2025 (8) TMI 1527 - HC - Income Tax

Deletion of income-tax addition upheld where assessing officer failed to establish nexus between consumption and presumed low yield

The Deputy Commissioner of Income Tax (Central), Raipur, Chhattisgarh Versus M/s Abhishek Steel Industries Ltd.,

The Deputy Commissioner of Income Tax (Central), Raipur, Chhattisgarh Versus M/s Abhishek Steel Industries Ltd., - 2025:CGHC:42527 - DB ISSUES PRESENTED AND CONSIDERED 1. Whether the Assessing Officer was justified in making an addition for alleged unaccounted production and sales by applying an estimated production yield of 89% in the Steel Melting Shop division. 2. Whether statistical/mathematical estimates derived from seized materials and variations in inputs (electricity, sponge iron, furnace oil) can, without tangible corroborative material, sustain rejection of books of account and additions under assessment provisions invoked after search and seizure. 3. Whether the principles that an assessing authority cannot act on mere suspicion or pure guess (as articulated in Dhakeswari Cotton Mills Ltd. and applied to assessment provisions) apply to additions made under the assessment provisions engaged in this matter. ISSUE-WISE DETAILED ANALYSIS Issue 1 - Legitimacy of Addition Based on an Estimated 89% Yield Legal framework: Additions under assessments post-search are governed by the relevant provisions of the Income Tax Act (assessment under the provisions invoked in the judgment). An Assessing Officer may estimate income/production where records are found unreliable, but such estimation must rest on material and not mere conjecture. Precedent Treatment: The Court referred to the constitutional bench decision establishing that an assessing authority is not entitled to make a pure guess; there must be more than bare suspicion to support an assessment based on estimation. Interpretation and reasoning: The AO adopted an 89% yield as a benchmark to compute alleged unaccounted production/sales. The record shows (i) failure to disclose the basis for arriving at 89% despite requests, (ii) lack of any comparable instance where 89% was the declared industry yield, and (iii) industry/peer comparison producing a different mean yield (approx. 81.35%) and yields declared by the assessee (approx. 83.94%). The CIT(A) and ITAT analysed seized materials, books, excise returns and other documents and found that the AO did not demonstrate a nexus between his statistical calculations and the purported 89% standard. The AO relied on variations in consumption figures without proving that such variations established unaccounted production or sales. The AO did not point to tangible discrepancies in the accounts or excise records that would render books unreliable. Ratio vs. Obiter: The conclusions that an addition cannot be sustained where the prescribed benchmark (89%) is unexplained and unsupported by record is ratio insofar as it was dispositive of the appeal and derived from concurrent factual findings. Conclusions: The AO's addition based solely on the unexplained 89% yield was unsustainable; the appellate authorities correctly set aside the addition because the AO failed to justify the adoption of that benchmark by reference to cogent material. Issue 2 - Reliance on Statistical/Mechanical Calculations from Seized Data to Reject Books of Account Legal framework: While an assessing authority may rely on material not admissible in court and use statistical analysis, rejection of books and consequential additions require tangible adverse material establishing unreliability of accounts; statistical inference cannot substitute for evidentiary proof of suppression. Precedent Treatment: The judgment relies on the principle from Dhakeswari Cotton Mills Ltd. that an assessment cannot be founded on bare suspicion; the assessing authority must have something more than conjecture. The coordinate-bench decisions dealing with similar facts (same search, same line of business) were recognized and treated as persuasive in the factual matrix. Interpretation and reasoning: The AO emphasized monthly variations in inputs and consumption without demonstrating that such variation was inconsistent with normal operations, capacity utilization, or documented production cycles (including planned shutdowns). The CIT(A) examined excise returns, day-to-day records and other seized material and found consistency with books. Witness statements and explanations from production department regarding measurement and estimation of burning loss were considered and not shown to be contradicted by the AO with tangible proof. The AO's approach relied disproportionately on generic statistics and mechanical computations unlinked to concrete discrepancies in bookkeeping. Ratio vs. Obiter: The holding that statistical variations, absent concrete evidence of falsity in records, cannot justify rejection of books is ratio as applied to the case; observations about proper weight to be given to statistical data are instructive but also operative in decision. Conclusions: Statistical and mathematical calculations based on seized material, without demonstrated nexus to inaccuracies in books or independent corroboration, do not justify additions or rejection of accounts; the appellate authorities correctly rejected the AO's estimate-driven additions. Issue 3 - Application of the Principle Against Acting on Mere Suspicion (Dhakeswari Principle) Legal framework: The constitutional-bench principle that an assessing officer cannot make assessments on pure guesswork governs the exercise of power where assessments are founded on estimation in the context of search-derived materials. Precedent Treatment: The Court expressly applied the Dhakeswari principle as binding and controlling on the use of estimation under the assessment provisions engaged. Interpretation and reasoning: The appellate authorities independently evaluated seized materials, reconciled loose slips with books and excise returns where possible, assessed witness statements, and found no tangible adverse material warranting rejection of books. Where the AO's conclusions rested on suspicion and conjecture, the principle forbids sustaining such additions. The Court found no perversity in the concurrent factual findings by the CIT(A) and ITAT that the AO proceeded on suspicion and that the books could not be impeached by tangible evidence. Ratio vs. Obiter: The application of Dhakeswari's requirement of 'something more than bare suspicion' to deny an addition based on unexplained estimation is squarely ratio of the Court's decision in this appeal. Conclusions: The Dhakeswari principle precludes sustaining AO's additions founded on unexplained statistical estimates and suspicion; the concurrent conclusions of the appellate authorities that the AO's action amounted to guesswork are valid and dispositive. Cross-References and Concluding Legal Findings 1. The three issues are interrelated: the legality of employing a particular estimated yield (Issue 1) depends on whether statistical/mathematical inferences from seized material can rebut books (Issue 2), which in turn is governed by the principle that estimation must be supported by more than suspicion (Issue 3). The appellate authorities applied this integrated analysis. 2. Concurrent factual findings by the first appellate authority and the tribunal that (a) the AO did not disclose the basis for 89% yield, (b) records including excise returns and books were consistent, and (c) no tangible adverse material was produced, are not perverse and constitute the foundation for setting aside the addition. 3. The appropriate legal conclusion drawn from the foregoing is that additions predicated on unexplained, benchmark statistical estimates and pure conjecture cannot be sustained; acceptance or rejection of books requires tangible imperfection, not mere variation or statistical suspicion.