Continuous compounding future value

Continuous compounding is what discrete crediting becomes when the slices stop being slices. Crediting interest yearly, then monthly, then daily, then every second: the balance’s growth factor climbs with each refinement, but by less each time, and the climb has a destination. That destination is the exponential function — the future value is the principal multiplied by the exponential of the rate times the time, and no finite schedule ever quite reaches it.

The second output translates the convention into a familiar one: the equivalent effective annual rate is what one year of continuous growth amounts to, an APY with the schedule taken to infinity. It sits above the stated continuous rate by a margin that grows with the rate — and comparing it against a real account’s APY shows how little of the ceiling is left unclaimed once crediting is merely monthly. The frequency race is, in practice, over almost as soon as it starts.

The convention earns its keep outside bank brochures. Pricing models, academic papers and professional rate quotes often state continuously compounded rates because they behave beautifully under arithmetic: growth over stacked periods multiplies, so continuous rates simply add across time. A figure quoted in that convention cannot be dropped into a discrete comparison as printed — translating it is exactly what this calculator is for.

Additivity is worth pausing on, because it is the property everything else trades on. Under discrete conventions, chaining two periods means multiplying growth factors — an operation that resists mental arithmetic and breaks the symmetry between gains and losses. Under the continuous convention the same chaining is a sum: the rate for a long stretch is the stretch’s rates laid end to end. That is why the convention shows up wherever returns are aggregated across time, and why a figure met in the wild with the word “continuously” in front of it deserves translation before comparison.

The result is exact arithmetic on an idealised convention, not a claim about any account. Real deposits are credited discretely, real rates change, and the output is nominal — inflation, tax and fees all live outside it. What the exponential frame offers is not realism but a ruler: the clean upper edge against which every messy real schedule can be read.

The previous lesson left a process running: credit yearly, then quarterly, then monthly, then daily, and watch the yield climb by less each time. A process that keeps improving and keeps improving less invites one question — does it stop? It does. However finely the year is sliced, the growth factor is bounded above by a specific number, and the sequence of ever-finer schedules walks toward that number without ever arriving.

The reason is worth holding in words rather than symbols. Slicing the year more finely credits interest sooner, which gives the credited amounts longer to earn — but it also makes each credited amount smaller, because the annual rate is being divided into more pieces. The two effects pull against each other, and their balance tightens as the slices shrink. What survives the tug is the exponential.

Reading the schedules against that limit reorders the whole frequency argument. Annual crediting sits visibly below the ceiling; quarterly closes much of the remaining gap; monthly closes most of what is left; daily leaves a margin too small to matter to any depositor. The race is effectively over almost as soon as it starts, and the only way to know that is to compute the finish line.

The future value here is the principal multiplied by the exponential of the rate times the time. That single expression carries two distinct claims. The first is about the principal: growth scales linearly in it, so doubling the sum exactly doubles the result and the principal is the one input whose behaviour needs no exploration. The second is about time, and it is far less obedient — time sits in the exponent, so doubling it does considerably more than doubling the gain whenever the rate is meaningfully above zero.

The second output translates the frame back into ordinary language. The equivalent effective annual rate is what one year of this growth amounts to on exactly the footing the previous lesson established, which makes it directly comparable with a real account’s advertised yield. It always sits a little above the stated continuous rate, and the margin between them widens as the rate rises: at small rates the two conventions nearly coincide, at large ones they part company noticeably.

That margin is the entire prize the infinite schedule wins over a single yearly crediting. Naming it precisely is what turns a piece of mathematical folklore — that continuous compounding is somehow enormously better — into a measured quantity, which at ordinary rates turns out to be a good deal less impressive than the phrase suggests.

Pricing models, academic papers and professional rate quotes reach for the continuous convention constantly, and not out of affectation. The convention has a property the discrete ones lack: rates add across time. Growth over a long stretch, under a discrete convention, means multiplying one growth factor by another — an operation that resists mental arithmetic, breaks the symmetry between gains and losses, and grows unwieldy the moment more than two periods are involved. Under the continuous convention the same chaining is a sum, and the rate for a long stretch is simply the stretch’s rates laid end to end.

Additivity is why the convention appears wherever returns are aggregated, compared across unequal periods, or fed into models that need to run forwards and backwards without accumulating convention errors. It is a bookkeeping virtue rather than an economic one — no account is credited this way — but bookkeeping virtues are what make large calculations survivable.

The practical consequence for a reader is a habit: a figure met in the wild with the word “continuously” in front of it cannot be dropped into a comparison as printed. It belongs to a different convention, and translating it — which is exactly what the equivalent annual rate output does — has to happen before any ranking is attempted.

Nothing pays continuously. Real deposits credit on a schedule, real rates move, and the figure this page computes describes an idealisation with no counterpart at any bank. Treating it as an achievable return would be a straightforward mistake, and the right way to hold it is as a ruler: a clean upper edge against which the messy real schedules can be laid, so their differences can be seen in proportion rather than argued about.

Used that way it settles arguments quickly. How much is a bank’s move from monthly to daily crediting actually worth? Compute the ceiling, compute the monthly figure, and read the gap — most of which the monthly schedule has already taken. What is the absolute best this rate could ever do under any schedule imaginable? That is the ceiling itself, and a promise exceeding it is a promise about something other than compounding.

The output remains nominal in the inflation sense, exactly like the effective yield before it. It is the outer bound of growth in currency, and currency is not buying power. The next lesson takes whatever rate this journey has arrived at — quoted, effective, or continuous — and asks the question none of them has yet answered: what survives prices.

The engine below arrives pre-filled with the pack’s own declared example: a round principal at a mid-single-digit continuous rate, left alone for a few years — the textbook setup, deliberately so. Read the equivalent effective annual rate against the stated rate first; that small margin is the whole prize of an infinite schedule. Then hold the rate still and stretch the years, and watch the future value outrun the proportional guess by more the longer it runs.

Read the future value as an upper bound rather than a forecast — no account is credited continuously — and the equivalent annual rate as the figure to carry back to the previous lesson for comparison against a real schedule. Both are computed live by the verified engine; this page stores neither.