Tools

Extra Tax When Accumulated Income Was Never Recorded

Income was taxed inside your holding every year and you never recorded it. What does that cost in extra tax when you sell?

The decision this page was built for, in the words it was recorded in: my accumulation units never paid me anything, but I have been taxed on income inside them every year. If I have not kept those records, how much extra capital gains tax do I pay when I sell, and after how many years does it matter? The calculation below is that question with the country, the tax and the kind of holding taken out of it, because none of the three changes the arithmetic and all three would have limited it to one place.

The surfaces that explain this problem all start from the same place: take your yearly records and add them to what the holding cost. That is correct and it is no help, because the reason anyone looks this up is that the records were never kept. The tools that do maintain the running adjustment sit behind an account you may no longer have, and they only cover what is held on that platform. So this page starts from a reconstruction instead, and it answers the half of the question nothing public answers: not only what the missing records cost, but after how many years the amount is worth doing anything about.

What this tool does not decide

  • Any of the eight numbers. The cost, the proceeds, the yearly income, the allowance and the rate are all yours, and no control on this page carries a default. That is not caution: it is the whole reason one page can be correct wherever this mechanism exists rather than correct in one country.
  • What the yearly income actually was. It is a reconstruction, the answer moves with it, and this page cannot check it. Calculate twice, once at each end of what you think it might have been, and treat the pair as the range rather than either figure as the answer.
  • What you owe. This is one difference between two tax figures on one disposal. Other disposals, other income, losses from another year and everything else in your position are absent, and any of them can change the actual amount.
  • Whether to sell, or what to file. An arithmetic is not a filing position. Where the amount matters, the figure to act on is one produced by someone who can see your actual records.
  • Anything about tax law, in any jurisdiction.

Your figures

Your holding, in eight numbers

All eight figures are yours and this page publishes none of them. It names no country, no tax, no scheme, no fund and no allowance that anybody legislates, and there is no field that could carry one. The yearly income figure is the hard one, and it is the reason this page exists: the surfaces that explain this arithmetic assume you kept the records, and the platforms that keep the running adjustment for you are behind an account you may no longer have. Take the yearly figure from the fund’s own published distribution history if you can, and treat the answer as an estimate that is exactly as good as that figure. Nothing you type is sent anywhere.

What it cost and what it fetched

What you actually paid, as your records state it. This is the figure the whole calculation corrects: income taxed inside the holding belongs in it, and if it was never added then this number is too low and every figure built on it is wrong in the same direction. There is no currency here and no field for one.

Enter 0 if it was written off. A figure below what you paid is not an error and is one of the three cases this page exists to separate: at a loss there is nothing to tax and the missing records cost nothing, which is the opposite of what the usual answer will tell you.

The income you were taxed on and never saw

One level figure, not a year-by-year series. That is deliberate: you are here because the year-by-year records are missing, and asking for them would be asking for exactly what is absent. Where a fund publishes its distribution history you can estimate from that. The answer moves with this number, so it is worth calculating twice — once at each end of what you think it might have been.

How many years the income above was taxed inside the holding. Enter 0 if it was sold in the year you bought it, which comes back with nothing owed and still answers the second half of the question.

What decides how much of it is taxed

Enter 0 if no such allowance applies to you. This is the figure that does the absorbing, and it is the reason the usual answer is often too high: once the corrected gain falls inside it, further years of unrecorded income cost nothing more. This page publishes no figure here — the allowance is a jurisdiction's and it moves.

Enter 0 if no rate applies, and the page will report that nothing is at stake rather than refusing to describe the position. There is no suggested figure here on purpose: the rate depends on the jurisdiction, on the kind of holding and often on the rest of your income, and this page knows none of the three.

What would count as material, and how much to draw

Your own threshold, and there is no suggested figure. Enter 0 if any amount matters to you. It is used for one thing: to report the first year at which the missing records cost more than that, which is a fact about the arrangement rather than about your own holding period.

This is the only control that changes no figure. Every number in the result is computed at your own number of years whatever you draw here. Drawing past your own year is the point rather than a side effect: the second half of the question is about years you have not held yet.

Result

Almost every explanation of this problem gives the same answer: apply the rate to everything you failed to record. That is right in one case out of three, and where it is wrong it is wrong in the direction that frightens people who are not exposed. Adding the missing income to what the holding cost reduces the gain, and a smaller gain can fall inside the amount you may realise tax-free, or below zero altogether. Past that point the cost of the missing records stops rising while the general rule keeps going. The holder with the largest omission is often the one least exposed to it, and a holder who sold at a loss owes nothing while being told they owe the rate on all of it.

Enter your eight figures above and select Calculate. Nothing is sent anywhere: the calculation runs in this browser, and no value is stored, shared or placed in the address bar.

The usual answer against the true one

Three rows and never one. The answer you would be given elsewhere is published so the correction can be seen, and it is never published without the difference beside it.
Which figureWhat it comes to
What the general rule says: the rate on everything unrecordedNot yet calculated
What it actually is, once the exempt amount is taken into accountNot yet calculated
What the general rule overstates it byNot yet calculated

What each further year of unrecorded income costs

Two lines against years held. The straight one is the general rule: the rate on everything unrecorded, rising by the same amount every year without limit. The other is the true answer, which follows it exactly and then flattens. The year they part company is the year the exempt amount starts absorbing the omission, and everything to the right of it is what the general rule gets wrong. Your own year is marked, and so is the first year at which the amount passes the threshold you set.

The chart appears here once you calculate. Every figure it draws is also published above it as a number, so nothing on this page depends on being able to see it.

What the holding actually cost

The income you were taxed on and never added to the cost
Not yet calculated
What the holding actually cost you, with that income added
Not yet calculated

The two gains, and the two taxes on them

The gain on the cost you have a record of
Not yet calculated
The gain once the missing income is added to the cost
Not yet calculated
The tax on the gain as your records state it
Not yet calculated
The tax on the gain as it actually is
Not yet calculated
What the missing records cost you
Not yet calculated

Which case you are in, and when it starts to matter

How much of the omission is actually taxed
Not yet calculated
Which of the three cases you are in
Not yet calculated
The year at which this passes your own threshold
Not yet calculated
Whether it ever reaches that threshold
Not yet calculated

There is no field here for how fast the holding grew, and no projection of what it might be worth later. Two reasons, and the second is the more useful one. A growth rate would be an expected return, which no authority publishes and this site does not invent. And it would answer the wrong question: what makes the missing records matter is the income that accumulated, not the price that moved, so the years are swept against a fixed disposal in order to isolate the one thing you are asking about. Growing the holding forward as well would have moved two things at once.

Every figure here is tax on one gain at one disposal. It is not the tax on your whole year, it is not your total bill, and it is not what you owe: other disposals, other income and everything else in your position are not in this arithmetic and cannot be added to it. What it isolates is one quantity — the difference between the tax on the gain your records state and the tax on the gain as it actually is.

This is arithmetic on eight numbers you supplied, and one of them is a reconstruction. The yearly income figure is taken as level across every year held, which no real fund’s distributions are: a fund that paid nothing for five years and a great deal afterwards has the same total on a level figure and a different one year by year. The answer moves with that figure, so it is worth calculating twice, once at each end of what you think it might have been. Nothing here handles units bought part-way through a distribution period, a loss carried to another year, part disposals, or averaging across separate holdings of the same thing.

This is not advice about whether to sell, what to file, or what to do about records you do not have. It is one difference between two figures, on numbers you supplied, and a reader whose reconstruction of the yearly income is wrong gets a confident answer to a different question. A disposal has consequences this arithmetic contains none of. Where the amount matters to you, the figure to act on is one a person who can see your actual records has produced.

How this is worked out

Why your cost base is too low

Income that a holding earns and pays out to you is taxed once, when you receive it. Income that a holding earns and keeps is taxed too, in the year it arises, even though nothing reaches you. Adding that income to what the holding cost is what stops it being taxed a second time when you sell.

If it was never added, you do not have a missing receipt. You have an understated cost and therefore an overstated gain, and the overstatement is the whole of the income you were already taxed on. That is the failure this page prices, and it is a failure of record-keeping rather than of arithmetic, which is why the surfaces that explain the arithmetic are no use: they start from records you do not have.

Why the usual answer is right once and wrong twice

The answer you will be given almost everywhere is the rate on everything you failed to record. It is what the arithmetic produces while both gains are comfortably above the amount you may realise tax-free. Adding the missing income to the cost pushes the gain down, and once the corrected gain falls inside that amount — or below zero — nothing further can be taxed. The extra tax stops rising. The general rule does not.

These are the calculation model's own canonical figures, at its own canonical holding: bought for 30,000, accumulating 900 a year, against an allowance of 3,000 and a rate of 20%. They are a worked example of the mechanism and not a suggestion about anybody's numbers.

One holding, three positions. The general rule is right in the first row, a quarter too high in the second, and the whole of a figure that is zero in the third.
The positionWhat the general rule saysWhat it actually isWhy
Ten years, sold at a gain1,8001,800Both gains clear the allowance, so every unit of the omission is taxed. This is the case the general rule is written for, and here it is right.
Twenty-five years, sold at the same gain4,5003,400The corrected gain has fallen through the allowance and out the other side into a loss. A quarter of the omission is absorbed, and the holder most alarmed by its size is the one least exposed to it.
Ten years, sold at a loss1,8000There is no gain to tax, so the records cost nothing at all. The general rule here is not merely too high: it is the whole of a figure that is zero.

The arithmetic

Eight numbers you supply, and nothing derived from anywhere else. The tax is computed twice, once on the gain your records state and once on the gain as it actually is, and the answer is the difference. One expression covers all three cases rather than each being special-cased, which is why the boundaries fall out correctly instead of having to be found.

omitted(y)        = yearly income x y

gain as recorded  = proceeds - cost
gain corrected(y) = proceeds - cost - omitted(y)

tax(g)            = rate x max(0, g - exempt amount)

extra tax(y)      = tax(gain as recorded) - tax(gain corrected(y))
                  = min( tax(gain as recorded), rate x omitted(y) )

starts to matter  = the first whole year at which extra tax passes your threshold

The second line of the extra-tax expression is the finding. It is a minimum of two things: the rate on everything unrecorded, which is what the general rule gives, and a ceiling equal to the whole tax that was payable on the gain your records state. The records cannot save more tax than was payable without them, and where the ceiling is the smaller of the two, every further year of unrecorded income costs nothing at all.

After how many years does it matter

The second half of the question is not about the answer's size but about when it becomes worth doing anything about, and it needs a threshold only you can set. The year reported is the first whole year at which the extra tax passes it.

That year is a fact about the arrangement rather than about you, and it does not move with how long you have held. A holder who bought this year gets the same answer as one who has held for twenty, which is deliberate: the reader at year zero is the only one who can act on it by keeping the records rather than reconstructing them.

Why there is no growth rate

The field this page most obviously lacks is a rate at which the holding grows, so that a future disposal value could be projected. It is refused twice over. It would publish an expected return, which no authority publishes and this site does not invent. And it would answer the wrong question: what makes the missing records matter is the income that accumulated, not the price that moved, so the years are swept against a fixed disposal in order to move one variable at a time.

What this does not model

  • an income figure that varies from year to year rather than being level
  • units bought part-way through a distribution period, and the adjustment that follows
  • any rate of growth, projected value or expected return
  • a loss carried back or forward, and any relief on one
  • part disposals, averaging across separate holdings, indexation and rebasing
  • any other disposal, any other income, and anything else in your position
  • any country, tax, scheme, wrapper, fund, allowance or published rate
  • anything you may do after the disposal
  • whether to sell at all

The limits of the calculation

  • Every figure is yours. This page publishes no cost, no proceeds, no allowance and no rate, and there is no field that could carry one. That is not caution: it is what makes one surface correct wherever this mechanism exists rather than correct in one country.
  • The yearly income figure is a reconstruction and the answer is exactly as good as it is. It is taken as level across every year held, which no real fund’s distributions are — a fund that paid little early and a great deal later has the same total and a different year-by-year shape. Calculate twice, once at each end of what you think it might have been.
  • Units bought part-way through a distribution period carry an adjustment this page does not model, and no free surface found by the research behind this Asset models it either. Where that applies to a meaningful part of your holding, the figure here is an approximation.
  • This is one disposal in isolation. Other disposals, other income, losses carried from another year, part disposals and averaging across separate holdings of the same thing are all absent, and each of them can change what is actually owed.
  • A corrected gain below zero is reported as a gain below zero and the tax on it is zero. What may be done with that loss is a different decision and this page has no view on it.
  • Nothing here says whether to sell, what to file, or what to do about records you do not have. Where the amount matters, the figure to act on is one a person who can see your actual records has produced.

A small amount that happens every year and is invisible in any single one is the subject ofa separate explainer on what a recurring deduction does over time. That explainer is about a charge that leaves the holding and this page is about income that stayed in it; what they share is the annual quantity nobody looks at until it is a decade deep.

Go deeper

  • The habit of mind this page needs

    How Investment Fees Compound Into Lost Wealth

    Why a small annual investment fee can create a much larger long-term difference, why the second component of that difference is signed, and where the arithmetic stops.

    An amount too small to notice in any one year, repeated for a decade, is the whole of the story. That explainer is about a charge that leaves the holding and this page is about income that stayed in it; what they share is the annual quantity nobody looks at until it is years deep.

    Read the explainer: How Investment Fees Compound Into Lost Wealth

Calculation model and corrections

Calculation model
Accumulated Basis Omission v1.0
Last reviewed
Why the usual answer is often too high
Because adding the missing income to what the holding cost reduces the gain, and a smaller gain can fall inside the amount you may realise tax-free or below zero altogether. The extra tax is the rate on everything unrecorded only while both gains clear that amount. Past that point it stops rising and the rule keeps going
Whose figures these are
Yours, all eight of them. This page publishes no cost, no proceeds, no allowance and no rate, and there is no field that could carry one. That is what makes the tool universal: the mechanism is the same wherever it exists and the numbers are the part that is not
What the yearly income figure is
Income taxed inside the holding each year and never paid out to you, as a single level figure. It is the reconstruction, and the whole answer moves with it. There is deliberately no year-by-year series: you are here because those records are missing, and asking for them would be asking for what is absent
Why the answer is a difference rather than a bill
Because it is the tax on the gain your records state minus the tax on the gain as it actually is. Both are computed the same way from the figures you supplied. Neither is your liability, and nothing else in your position is in the arithmetic
What happens at a loss
Nothing is taxed and nothing is owed, and this is reported rather than refused. It is one of the three cases the calculation exists to separate, and it is the one where the usual answer is most alarming and least true
Why there is no growth rate
Because what makes the omission matter is the income that accumulated rather than the price that moved. The years are swept against a fixed disposal so that one variable moves at a time, and a growth rate would additionally publish an expected return, which no authority publishes
What the year at which it starts to matter is about
The arrangement, not you. It does not change with how long you have held, which is why a holder at year zero still gets an answer: they are the one reader who can act on it by keeping the records rather than reconstructing them
The share that is taxed
How much of the omission is actually taxed, from none of it to all of it. Where no rate applies there is no share to publish and the page says so, rather than printing zero — a share that does not exist and a share that is zero are different facts
How far the chart is drawn
As far as you ask, and it changes no figure. Every number above it is computed at your own number of years. The sweep deliberately runs past your own year, because the second half of the question is about years you have not held yet
The unit
Whatever unit you entered the cost in, throughout. This page names no currency and there is no field for one
Excluded
Every country, tax, scheme, wrapper, fund, allowance and published rate; an income series that varies year to year; units bought part-way through a distribution period; losses carried back or forward; part disposals, averaging and indexation; every other disposal and every other kind of income; and any view on whether to sell
Rounding
Display only; intermediate values remain unrounded

Correction history

  • The year at which the omission starts to matter is published for a holder who has held for no years at all, and the first draft of the model did that without having argued for it. Writing the canonical case forced the argument, and it changed what the figure means rather than what it is: it is a fact about the arrangement and not about this holder’s exposure, and a holder at year zero is precisely the one who can still act on it. Returning nothing there would have withheld the answer from the only reader who could use it to keep the records rather than to reconstruct them. The behaviour is unchanged and is now an invariant, asserted across every year from zero to the end of the sweep.
  • A share of the omission that does not exist and a share that is zero were nearly published as the same thing. A holding with no rate has no share to report, because nothing defines it; a holding that has a rate and an omission, none of which is taxed, has a share and it is genuinely zero. Reporting both as absent would have merged a quantity that does not exist with one that exists and is zero, and a reader in the second case would have been unable to tell that the arithmetic had reached an answer at all.