foundation is Andaman Timber Industries v. Commissioner of Central Excise, Civil Appeal No. 4228 of 2006 decided on 2 September 2015, reported at (2015) 127 DTR 241:

not allowing the assessee to cross-examine the witnesses by the Adjudicating Authority though the statements of those witnesses were made the basis of the impugned order is a serious flaw which makes the order nullity inasmuch as it amounted to violation of principles of natural justice because of which the assessee was adversely affected.

Two qualifications should be made when citing it. It is a central excise decision, applied to income-tax by analogy, and the article that presents it as an income-tax authority invites correction. And it is conditional: it bites where the statement was “made the basis of the impugned order”.

Behind it stands Kishinchand Chellaram v. CIT (1980) 125 ITR 713, which is the closer analogy on the facts. There the Revenue relied on a bank manager’s letter never shown to the assessee. The Supreme Court held the statements were in the nature of hearsay, that no reliance could be placed on the letter since it was not shown to the assessee, and that before the authorities could rely on it “they were bound to produce it before the assessee so that the assessee could controvert the statements”. An Investigation Wing appraisal report, an operator’s statement and a SEBI order withheld from the assessee are on all fours with that undisclosed letter.

For the assessee. The Orissa High Court in Dipansu Mohapatra held that reliance on statements of “so called entry operator” recorded “in some other proceedings not connected with the Assessee”, with no opportunity to challenge them or to cross-examine, was a failure of natural justice that “went to the root of the matter”. The Supreme Court declined to interfere in the companion matter of Kuntala Mohapatra. The Gujarat High Court has repeated the point in Parasben Kasturchand Kochar (“neither statement was supplied to the assessee nor cross examination was allowed”), in Divyaben Parmar and in Vinod Premjibhai Gangani (“the statements were recorded at the back of the assessee. The assessee was deprived off to cross examine the witness”). The Madhya Pradesh High Court in Aniruddha Nyati went a step further and treated non-furnishing of the report itself as a denial of the right to examine and cross-examine.

Against the assessee. Swati Bajaj holds there is “no vested right for the assessee to cross examine the persons who have not deposed anything against the assessee”, and that formal cross-examination is not part of natural justice so long as the party has a fair opportunity to see, comment on and criticise the evidence. The Delhi High Court took the same view in Sanjay Kaul v. PCIT, ITA 203/2020 decided on 29 July 2020, holding that the right to cross-examine “is not an invariable attribute” of a quasi-judicial enquiry.

And it is live. The Bombay High Court admitted CIT v. Deepak Valji Karia, ITA No. 38 of 2024, on 8 October 2025 on the express question whether the Tribunal was perverse in quashing an assessment on procedural grounds, namely lack of cross-examination and non-supply of statements, rather than examining the substantive merits. That question has not been answered.

The fair statement of the law is therefore this. Where the addition rests solely or mainly on a statement adverse to the assessee that was never put to him, Andaman Timber applies and the ground is very strong. Where the Department can characterise the statements as being about the scrip and the operators rather than about this assessee, the Calcutta answer is available to it. The way to keep the ground alive is to demonstrate, on the assessment order itself, that the statements were the foundation.

What weight does a SEBI order carry, and what about exoneration?

This is the most underused evidence in the whole field, and the reason is that practitioners cite the SEBI order the Department relies on without tracing what happened to it afterwards.

Take Kailash Auto Finance Ltd, a scrip that appears in assessment orders across the country. The history is entirely documentary:

Date	Order	Effect
29 March 2016	Ex parte ad interim order WTM/RKA/ISD/42/2016	246 persons and entities restrained
15 June 2016 to 13 July 2017	Confirmatory orders	Restraints confirmed
21 September 2017	SEBI/WTM/MPB/EFD-DRA-I/31/2017	Directions revoked against 244 entities, SEBI recording no adverse findings on their role in the manipulation
5 December 2018	WTM/AB/EFD-1/DRA-1/04/2018-19	Only 7 noticees held to have violated securities laws

Of the 246 entities restrained on a prima facie view, the regulator itself cleared 244 within eighteen months. The final order sanctioned seven noticees, but only two of those seven were among the 246 originally restrained; the other five were brought in later, by a show cause notice of December 2017. So of the people caught by the order the Assessing Officer relies on, SEBI ultimately proceeded against two. An Assessing Officer who, in an assessment framed after September 2017, relies on the interim order of March 2016 is relying on a document the regulator has withdrawn as against all but two of those it named. That is not an argument; it is a matter of record, and the orders are public.

The pattern repeats. SEBI’s suspension of trading in Risa International Ltd was set aside by the Securities Appellate Tribunal on 30 March 2016, a fact recorded by the Madras High Court in Sohanraj Uttamchand, which also noted that “no adverse action of any kind was ever taken by SEBI, the stock exchanges or any other regulatory or investigative agency against either of the assessees”.

The Tribunal has given exoneration decisive weight. The Mumbai bench in ITO v. Sunita Chaudhary deleted additions under both section 68 and section 69C where SEBI had exonerated the assessee and removed the restraint orders. The Mumbai bench in ACIT v. Pravin Selvadia recorded that SEBI had cleared the shares. The Jaipur bench in Birendra Singh Nirbhay treated SEBI’s subsequent revocation of restrictions as supporting legitimacy. The Delhi bench in Gaurav Aggarwal, decided on 16 September 2026, made the finest distinction of the lot: SEBI had found 29.88 per cent of the transactions questionable and had cleared 70.12 per cent, and the assessee’s name appeared in no SEBI charge-sheet.

Absence from the list works the same way. The Gujarat High Court in Divyaben Parmar recorded that Sunrise Asian Ltd was not among the SEBI-identified rigged scrips, and in Sangitaben Jagdishkumar Shah that “as per the SEBI report the script VAS Infrastructure Ltd. was not blacklisted and was not termed as penny stock by the Stock Exchange Regulatory Authority”. That reasoning crossed state lines when the Bombay High Court adopted it in Komal Kumarpal Shah. At Kolkata, benches have deleted additions on Bakra Pratishtan Ltd on the footing that the scrip is not in the official list of 84 at all.

The contrary use must be acknowledged. Where SEBI has made findings about the assessee, or where an adjudication order records synchronised trading and rigging, benches have used it against him: the Ahmedabad bench did so in Hanumanprasad Lalchand Gupta, the Bangalore bench in Pradeep Kothari and the Patna bench in Bhupesh Nemani. The point is not that SEBI material always helps the assessee. It is that the SEBI file has to be read to the end.

How are Sumati Dayal and Durga Prasad More answered?

Every assessment order and every Departmental submission leads with these two, and they cannot be brushed aside. They can, however, be placed.

CIT v. Durga Prasad More (1971) 82 ITR 540 held that the taxing authorities “were not required to put on blinkers while looking at the documents produced before them” and were entitled to examine surrounding circumstances, and that courts and tribunals “have to judge the evidence before them by applying the test of human probabilities”.

Sumati Dayal v. CIT (1995) 214 ITR 801 applied that to a claim of race winnings, adopting the proposition from Durga Prasad More itself that “apparent must be considered real until it is shown that there are reasons to believe that the apparent is not the real” and that the authorities are entitled to look into the surrounding circumstances and to apply the test of human probabilities.

Four answers are available, and the strongest article uses all four.

First, what those cases were about. Sumati Dayal concerned an assessee whose own account of winning on thirteen occasions in one year, ten of them jackpots, and twice more in the next, with no record of race-course expenditure, no drawings on race days, no claimed losses and a sudden loss of interest in racing, was incapable of belief on her own evidence. It was also, and this is routinely missed, an appeal against an order of the Settlement Commission, so the Court was asking whether the majority had acted unreasonably or on no evidence, not rehearing the merits. Durga Prasad More concerned self-serving recitals in a deed. Neither is authority that a third party’s manipulation can be attributed to an investor who did nothing.

Second, the same judgments cut the other way. Sumati Dayal itself says that “while considering the explanation of the assessee the Department cannot, however, act unreasonably”. And the closing words of Durga Prasad More are that the decision of the final fact-finding authority on the reliability of evidence “is made conclusive by law”, which is the very proposition on which the Tribunal’s findings have been upheld in Krishna Devi, Indravadan Jain and Parasben Kochar.

Third, the Madras formulation. In Sohanraj Uttamchand the court framed the answer better than it has been framed anywhere else:

The ‘test of human probabilities’ enunciated in those decisions is undoubtedly a valid and useful tool of appreciation of evidence, but it does not dispense with the requirement that the Revenue must, at the very least, bring on record some positive material connecting the assessee with the alleged design, particularly where the primary documentary evidence of the transaction is not under challenge.

The same court distinguished both decisions expressly: Sumati Dayal concerned unaccounted funds under the guise of winnings from races, “a matter of pure chance”, and Durga Prasad More unexplained cash credits unsupported by any documentary trail comparable to that in a listed-share transaction.

Fourth, the counterweight from the Supreme Court itself. Omar Salay and Umacharan Shaw are of the same vintage and higher pedigree on the specific point that suspicion is not proof. The Department’s reliance on human probabilities is an invitation to decide on inference; the assessee’s answer is that inference is permissible only within the evidence, not in place of it.

Does the price rise alone prove anything?

No, on the weight of authority, though it is always the lead point.

The Madhya Pradesh High Court in Aniruddha Nyati held that a “mere sudden increase in share prices on the stock exchange, does not ipso facto determine that shares are bogus”. The Gujarat High Court in PCIT v. Mamta Rajivkumar Agarwal held that price movement “cannot be a sole criteria for reaching to the conclusion that the bogus long-term capital gain was generated”. The Delhi High Court in Krishna Devi accepted a 4,849 per cent jump and still deleted. The Surat bench in Subhashchandra Patel held that mere price variation is not unexplained income absent direct proof of the assessee’s link to manipulation, and the Rajkot bench in Meena Ashok Shah that “subsequent appreciation in market value cannot establish that original acquisition was fictitious”.

The honest qualification is that price rise plus assessee-specific material is a different case altogether, and that is what separates the two lines. In Suman Poddar there were no more than contract notes and no evidence of actual sale. In Udit Kalra the company had meagre resources and reported consistent losses, the findings of all three authorities were concurrent, and the court declined to reopen them; the scrip, Kappac Pharma, has since been the subject of a deletion upheld by the Gujarat High Court on different facts, which is a reminder that the scrip is not the case. In Sanjay Bimalchand Jain the shares were of unlisted private companies bought off market through two intermediaries sharing an address and signatories. In the Madras group led by Manish D. Jain (HUF) the purchase was off market, the seller untraceable, and a group of family members did the identical trade. Those are not cases about an investor with a clean chain; they are cases where the chain itself was missing or broken.

What internal contradictions are worth taking?

Three appear in most orders and each has judicial support.

Cost of acquisition allowed while the sale is called sham. The Gujarat High Court in Mamta Rajivkumar Agarwal put it exactly: “On one hand, the AO is treating the entire transaction as sham transaction and on the other hand he’s allowing the benefit of the cost of acquisition for the shares.” The Punjab and Haryana High Court noticed the same thing in Hitesh Gandhi, that the officer “himself gave credit to assessee for indexed cost of acquisition”.

The gross receipt taxed rather than the gain. Where the investment was made in an earlier year and is disclosed in that year’s balance sheet, only the gain can be brought to tax in the year of sale. The Jharkhand High Court so held in PCIT v. Manoj Kapoor, reducing an addition of Rs 10,45,266 to the gain of Rs 5,05,265 because the cost “was duly shown in balance sheet of the Assessee in previous year and was made out of past earning and savings”. That is a quantum point available even where the genuineness argument fails, and in that case it was the only point the assessee won.

Receipt and payment asserted together. If the sale consideration is the assessee’s own unexplained money coming back to him under section 68, and he also paid a commission under section 69C to arrange it, the Department is characterising the same money in two mutually inconsistent ways in the same order. That argument is developed below.

What do the recurring fact patterns prove?

Four patterns recur and each has been litigated.

Long holding. A holding of several years is close to dispositive in practice, because an accommodation entry is priced and timed and does not sit in a demat account for a decade. The Gujarat High Court in Jagat Pravinbhai Sarabhai deleted where the shares had been retained for more than ten years. The Mumbai bench in Elara India Opportunities Fund held that shares held more than ten years cannot be a penny stock holding. The Gujarat High Court affirmed deletion in Shivani Ashokbhai Shah where the shares had been acquired by the assessee’s mother and held over fifty-five months, and in the Gangani matters on a holding of about two and a half years.

Subscription through a public issue. The Jaipur bench in Birendra Singh Nirbhay held that “every company making an IPO cannot be termed a shell company”, and that the Investigation Wing’s conclusions about brokers and entry operators do not apply to a direct subscription. The Nagpur bench reached the same result on an IPO subscription in GCM Securities.

Preferential allotment and amalgamation-linked swaps. This is the harder pattern, because it is the precise modus SEBI found in First Financial Services Ltd, where the price of the scrip rose from Rs 5 to Rs 263, an increase of 5,160 per cent in 115 trading days, and eighty preferential allottees turned a collective investment of about Rs 14.50 crore into collective profits of about Rs 172 crore. Expect the Department to lead with it. The Mumbai bench nevertheless deleted in Dipesh Ramesh Vardhan, where Santoshima Tradelink shares acquired by preferential allotment were swapped on amalgamation into Sunrise Asian, holding that the “vital link” between the assessee and the other entities was never established. The Bangalore bench went the other way on Sunrise Asian in Pradeep Kothari, where both an Investigation Wing report and a SEBI adjudication report found manipulation and the assessee did not appear.

Loss cases rather than gain cases. Where the transaction produced a loss, the motive theory collapses outright, because there is no unaccounted money being converted into an exempt gain. The Delhi bench deleted on that footing in Usha Jain, where the transaction in fact yielded a long-term capital loss. But the argument has its limits, and the limit is instructive. The Kolkata bench distinguished Swati Bajaj on exactly this ground in Zulu Merchandise, ITA No. 553/Kol/2024 decided on 23 September 2024, holding that a business loss was a different case from an implausible exempt gain. The Calcutta High Court set that order aside on 1 August 2025. So in West Bengal the loss distinction has been tried and has failed, and the same judgment is now the Revenue’s principal recent weapon. Elsewhere the point remains available.

What does Swati Bajaj decide, and how far does it reach?

Any honest treatment of this subject has to take Swati Bajaj on its own terms, because it is the decision that laid down a rule rather than declining to interfere with a finding of fact, and because a line of Calcutta decisions has since applied it to the same effect.

PCIT v. Swati Bajaj and Others, ITAT No. 6 of 2022 and about ninety connected appeals, was decided by the Calcutta High Court on 14 June 2022 and is reported at (2022) 446 ITR 56. The lead facts were 50,000 shares of Surabhi Chemicals and Investment Ltd bought for Rs 1,00,000 and sold for Rs 29,23,500 within seventeen to twenty-one months. The court allowed the Revenue’s appeals, set aside the Tribunal’s common order and restored the assessments as affirmed by the Commissioner (Appeals). It did not remand.

Its holdings, compressed:

• the onus is on the assessee to bring itself within the exemption, and to establish that the rise in price within a short period was genuine;
• the documentary chain proves process, not economic reality;
• the test is the preponderance of probabilities, and the allegation may be inferred “by a logical process of reasoning from the totality of the attending facts and circumstances”;
• the Investigation Wing report was properly relied on, supply of it is not mandatory, and the test is prejudice;
• there is no vested right to cross-examine persons who have not deposed against the assessee;
• where the price rise was achieved by manipulative practices, “whatever resultant benefits which accrue from such manipulative practices are also to be treated as tainted”.

That last proposition is the one that matters, and it is the true point of departure from the rest of the country. On the Swati Bajaj approach, the absence of a link between the assessee and the rigging is not the answer; the presence of rigging at the other end is enough unless the assessee disproves taint.

How far does it reach? Three things can fairly be said.

Within Calcutta it governs, and it has been applied repeatedly: in Rajesh Kumar Damani, in B L Tak and Sons HUF, and most importantly in PCIT v. Zulu Merchandise Pvt Ltd, ITAT No. 88 of 2025 decided on 1 August 2025. Zulu is the decision the Department will put to you, because it expressly declines to follow Krishna Devi on the ground that the Delhi judgment was “rendered on January 15, 2021 prior to the decision in Swati Bajaj”, and it holds that Swati Bajaj “appears to have not been placed before the Hon’ble Court when it dealt with the matter in Dipansu Mohapatra”.

Even within Calcutta it is not absolute. The same court carved out Bakra Pratishtan Ltd, a scrip not in the list of 84, in Deepa Agarwal and Abhay Kumar Jain. It dismissed the Revenue’s appeal in Sawankumar T Jajoo in January 2025 where the assessee demonstrated the scrip did not appear in the AIR report. And in Pradip Kumar Jajodia HUF in August 2025 it upheld the Tribunal on invalid reopening notwithstanding Swati Bajaj being cited.

Outside Calcutta it is a non-jurisdictional decision, and the ordinary rule applies that the Tribunal follows its own High Court. Benches at Mumbai (Kush N. Shah), Indore (Narayan Lal Jagetiya), Nagpur (Indrakumar Ghisulal Agrawal), Surat and Ranchi (Bijoy Kumar Agarwal) have said so in terms, preferring their own High Court. The Gujarat High Court settled the question for Gujarat on 31 August 2026 in PCIT v. Shivani Ashokbhai Shah, holding that “the observation and final decision of the judgment passed by the High Court at Kolkata has no manner of application in the instant case due to difference in the facts”.

The contrary illustration is Bhupesh Nemani, decided by the Patna bench on 31 August 2026, which applied Swati Bajaj although the Calcutta High Court is not its jurisdictional High Court. Where a bench has no decision of its own High Court to follow, the pull of Swati Bajaj is real.

Where does each High Court stand?

What follows is a court-by-court map. Two structural cautions should be read with it.

The first is that most of the favourable corpus consists of refusals to interfere with findings of fact under section 260A. It establishes that these additions do not survive where the Tribunal finds no evidence. It does not establish a rule of law that they can never survive. The Bombay formulation is the clearest, from PCIT v. Minkal Doshi: the questions proposed would require the court to “re-assess or re-evaluate the evidence on record, as if we were exercising the first appellate powers”, and this is “not a case of either no evidence or exclusion of relevant evidence”.

The second is that the same is true of most of the adverse corpus. Udit Kalra, Suman Poddar, Sanjay Kaul and Chandan Gupta all turn on the Tribunal’s findings not being perverse. The symmetry is almost exact: in Punjab and Haryana the same judge wrote Chandan Gupta against the assessee and Hitesh Gandhi for the assessee, and in both the reason was that the Tribunal’s finding of fact was not perverse. Swati Bajaj is the outlier in that respect, and that is precisely why it dominates: it is the decision that answered questions of law rather than refusing to reopen findings of fact, and the Calcutta decisions that follow it, including Hill Queen Investment, Usha Devi Modi, Bina Gupta and Zulu Merchandise, have done the same.

Bombay. A settled line of dismissals on the section 260A threshold: Shyam R. Pawar (2015) 229 Taxman 256, Jamnadevi Agrawal (2010) 328 ITR 656, Mukesh Ratilal Marolia, Ziauddin A Siddique (ITA No. 2012 of 2017, 4 March 2022), Indravadan Jain (HUF) (ITA No. 454 of 2018, 12 July 2023), and a run of 2025 decisions including Minkal Doshi, Komal Kumarpal Shah and Kishore Hargovind Patel (HUF). Against that, Sanjay Bimalchand Jain (Nagpur Bench, 10 April 2017) went for the Revenue on facts. The important qualification is that Bombay is not closed. Between October 2025 and June 2026 the court admitted several Revenue appeals in penny stock matters, including CIT v. Deepak Valji Karia, ITA No. 38 of 2024 admitted on 8 October 2025 on the cross-examination question, and a group of Shah matters framing section 68 and the section 69C commission together. Those admissions are not decisions and none of that group has yet been decided, though the court has since dismissed the connected Jiwarajka appeals, which is dealt with below.

Gujarat. The most developed and most consistently favourable body of authority: Parasben Kasturchand Kochar (SLP dismissed on the merits formula), Jagat Pravinbhai Sarabhai, Mamta Rajivkumar Agarwal, Divyaben Parmar, Vinod Premjibhai Gangani, Ambalal Chimanlal Patel, the VAS Infrastructure trio, Champalal Gopiram Agarwal, and Shivani Ashokbhai Shah confining Swati Bajaj. Note the distinction Gujarat itself maintains: writs against reopening in penny stock cases have been dismissed, for example in Vilas Vrajlal Parekh HUF, so the merits line and the reopening line must be kept apart.

Madras. Both the strongest recent reasoning in the country and a settled line the other way. CIT v. Sohanraj Uttamchand, T.C.A. Nos. 714 and 721 of 2018 with T.C.A. Nos. 8 and 9 of 2023, decided 28 August 2026 and carrying the neutral citation 2026:MHC:3510, contains the fullest reasoned answer against the Revenue on the merits rather than a dismissal at the threshold, and it also upheld deletion of the section 271(1)(c) penalties. Against it stands the Bakra Pratisthan group of 2020 and 2021, Manish D. Jain (HUF), Rakhee Jain, Leela Devi, Vinod Kumar Kundammal and Prabha Jain, where the purchase was off market, the seller untraceable and a family group did the same trade. Tharakumari, frequently cited against assessees, is a non-rebuttal case: the assessee missed five hearings and filed nothing.

Delhi. Krishna Devi (ITA 125/2020 with ITA 130/2020 and 131/2020, 15 January 2021), Karuna Garg (ITA 477/2022, 23 November 2022) and Reeshu Goel (ITA 173/2021, 14 December 2021) for the assessee; Udit Kalra, Suman Poddar and Sanjay Kaul against. Delhi has never reconciled the two lines expressly, and Krishna Devi does not cite the earlier three. The reconciliation is evidentiary: in the adverse cases the officer had assessee-specific material and there were concurrent findings against the assessee; in the favourable cases the Tribunal found for the assessee on the evidence and the officer had made no independent enquiry.

Calcutta. Swati Bajaj and its progeny, as above, against an earlier favourable line (Shreyashi Ganguli, Rungta Properties, Classic Growers, Alpine Investments, Carbo Industrial Holdings, Emerald Commercial, Lakshmangarh Estate, “suspicion can never be taken as proof”) that consists almost entirely of short unreported Division Bench orders.

Punjab and Haryana. Hitesh Gandhi (ITA No. 18 of 2017, 16 February 2017) is the parent authority and is a search case, which makes it valuable: nothing gathered in the search or the post-search investigation showed a sham. Prem Pal Gandhi follows it expressly. Chandan Gupta (2014) went the other way on concurrent findings. There is no post-2018 decision of that court on the issue and none engaging with Swati Bajaj.

Orissa. The most favourable of all, and the court on which the Supreme Court has been asked to speak most often. Dipansu Mohapatra and its batch, and Kuntala Mohapatra, all turning on denial of cross-examination. Note the one qualification: in Bimala Devi Singhania the court admitted the Revenue’s appeal, holding that the section 10(38) question did raise substantial questions of law where the group had admitted before the authorities, after a survey, to having paid tax on bogus long-term capital gain. No final judgment in that matter has been traced.

Madhya Pradesh. Strongly favourable and the single largest body of authority anywhere: the Indore batch of thirty-three appeals led by Aniruddha Nyati, 30 April 2024, and Aditya Agrawal at Jabalpur, 1 April 2026. No adverse Madhya Pradesh decision was found. It should not, however, be presented as settled: the Supreme Court issued notice on the Revenue’s petitions against the Aniruddha Nyati batch on 30 June 2025.

Allahabad. Renu Aggarwal, whose SLP was dismissed with the question of law kept open, and Nita Rastogi, decided 13 May 2026, on borrowed satisfaction in the reopening.

Rajasthan. Pooja Agarwal (2017) for the assessee, and PCIT v. Sanjay Chhabra, reported at (2023) 453 ITR 516, where the Revenue’s appeals were dismissed because the material gathered was not disclosed and cross-examination was denied. The Supreme Court proceedings against Sanjay Chhabra were restored to the file in November 2025 and are pending, so it should not be described as having survived the Supreme Court.

Jharkhand. PCIT v. Manoj Kapoor, 16 August 2023, is a mixed result and, read honestly, an adverse data point on genuineness: the court did not disturb the finding that the gain was bogus and allowed the appeal only in part, on the quantum point that the cost of acquisition disclosed in an earlier year could not be added in the year of sale.

Gauhati. PCIT v. Rohit Karan Jain, ITA/5/2023 decided in March 2025, the only decision from the north-east, deleting an addition of Rs 4,23,30,000 on the footing that other than a retracted statement there was no material, and that no addition can be made in a completed assessment without incriminating material.

Karnataka, Kerala, Telangana, Andhra Pradesh, Chhattisgarh. No decision on this issue has been traced in any of them. Karnataka does have a decision worth knowing for a different reason: in the Bagaria group, decided 20 December 2023, prosecutions under section 276C arising out of penny stock claims were quashed for want of mens rea, which is a reminder that the addition and the prosecution are different questions with different standards.

Patna, Himachal Pradesh, Uttarakhand, Sikkim, Jammu and Kashmir, Manipur, Meghalaya, Tripura. As at September 2026, no reported decision of any of these courts on this issue has been traced.

At a glance:

High Court	Leading authority for the assessee	Leading authority for the Revenue
Bombay	Indravadan Jain (HUF), Ziauddin A Siddique, Shyam R. Pawar	Sanjay Bimalchand Jain; six appeals admitted 2025-26, undecided
Gujarat	Parasben Kochar, Divyaben Parmar, Shivani Ashokbhai Shah	None on the merits; reopening writs dismissed
Madras	Sohanraj Uttamchand (merits, 2026)	The Bakra Pratisthan group, 2020-21
Delhi	Krishna Devi, Karuna Garg, Reeshu Goel	Udit Kalra, Suman Poddar, Sanjay Kaul
Calcutta	Sawankumar T Jajoo; the pre-2022 line	Swati Bajaj; Zulu Merchandise, Rajesh Kumar Damani
Punjab and Haryana	Hitesh Gandhi, Prem Pal Gandhi	Chandan Gupta (2014)
Orissa	Dipansu Mohapatra, Kuntala Mohapatra	Bimala Devi Singhania (admitted; no final order traced)
Madhya Pradesh	Aniruddha Nyati batch, Aditya Agrawal	None traced
Allahabad	Renu Aggarwal, Nita Rastogi	None traced
Rajasthan	Pooja Agarwal, Sanjay Chhabra (SLP pending)	None traced
Jharkhand	Quantum point only in Manoj Kapoor	Manoj Kapoor on genuineness
Gauhati	Rohit Karan Jain	None traced
Karnataka, Kerala, Telangana, AP, Chhattisgarh	No decision traced	No decision traced
Patna, HP, Uttarakhand, Sikkim, J&K, Manipur, Meghalaya, Tripura	No decision traced	No decision traced

Where does the Tribunal stand, bench by bench?

The Tribunal is where these cases are actually won and lost, and the pattern is consistent enough to tabulate. The selection below is of decisions whose orders have been read, weighted towards 2025 and 2026, and it deliberately includes the benches that went the other way.

Bench	Case	Outcome and ground
Mumbai	Kush N. Shah, ITA 4387/Mum/2023, 10.09.2026	Deleted. Report never confronted, no cross-examination, “a fatal flaw”; jurisdictional High Court preferred over Calcutta
Mumbai	ACIT v. Pravin Selvadia, ITA 2571 and 4329/Mum/2024, 31.12.2025	Deleted. SEBI had cleared the shares; no direct material
Mumbai	ITO v. Sunita Chaudhary, ITA 2124/Mum/2024, 19.01.2026	Deleted. SEBI had exonerated the assessee and removed the restraint
Mumbai	Ramprasad Agarwal, ITA 4843/M/2018, 30.11.2018	Deleted. Single statement, no cross-examination
Mumbai	Dipesh Ramesh Vardhan, ITA 7648/Mum/2019 and others, 11.08.2020	Deleted. Preferential allotment and amalgamation swap; “vital link” never established
Delhi	Gaurav Aggarwal, ITA 3009/Del/2026, 16.09.2026	Deleted. SEBI cleared 70.12 per cent; assessee in no charge-sheet
Delhi	Bhawna Kapoor, ITA 632/Del/2020, 09.04.2026	Deleted. Penny stock label plus appreciation does not establish fraud
Delhi	Usha Jain, ITA 4605/Del/2025, 25.08.2026	Deleted. Transaction in fact produced a loss
Delhi	Meenu Goel, ITA 6235/Del/2017, 19.03.2018	Deleted. Authorities relied only on the Investigation Wing report
Delhi	Pooja Ajmani, ITA 5714/Del/2018, 25.04.2019	Sustained. Price jump in a company of unknown credentials
Ahmedabad	ITO v. Sahil Vinod Agarwal (HUF), ITA 2008 and others/Ahd/2025, 23.09.2026	Deleted. No link established to the group indulging in rigging
Ahmedabad	DCIT v. Girishkumar Amratlal Bhandari HUF, ITA 977 to 979/Ahd/2025, 04.11.2025	Sustained. Price Rs 2 to Rs 469, known operators, premeditated structure
Ahmedabad	Hanumanprasad Lalchand Gupta, ITA 1514/Ahd/2025, 14.09.2026	Sustained. Cumulative circumstances including a SEBI suspension circular
Kolkata	Chandu Somani, ITA 2029 and 2503/Kol/2024, 26.02.2026	Deleted. Scrip absent from the official list of 84; borrowed satisfaction
Kolkata	Shyam Sunder Somani, ITA 2395/Kol/2024, 16.09.2026	Deleted. Same ground
Kolkata	Manoj Jain (HUF), ITA 1782/Kol/2018, 21.09.2023	Sustained. Swati Bajaj treated as binding jurisdictional precedent
Chennai	Sohanraj Uttamchand, ITA 1787/CHNY/2017, 28.02.2018	Deleted; affirmed by the Madras High Court on 28.08.2026
Bangalore	Pradeep Kothari, ITA 2388/Bang/2024, 13.08.2026	Sustained. Investigation report and SEBI adjudication both found manipulation
Jaipur	Birendra Singh Nirbhay, ITA 704/JPR/2024, 09.10.2025	Deleted. IPO subscription; SEBI revoked restrictions
Hyderabad	Smark Securities, ITA 2194/Hyd/2025, 22.07.2026	Deleted. SEBI action against the company alone insufficient
Indore	Narayan Lal Jagetiya, ITA 953/Ind/2025, 24.06.2026	Deleted. No cross-examination on the AIMS data; Swati Bajaj territorially inapplicable
Nagpur	DCIT v. Indrakumar Ghisulal Agrawal, ITA 220/Nag/2023, 10.04.2026	Deleted. IPO allotment; reports “general in nature”
Surat	Subhashchandra Patel, ITA 92/SRT/2025, 10.09.2026	Deleted. Ten-year holding
Rajkot	Meena Ashok Shah, ITA 656/RJT/2025, 16.09.2026	Reassessment quashed and, in the alternative, addition deleted on merits
Raipur	Milly Sewak, ITA 555/RPR/2026, 07.07.2026	Deleted. No nexus with entry providers established
Ranchi	Bijoy Kumar Agarwal, ITA 310/Ran/2025, 06.01.2026	Deleted. Jurisdictional High Court authority wrongly ignored below
Visakhapatnam	Ashok Kumar Agrawal, IT(SS)A 10/VIZ/2025 and others, 13.06.2025	Deleted. SEBI penalised the brokers; no evidence the assessee acted in connivance
Patna	Bhupesh Nemani, ITA 436/PAT/2025, 31.08.2026	Sustained. SEBI adjudication recorded synchronised trading; Swati Bajaj applied
Chandigarh	Deepak Jain, ITA 1220/CHD/2025, 18.09.2026	Quashed. Mismatch between the recorded reasons and the transaction assessed
Pune	Madhu Hareshkumar Jain, ITA 1698/Pun/2026, 23.09.2026	Quashed on approval under section 151, not on merits
Kolkata	Mukesh Kumar Jhawar, ITA 1632/Kol/2026, 03.09.2026	Quashed. Notice issued 31.03.2021 but served 01.04.2021

Four things are visible in that table and are worth saying out loud.

The overwhelming majority of reasoned Tribunal decisions delete, and they delete for the same reason. The decisions that sustain are not aberrations: they are cases with assessee-specific material, or where the assessee did not appear, or where the bench considered Swati Bajaj binding on it. A significant and growing proportion of recent wins are on jurisdiction and limitation rather than on the merits, which is a reminder that the reopening deserves as much attention as the addition. And no Special Bench or Third Member reference on this issue has been traced, which is itself telling about why the divergence persists.

The section 69C commission addition

This deserves separate treatment because it is the least defensible part of the assessment and it is very often conceded by default, on the footing that a few lakh rupees is not worth fighting about. With section 115BBE applied to it and penalty under section 271AAC riding on it, that calculation is usually wrong.

What the section says. Section 69C applies “where in any financial year an assessee has incurred any expenditure and he offers no explanation about the source of such expenditure”. Three textual points follow. The charge is on expenditure incurred; the incurring is the jurisdictional fact and the source is only the second enquiry. The section says the amount “may be deemed” to be income, so it is discretionary. And the proviso bars deduction; it does not authorise estimation.

What the Assessing Officer actually does. He takes a percentage of the sale consideration and adds it. The observed percentages across the orders are two, three, four, five, six and eight per cent, sometimes two different percentages in the same order. There is no bank entry, no cash book, no seized paper and no statement recording any payment by the assessee. The expenditure is inferred from the alleged scheme, and the alleged scheme is inferred from the price chart.

The first answer is consequential. Section 69C presupposes an accommodation entry to have been arranged. Once the section 68 addition goes, there is nothing left for a commission to have been paid on. Benches have deleted on exactly that basis: Mumbai in Pravin Selvadia (“This addition was consequential to the section 68 addition. Once the primary addition was deleted, this commission charge lacked foundation”) and in Kush N. Shah, Delhi in Gaurav Aggarwal, Jaipur in Birendra Singh Nirbhay, Nagpur in Indrakumar Ghisulal Agrawal, Jodhpur in Ram Niwas Chouhan and Ahmedabad in Ice Worth Reality LLP, where a notional commission of Rs 3,04,38,743 went with the principal addition. At High Court level the Bombay court dismissed an appeal in which the commission question was expressly framed, in CIT (Central)-3 v. Dilip B. Jiwarajka, ITA No. 849 of 2023 decided on 15 April 2026. Question (E) was in terms about “alleged commission/expenditure paid @ 5 to 7 % by invoking the provision of Section 69C of the Income tax Act on the non genuine sale value of shares without any cogent material on record”. The court did not analyse the commission separately: it dismissed the whole appeal as raising no substantial question of law, following the connected matter of Surendra B. Jiwarajka. That is the closest thing to High Court authority on the point, and its limits should be acknowledged.

The second answer is independent and survives even if the principal addition stands. The Department must first show that expenditure was incurred. The Nagpur bench said so in Indrakumar Ghisulal Agrawal: the addition “lacked evidentiary basis: onus is on the Assessing Officer to show first that the expenditure is incurred”. The Bombay High Court has stated the general principle for section 69C in PCIT v. Vaman International Pvt Ltd (2020) 422 ITR 520, which is a bogus purchases case and not a penny stock case: without causing further enquiries and giving an opportunity to examine or cross-examine, it is not open to the officer to make an addition under section 69C. And in the reopening context the Bombay High Court in Gaurang Manhar Gandhi described an assumed brokerage of Rs 33,56,500 as “purely speculative”, there being no evidence that any such payment was made.

The third answer is that the Department’s own institutions do not support the practice. The Comptroller and Auditor General, in Report No. 11 of 2020, Chapter VI, selected 547 penny stock cases across twenty-nine Commissioner charges in Mumbai and audited 499 of them, and found, on the commission, that “there was no consistency in the approach in disallowing the same”, the disallowance varying from 0.5 per cent to 5 per cent across the sixty-nine cases in which it was made, with a further forty cases in which no commission was disallowed at all. That is not an advocate’s complaint about arbitrariness; it is the constitutional auditor’s finding. And the Board itself, by letter F.No. 225/26/2026-IT A.II dated 21 May 2026 addressed to all Principal Chief Commissioners and Directors General, has directed that in relation to sections 68, 69A, 69B, 69C and 69D “the Assessing officer has to satisfy himself as to the true nature and source of the amounts for which such sections are invoked”, that the enquiry powers in section 133(6) may be used for the purpose, and that the relevant provisions are to be invoked “based upon inquiry and facts of the case”. A percentage of turnover is neither an inquiry nor a fact of the case, and it does not satisfy anybody as to the true nature and source of anything.

The letter should be cited for what it is. It was issued on the Comptroller and Auditor General’s draft report and it is principally revenue-protective: its remaining directions are that section 115BBE be applied to such additions, that set-off be denied, and that the computation columns be filled correctly on the Board’s own system. It is not an instruction to make fewer additions. What it does supply is the Board’s own statement of the standard an Assessing Officer must meet before invoking section 69C, and that standard is inconsistent with a percentage plucked from the air.

The fourth answer is the internal contradiction. If the sale consideration is the assessee’s own unexplained money returning to him, taxed under section 68, then he did not receive an accommodation entry from anyone; and if he paid a commission to obtain one, then the money was not his. The Department cannot have both in the same order, and the order that asserts both is asserting two mutually inconsistent characterisations of a single receipt.

On the other side, no penny stock decision has been traced in which a section 69C commission was sustained after the section 68 addition was sustained. In the sustained decisions reviewed for this note no commission appears to have been added at all. That is a bounded negative rather than an absolute, and it should be stated as such; the Jiwarajka question, framed on a commission of five to seven per cent, shows that commissions are added readily enough in cases the Department regards as strong. The nearest sustained-commission authority is not a penny stock case: in A and R Buildmart Pvt Ltd the Delhi bench reduced a 3 per cent commission from Rs 1,80,000 to Rs 90,000 to match the entry actually proved, which is itself authority that the commission cannot exceed and cannot outlive the proved principal.

What has the Supreme Court actually said?

Less than either side claims.

There is no Supreme Court judgment deciding this question. As at the date of this note, every intervention by the Court in this field has been at the special leave stage: short orders refusing leave, or orders issuing notice. No reported judgment decides whether, and when, the sale consideration on listed shares may be assessed under section 68 where exemption is claimed under section 10(38) or section 112A.

That matters because of Kunhayammed v. State of Kerala (2000) 245 ITR 360, which is the authority to cite whenever an order refusing leave is pressed as approval:

An order refusing special leave to appeal may be a non-speaking order or a speaking one. In either case it does not attract the doctrine of merger. An order refusing special leave to appeal does not stand substituted in place of the order under challenge.

Only where leave is granted and the appeal decided does merger operate. With that in hand, the orders in this field sort into categories of quite different weight:

Category	Examples
Revenue’s petition withdrawn	Krishna Devi, SLP (C) No. 4068 of 2022, withdrawn by the Revenue on 30 January 2025
Dismissed on the “not inclined to interfere” formula	Parasben Kasturchand Kochar (2021), Kuntala Mohapatra (2024), Mukesh Ratilal Marolia (2014), Sanjaykumar Damjibhai Gangani
Dismissed with the question of law expressly kept open	Renu Aggarwal (2023) and much of the Orissa cluster
Dismissed on limitation alone	Sitansu Sekhar Mohapatra, a delay of 309 days not condoned
Assessee’s petition dismissed in limine	Suman Poddar (2019)
Pending	Bindu Garg, SLP (C) No. 419 of 2022, the lead matter, still at the notice stage and listed for 1 October 2026; Aniruddha Nyati, notice issued 30 June 2025; Sanjay Chhabra, restored to the file in November 2025

Three observations follow.

The strongest single point for the assessee is not a dismissal at all. The Revenue challenged Krishna Devi, notice was issued, and then the Revenue withdrew its own petition on 30 January 2025, a fact recorded by the Supreme Court in a later order in a connected matter. A judgment whose challenge has been abandoned stands on stronger ground than one whose challenge was refused, because there is no Kunhayammed objection to make.

The Orissa cluster is the cleanest illustration of why dismissals prove nothing. The same line of High Court reasoning produced a dismissal on the merits formula in Kuntala Mohapatra and a dismissal purely on a 309-day delay in Sitansu Sekhar Mohapatra, weeks apart. The Department’s appeals there failed mostly on limitation, which is worth nothing to anyone on the merits.

And the Department’s use of Suman Poddar is wrong. A non-speaking dismissal of the assessee’s own petition declares nothing and approves nothing.

Of the leading High Court decisions on the favourable side, no Supreme Court order at all has been traced in Ziauddin A Siddique, Indravadan Jain (HUF) or Prem Pal Gandhi; and, notably, none has been traced against Swati Bajaj either. It is not correct to say that Swati Bajaj has been approved by the Supreme Court, and it is equally not correct to say it has been set aside.

Where is the law going?

Three things are worth watching, and a practitioner writing grounds today should know about all of them.

The lead Supreme Court matter is PCIT v. Bindu Garg, SLP (C) No. 419 of 2022, against the same Delhi High Court judgment of 15 January 2021 that decided Krishna Devi. It has been pending since January 2022 and as recently as 24 August 2026 had still not got past service of notice: the Registrar recorded that the Revenue had not filed the spare copy required for notice to issue, granted a week’s last opportunity, and directed that on compliance notice issue and the matter be listed on 1 October 2026. A substantial batch of matters from Madhya Pradesh, Rajasthan and elsewhere is tagged behind it. Two civil appeals that were once tagged with it were de-tagged in April 2023 as involving different issues, so it should not be assumed that leave has been granted anywhere in this cluster.

The second live cluster is less well known and matters just as much. The Revenue has challenged the Madhya Pradesh batch: in PCIT v. Aniruddha Nyati, SLP (C) Diary No. 20425 of 2025, the Supreme Court issued notice on 30 June 2025 on the applications for condonation of delay and on the petitions themselves, directing that they be listed with another pending matter. The largest body of favourable High Court authority in the country is therefore under challenge, and an article or a submission that presents it as settled is overstating it.

The Bombay High Court has admitted at least six Revenue appeals since October 2025, framing a standard set of questions on the Investigation Wing statements and SEBI orders, on Sumati Dayal and Durga Prasad More, on the section 68 and section 69C additions together, and on Sanjay Bimalchand Jain. Anyone telling a Bombay client that the question is settled in his favour is ahead of the record.

And the constitutional auditor is pressing in the other direction. The Board’s letter of 21 May 2026 was issued on a draft report of the Comptroller and Auditor General on the application of sections 68, 69A, 69B, 69C and 69D read with section 115BBE, which, as the letter records, pointed to variation in invoking those provisions and to incorrect application of the rate of tax. Expect assessments to become better documented, not fewer.

What is the position under the Income-tax Act, 2025?

Of prospective interest only, but worth stating, because the recodification does not change the argument.

The 2025 Act came into force on 1 April 2026. Section 10(38) was in any event inoperative from assessment year 2019-20, having been withdrawn by the Finance Act 2018, and it remains the live provision for assessment years 2005-06 to 2018-19, which is where the entire body of case law sits. Every penny stock appeal now in the system is a matter under the 1961 Act. For the position on which enactment governs a given proceeding, see which Act governs your appeal.

Looking forward, cash credits under section 68 become section 102, headed unexplained credits. Unexplained expenditure under section 69C becomes section 105, and one change there is worth noting: where section 69C said the amount “may be deemed” to be income, section 105 says it “shall be deemed”, removing the officer’s discretion. What has not changed is the jurisdictional fact, which remains that expenditure “has been incurred by the assessee”, so the onus argument set out above survives the recodification intact. The provision corresponding to section 115BBE is section 195, and here there is a change of substance that cuts the other way. As enacted, section 195 carried the same sixty per cent rate. The Finance Act 2026 substituted thirty per cent for sixty per cent with effect from 1 April 2026, and the standalone penalty provision corresponding to section 271AAC was omitted and subsumed into the general penalty provision. So for tax years under the new Act the consequence argument made above, that a notional commission is not a rounding item because of the rate and the penalty riding on it, is materially weaker. It remains good for every year still in dispute under the 1961 Act, which is all of them.

How should the defence be built?

Work in this order, because the earlier steps dispose of more cases than the later ones and cost less to establish.

Assemble the chain before arguing anything. Contract notes for purchase and sale, the broker’s ledger, bank statements showing the debit for the purchase and the credit for the sale, the demat statement showing the date of credit and the holding period, and the securities transaction tax evidence. Anything missing from that chain is where the case will be lost, and it is better to know at the outset.

Ask for the Department’s material in writing, and early. The Investigation Wing report, the appraisal report, every statement relied on, and the SEBI order. Ask by reference to the specific document named in the show cause notice. If it is not supplied, record that, because the strength of the cross-examination ground depends on showing that the material was the foundation of the order and was withheld.

Establish whether the assessee is named anywhere. In the report, in any statement, in any SEBI order, in any list. In the great majority of these cases he is not, and that single fact is the case.

Trace the SEBI history of the scrip to its end. The interim order is the beginning of the story, not the end. Revocation orders, final orders and Securities Appellate Tribunal decisions are public, and the arithmetic in Kailash Auto Finance, where 244 of the 246 restrained were cleared and only two of them were ultimately sanctioned, shows what is often waiting there.

Demand cross-examination in writing, specifically, and repeat it. Name the person, refer to the statement, and put it on record at the assessment stage. The ground is far weaker when raised first in appeal.

Test the reopening and the approval. A very large and growing share of recent wins are on jurisdiction rather than merits: borrowed satisfaction, reasons that are conclusions, notices that do not specify the year, mismatches between the recorded reasons and the transaction assessed, approval by the wrong authority under section 151, and notices served a day late. For the approval and procedural machinery see section 148A, and for the treatment of statements recorded during search see section 132(4).

Take the section 69C commission separately and on its own ground. Do not let it ride in the slipstream of the principal addition, and do not concede it as immaterial. Ask what evidence exists that any expenditure was incurred, and put the C&AG’s finding and the Board’s letter of 21 May 2026 on the record.

Know which High Court binds the bench. Outside Calcutta, Swati Bajaj is a non-jurisdictional decision, and where the jurisdictional High Court has spoken the other way that should be the first submission, not an afterthought. Where the bench has no jurisdictional decision of its own, expect Swati Bajaj to carry real weight and prepare to distinguish it on the facts rather than to dismiss it.

Finally, be candid with the client about the split. On identical facts, the outcome today still depends materially on where the appeal is heard. That is an uncomfortable thing to say and it is true, and it will remain true until the Supreme Court decides the matter that is currently waiting for a hearing date.

Related reading: after an income-tax search, from panchnama to block notice; cross-examination and on-money additions on the natural-justice ground in a different setting; and when a penalty notice must specify the limb.

This note is general commentary on the law as at 27 September 2026 and is not advice on any matter. The position in a particular case depends on its own facts.