Depreciation

Two regimes, and why they disagree on purpose

The books follow the Companies Act 2013, and specifically Schedule II. Schedule II prescribes a useful life for each class of asset. It does not prescribe a rate. The company chooses its method — straight line or written down value — applies it over that life, and leaves a residual value standing at the end. The result reaches the Profit and Loss account and reduces the carrying amount on the Balance Sheet.

The return follows the Income Tax Act, section 32, read with Rule 5 of the Income Tax Rules and Appendix I to those Rules. Here there is no individual asset at all. Assets are pooled into a block, additions go into the pool, sale proceeds come out of it, and the whole pool is written down at one prescribed percentage on the reducing balance. The figure never touches the books; it exists only in the computation of income.

So the same machine is being depreciated twice, on different facts, by different arithmetic, to different totals in any given year. Neither number is wrong. Attempting to make them agree — by adopting Appendix I rates in the books, which was tolerable practice before 2014 and is not now — produces accounts that do not comply with Schedule II and a computation that was never independently checked.

A great deal of older guidance describes the Companies Act as prescribing rates, because Schedule XIV to the Companies Act 1956 did exactly that. It was replaced by Schedule II, which prescribes lives. Material written against the 1956 Act is not a safe guide to the current position.

A life or a rate — which one you have been handed

This is the point most often held backwards, and holding it backwards produces a figure that is wrong in both regimes at once. The two schedules hand you different kinds of number, and each is complete on its own terms.

Schedule II hands you a life. There is no rate anywhere in it, and there has not been one since Schedule XIV to the 1956 Act was replaced. You take the life, you take your own residual value policy, you choose the method, and the rate is arithmetic you perform yourself. Two companies owning identical machines can properly carry different rates in their books, because they have made different method choices over the same prescribed life.

Appendix I hands you a rate. There is no life anywhere in it, and how long the asset is expected to last is simply not an input to the computation. A machine expected to last five years and one expected to last twenty are both written down at the general plant and machinery rate, in the same block, at the same percentage, for as long as the block exists.

• In the books, start from the life. Find the class in Schedule II, read the life, and derive the rate from it. The derived rate has to be recomputed whenever the life or the residual convention changes.
• In the computation, start from the block. Find the entry in Appendix I that describes the asset, and the percentage on that entry is the rate. Nothing about the asset's expected life enters the calculation.
• The two group assets differently, so the mapping is many-to-one in places. End user computers carry a 3-year life and servers and networks 6, while both sit in the same 40% Income Tax block. Office equipment carries a 5-year life and has no Income Tax head of its own at all, so it falls into general plant and machinery at 15%.
• Neither figure substitutes for the other. An Appendix I rate applied in the books produces accounts that do not comply with Schedule II; a life-derived rate applied in the computation is not the allowance section 32 gives.

The one question to ask of any depreciation percentage handed to you is which of the two produced it. A rate quoted without saying whether it was derived from a life or read off Appendix I cannot be checked, and cannot safely be used in either place.

Companies Act — deriving a rate from a life

Because Schedule II gives a life and not a rate, the rate has to be worked out. The two formulas are short, and both fall out of the same requirement: the method must consume the depreciable amount, being cost less residual value, over exactly the prescribed life.

Under the straight line method the depreciable amount is spread evenly, so every year takes the same slice of original cost:

SLM rate % = (100 − residual %) ÷ useful life, applied to original cost.

The same charge can be reached without the percentage at all, working straight from the money: Depreciation per year = (Cost − Residual value) ÷ Useful life. The two routes give the identical figure, because the rate was only ever the depreciable amount expressed as a percentage of cost.

Under the written down value method the rate is applied to a shrinking balance, so it has to be the rate that lands exactly on the residual value after the prescribed number of years. That makes it a root, not a division:

WDV rate % = [1 − (Residual value ÷ Cost) ^ (1 ÷ Useful life)] × 100, applied to opening book value. The ^ is a power — on a calculator, the x^y key — so with a 15-year life the bracket holds the fifteenth root of the residual fraction.

Each year's charge then follows from the balance rather than from cost: Depreciation for the year = Opening written down value × WDV rate %. The opening balance is smaller every year, which is why the charge tapers and why the rate must never be applied to original cost.

The WDV formula has no solution where the residual value is zero, because no percentage of a balance ever reaches zero — the root of zero is zero, and the rate comes out at a flat 100%, which would write the asset off entirely in year one rather than over its life. A company intending to write an asset down to nothing has to use the straight line method.

Whichever method is chosen, a part year is then reduced by days: Charge for the year × (days held ÷ days in the year) — × days held ÷ 365 in an ordinary year, 366 in a leap year — counting from the date the asset was available for use to the year end.

• Depreciation runs pro rata from the date the asset is available for use, not from the date of the invoice and not from the start of the year. An asset commissioned on 1 January carries three months of charge in that year.
• The residual value is a floor, not a target. The charge in any year is whatever the method gives, or whatever headroom remains above the residual, whichever is smaller.
• The method is a policy choice, applied consistently to a class of asset and disclosed. It is not chosen asset by asset to suit a result.

Every formula, in one place

Collected so that a figure can be checked without reading the argument around it. Each line is written to be keyed straight into a calculator, and each belongs to one regime only — the first six are the books, the four after them are the computation, and the last is the entry that reconciles the two.

• Everything in the first six lines is a percentage you compute. Nothing publishes it, and it moves the moment the life or the residual convention moves.
• Everything in the four lines after them is a percentage you look up. Nothing about the asset's expected life changes it.
• The rates, the 20%, the 5% residual cap and the 180-day threshold are all figures the legislature moves. Each one has to be confirmed against the current schedule before it reaches a signed account or a filed return.

The ^ in the WDV formula is a power, reached with the x^y key. Dividing where the formula roots — reading (0.05) ^ (1 ÷ 15) as 0.05 ÷ 15 — is the ordinary slip, and it produces a rate small enough to look plausible on the page and wrong in every year of the schedule.

The example, carried through

Take plant and machinery, general — the class Schedule II gives a useful life of 15 years and Appendix I pools into a block written down at 15%. Cost ₹9,00,000, available for use on 1 April so that the first year is a full one, straight line method, residual value at the 5% convention.

Residual value is 5% of ₹9,00,000, being ₹45,000. The depreciable amount is therefore ₹8,55,000. The straight line rate is (100 − 5) ÷ 15 = 6.33% of original cost, and the annual charge is ₹8,55,000 ÷ 15 = ₹57,000, unchanging for fifteen years.

• Had the company chosen written down value instead, the derived rate on the same life and residual would be 1 − (0.05) ^ (1/15) = 18.10% of opening book value.
• That produces a first-year charge of about ₹1,62,900 rather than ₹57,000 — nearly three times as much — on the same asset, the same life and the same schedule.
• The total over fifteen years is identical under both methods, at ₹8,55,000. Only the distribution across years differs, and with it every intervening year's reported profit.

Because the WDV rate is derived from the life, it must be recomputed whenever the life or the residual convention changes. A WDV rate typed into a fixed-asset register as a constant will keep running unchanged after the life beneath it has been amended, and nothing in the register will report the contradiction.

Three things Schedule II asks that a rate table never did

The move from prescribed rates to prescribed lives brought three obligations that had no equivalent under the old schedule, and all three are still commonly missed.

• Residual value. Schedule II caps residual value at ordinarily not more than five per cent of original cost. Most companies adopt the cap itself as policy, because a different estimate has to be supported and explained to the auditor. It is a convention with a justification behind it, not a default that applies itself.
• Componentisation. Where a part of an asset is significant in cost relative to the whole and has a useful life different from the remainder, that part must be identified and depreciated separately over its own life. A building's lifts and air-handling plant do not last as long as its frame, and treating the building as one indivisible asset understates the charge in early years and overstates it later.
• Revising a life. A useful life is an estimate, and estimates get revised when the facts change. A revision is treated as a change in accounting estimate and applied prospectively: the carrying amount remaining at the date of revision, less the residual value, is spread over the revised remaining life. Prior years are not restated and no catch-up entry is put through the opening reserves.

Componentisation is an accounting requirement, not a tax one. Splitting a building into components changes the book charge and changes nothing at all in the computation of income, because the whole building has gone into one block regardless.

Income Tax — the block of assets

Section 32 allows depreciation not on assets but on blocks. A block is the group of assets falling within a class in respect of which the same rate of depreciation is prescribed. Once an asset enters a block, it stops having a separate existence for tax purposes: it has no individual written down value, no individual charge, and no individual profit or loss.

The written down value of a block at the year end is computed under section 43(6): the opening written down value, plus the actual cost of assets acquired during the year and put to use, less the moneys payable in respect of assets sold, discarded, demolished or destroyed during the year — with that reduction limited so the block does not go below nil. Depreciation at the Appendix I rate is applied to what results.

Written as two lines that can be keyed in: Closing written down value = Opening written down value + actual cost of assets acquired and put to use during the year − moneys payable in respect of assets sold, discarded, demolished or destroyed. Then Depreciation for the year = Closing written down value × the Appendix I rate for that block.

The method is written down value only. There is no straight line election, no residual value, and no pro rata by days. There is exactly one refinement on timing, and it is the half-year rule below.

• Actual cost means the purchase price plus everything required to bring the asset to its present location and condition and into use, less any portion met by another person or by a subsidy.
• The test for the first year's claim is that the asset was acquired and put to use, not merely owned. An asset bought in February and commissioned the following April is not put to use in the earlier year at all.
• Depreciation under section 32 is mandatory. It is allowed whether or not the assessee claims it, so it cannot be forgone in a loss year to preserve a higher written down value.
• Depreciation that cannot be absorbed against income of the year is carried forward under section 32(2) and, once carried forward, is treated as part of the following year's allowance.
• Where a business is succeeded or amalgamated during the year, the allowance for that year is apportioned between predecessor and successor in the proportion the Act specifies rather than claimed twice.

Because the block absorbs the asset, the fixed-asset register and the tax computation stop corresponding almost immediately. A block carrying a written down value long after the books have written every asset in it down to residual is normal, not a reconciliation error.

The half-year rule

Where an asset is acquired and put to use for less than 180 days in the year, exactly half the Appendix I rate is allowed on it in that year. The full rate applies from the following year. There is nothing between the two — the rule is a switch, not a proportion, and 179 days and 3 days attract the identical treatment.

As arithmetic: where the asset ran for 180 days or more, Depreciation = Closing written down value × rate. Where it ran for less, Depreciation = Closing written down value × rate ÷ 2, and the undivided rate returns the following year. Nothing else in the year changes.

On the ₹9,00,000 machine above, in the 15% block, this is the difference between the two lines below.

• The half rate applies only in the first year. It is not a permanent handicap on the asset; the block catches up from the second year onwards.
• The rule bites on the asset, but the arithmetic happens inside the block, so a block containing both long-held and newly added assets is computed in two parts in that year.

Counting 180 days from the invoice date rather than from the date the asset was put to use is the ordinary version of this mistake, and it usually appears on assets bought late in the year — which is precisely where the rule is decided.

Additional depreciation under section 32(1)(iia)

Beyond the normal allowance, a further deduction is available on new plant and machinery, stated as a percentage of actual cost and allowed once. It is an industrial incentive rather than a measure of consumption, which is why its conditions are narrow and why so many assets that look eligible are not.

The deduction is confined to an assessee engaged in the manufacture or production of an article or thing, or in the generation, transmission or distribution of power, and to new plant and machinery acquired and installed by that assessee.

Where it is available, the arithmetic is a single line: Additional depreciation = Actual cost × 20%, claimed once, in addition to the normal charge on the block. Where the asset was put to use for less than 180 days it is halved on the same test the normal charge uses — Actual cost × 20% ÷ 2 in that year, and the remaining half in the year after. Note that it runs on actual cost throughout, not on the written down value of the block, so it does not taper the way the normal charge does.

• Second-hand plant — anything used by any other person, in India or abroad, before its installation — is excluded.
• Plant installed in office premises, in residential accommodation or in a guest house is excluded.
• Office appliances and road transport vehicles are excluded, which removes cars, lorries, computers used in an office, printers and air conditioners from the claim.
• Ships and aircraft are excluded.
• Anything whose entire actual cost is already allowed as a deduction, whether as depreciation or otherwise, in computing income of any one year, is excluded.
• Where the asset was put to use for less than 180 days, only half the additional depreciation is allowed in that year, and the unclaimed half is allowed in the year following. It is the only element of the computation that deliberately straddles two years.

Additional depreciation is not available to a company that has opted into a concessional corporate tax regime — the reduced rate is granted in exchange for giving up specified deductions, and this is one of them. Check which regime the company is in before the claim is built into a projection. The percentage itself, and the list of excluded assets, are both revised from time to time.

What happens when an asset is sold

This is where the block concept produces its most counter-intuitive result, and it is worth being precise about it. On a sale, the moneys payable are deducted from the block. That is the whole of the treatment. There is no profit on sale and no loss on sale in the computation of income, however large the difference between the sale price and what the asset was carrying.

Suppose a 15% block opens at ₹5,00,000, a machine is sold during the year for ₹1,20,000, and another is bought and put to use for ₹2,00,000. The block becomes ₹5,00,000 + ₹2,00,000 − ₹1,20,000 = ₹5,80,000, and depreciation is 15% of that. Whether the machine sold had been in the books at ₹40,000 or at ₹4,00,000 makes no difference at all to the computation. The books will show a profit or a loss on that sale; the computation will not, and the book figure is added back.

There are exactly two situations in which a gain or loss does surface, and section 50 deems both to be short term regardless of how long the assets were held.

The gain is short term even where the asset was held for twenty years. That closes off indexation and the concessional long-term rates, and it is a genuinely expensive surprise on a building or a piece of land-adjacent plant sold at a gain when its block empties.

The two regimes side by side

Collected in one place, because the differences are systematic rather than incidental.

• The asymmetry is not only in method. The two schedules also group assets differently, and that alone produces divergence.
• Schedule II gives end-user computers a life of 3 years and servers and networks 6; Appendix I pools both into a single 40% block.
• Schedule II has a 5-year life for office equipment; Appendix I has no office equipment head at all, so printers, photocopiers and air conditioners fall into general plant and machinery.
• A ₹3,00,000 set of computers is fully written down in the books after three years, standing at ₹15,000 residual. In the 40% block the same cost has attracted ₹1,20,000, ₹72,000 and ₹43,200, leaving ₹64,800 still to be written down, and ₹38,880 the year after that.

Deferred tax, briefly

Because the two regimes allow the same total in different years, tax expense computed on accounting profit does not equal the tax actually payable. Deferred tax is the accounting entry that reconciles the two, so that the charge in the Profit and Loss account relates to the profit reported beside it rather than to whatever the return happened to produce.

The mechanics for depreciation are simple enough to hold in the head. Where the tax computation has written the asset down faster than the books, the carrying amount in the accounts exceeds the tax written down value, tax has been deferred rather than saved, and a deferred tax liability is recognised. Where the books have run faster, the position reverses into a deferred tax asset.

The liability is measured by applying the tax rate to the difference between the two carrying amounts, not by taking the difference in the depreciation charge. On the ₹9,00,000 machine at straight line in the books and 15% in the block:

• At the end of year 1, ₹78,000 multiplied by the applicable tax rate is the deferred tax liability to be carried. The movement from the previous year's balance is what goes through the charge for the year.
• The difference widens while the faster regime is ahead, peaks, and then closes again as the slower one catches up. Over the whole life of the asset it returns to nothing, which is what makes it a timing difference and not a permanent one.
• A deferred tax asset is recognised only where its recovery meets the recognition test in the applicable standard — a stricter test where it rests on carried-forward losses or unabsorbed depreciation than where it rests on ordinary timing differences.
• The rate used is the rate enacted or substantively enacted for the period the difference is expected to reverse in, so a company moving between tax regimes has to remeasure its balance rather than roll it forward.

Deferred tax is recognised on the difference in carrying amounts, not on the difference between two depreciation charges. Computing it from the charges gives the movement for the year and no balance, and the error only becomes visible when the asset is sold or the rate changes.

What goes wrong most often

In roughly the order a practitioner meets them.

• Depreciation started from the invoice date rather than the date the asset was available for use.
• Appendix I rates applied in the books, producing a charge that does not comply with Schedule II and a computation that was never independently prepared.
• The 180-day test judged on ownership rather than on being put to use.
• Additional depreciation claimed on office equipment, computers or a vehicle, none of which qualify — or claimed by a company that has opted into a concessional regime and given the deduction up.
• Book profit or loss on the sale of an asset carried into the computation instead of being added back, with the block left unreduced.
• A revised useful life applied retrospectively, with a catch-up charge that belongs in no year at all.
• Components never separated, so a building depreciates as one asset for sixty years while its lifts are replaced twice.
• Deferred tax computed once at the acquisition of an asset and then left unadjusted as the difference between the two bases moves every year.