The Outcome Ledger

The Outcome Ledger is a free calculator for working out what money turns into over time. There is no account to make and no email to hand over. Nothing you type is sent anywhere — every figure stays in your own browser, which is also why your settings are still here when you come back and why nobody else can see them.

All eight tabs answer one question in different forms: what does this money actually become, once the things that quietly eat it have been counted? That means inflation, tax, fees, maintenance, depreciation and the years where the number goes down. A projection that leaves those out is not optimistic, it is just wrong, and it is wrong in the direction that costs you money.

The numbers each tab opens with are not picked at random and they are not predictions. They are long-run averages and current rates. The S&P 500 has returned about 10% a year since 1928 with dividends reinvested, which is nearer 7% once inflation is taken off. The Stocks tab opens at that historical rate: 8% price growth with a 2% dividend on top. These are nominal figures, so if you want the real one, tick “Show in today’s dollars” rather than lowering the growth rate — doing both takes inflation off twice. Inflation starts at 2.5%, roughly what central banks aim at, and the house price, mortgage rate and rent are close to the US medians for 2026. Every field’s i button says where its starting number came from, and every one of them is meant to be replaced with yours.

Where something is deliberately left out of the maths, the tab’s own explainer says so and says why. None of this is financial advice, and none of it knows anything about your situation.

Every rate this calculator opens with is listed below, with what it is based on. This is a calculator, not financial advice: it produces estimates for thinking with, it knows nothing about your situation, and it is not a substitute for an accountant or a licensed advisor.

The starting numbers are long-run averages and current rates, not predictions, and every one of them is meant to be replaced with yours. Listed here are the rates that actually drive the results. The remaining figures — repair costs, replacement cycles, closing and selling costs, lease residuals — each carry their source in the field’s own i button.

The growth rates are nominal. That is the one convention worth knowing before reading any result. Every rate above is before inflation, and “Show in today’s dollars” is what converts a result into what it would buy now. Entering a lower rate to allow for inflation and ticking that box takes inflation off twice, which understates every projection on the site.

What is deliberately left out. Each tab’s own explainer names what that tab does not model, and those are the honest parts. The Bonds tab does not show what happens to the price of a bond fund when rates move, because it models holding to maturity, where a rate change does not touch you. The Portfolio tab blends averages, which cannot show that real asset classes fall together in the worst weeks — historical mode is the better test, since it replays years that actually happened. The Debt tab’s invest-instead view compares a guaranteed return against a hoped-for one as though they carried the same certainty; they do not, and that should weigh toward clearing the debt. The Real Estate tab ignores the interest some places require on a held security deposit, which is small next to the compounding the deposit gives up. Real estate is left out of the Portfolio blend entirely, because a mortgage makes it a borrowed position that does not reduce to a percentage.

Worth knowing if you saved a scenario earlier: a changed starting number only reaches new visitors and anyone who uses “Reset this tab”. Your settings are kept in your own browser and are overwritten only where they already hold a value, so an older share link can produce slightly different figures than the same link made today. Nothing you type is sent anywhere. If a figure here is stale or simply wrong, it is worth saying so, so it can be corrected.

One pot of money compounded once a year. The rate it grows at is price growth plus the dividend yield, less the fund fee, so the three fields work together rather than one overriding the others. If dividends are reinvested more often than yearly, the yield is compounded at that frequency before it’s added, which is why quarterly reinvestment ends up slightly ahead of annual at the same yield. Contributions land throughout the year rather than all on 1 January, so a year’s deposits earn roughly half a year of growth; assuming otherwise is the most common way these projections quietly overstate things.

Contributions can run in up to three phases, so the amount can change as your life does. An employer match is a percentage of what you put in, capped, and the cap rises with inflation when your contributions are set to rise with it — a plan’s cap tracks salary, so freezing it would make the match quietly bite harder every decade.

Withdrawals can begin in a later year, and they come in three shapes. A fixed amount, a percentage of the balance, or the dividends alone. Only the last leaves the principal untouched; the other two sell from the pot to pay you, which is why they can empty it and dividends-only cannot.

Two of the switches change the method rather than the numbers. The Monte Carlo fan runs the whole projection a thousand times, with each year’s return drawn at random around your growth rate, spread by the volatility slider, then shows the tenth, fiftieth, and ninetieth percentiles instead of a single line. Where withdrawals are a fixed amount, it also reports how many of those thousand runs ran out of money. Historical mode does something different and stricter: it replays actual S&P 500 years from 1928 to 2024 in the order they happened, starting from three random years, so crashes arrive in real sequences rather than as an average.

The volatility slider only sets the width of that fan. It does not move the main projection line at all.

What the fan cannot show is that its returns are drawn each year independently and follow a bell curve. Real markets have fatter tails than that, and bad years arrive in clusters rather than politely spaced out. Historical mode is the honest test. Compounding here is annual rather than monthly, fees come off the growth rate as a flat percentage rather than as separately modelled trading costs and spreads, and the crash option is a single-year drop in a year you pick — a stress test, not a forecast.

Rates here are before inflation. Tick “Show in today’s dollars” for the real figure rather than lowering the growth rate yourself.

Two lines that answer different questions. One is what the car is worth as it depreciates: roughly a fifth gone in the first year, about 12% a year through year five, and around 7% a year after that. Buying three years old starts you further along that curve, so the same money drops more slowly from the day you buy.

The other line is what the car has cost you all in, and it is already net of the vehicle you still hold. Depreciation is the real expense, not the sticker price. Added to it: the loan interest, insurance and fuel, and repairs. Repairs are treated as zero for as long as the warranty runs, then start at the figure you set and compound at that rate every year after that, because an ageing car does not fail on a flat schedule. Insurance and fuel are charged at the same rate whether you own or lease, and they do not climb with age, so they lift both totals equally and never decide the answer.

With leasing switched on, the payment can be computed the way a dealer computes it: the depreciation you use up over the term, plus a finance charge on the money tied up in the car. Or you can type in a real quote. If that quote was given to you per week or every two weeks, it is first converted to a true monthly figure. Every two weeks is 26 payments a year, not 24, and reading it as half the monthly payment understates the cost by a full month of payments every year.

Leasing then splits in two. Lease again, and the payment steps up at each renewal; the car is always new, so the repair clock restarts every term, and if the term is shorter than the warranty, you never pay a repair bill at all. Buy it out, and you pay the residual plus the purchase fee, after which it is one continuous car costed exactly like the owned one; the buyout price is checked against what that car is actually worth on the open market, which is the only place a lease can be a bargain or a trap.

What this cannot know is your car in particular. The depreciation curve is a typical one, not any model’s resale record. Lease mileage limits, the charge for going over them, and the wear-and-tear bill at handback are not modelled, and they are where real leases go wrong. Neither is sales tax, on the purchase or on the payments, nor a trade-in, a lease down payment, or an accident. Loan interest is spread evenly across the term rather than front-loaded the way a real amortization schedule does it, which flatters the early years slightly. And nothing here invests the difference between the two payments for you.

The simplest engine on the site, and deliberately so. Your balance earns the yield you set; the interest is added once a year, and it compounds. That is genuinely how a GIC or a CD behaves when you hold it to the end, which is why this tab has fewer switches than the others and why the “doubles in” figure is a real answer rather than an estimate.

Monthly additions are added at the end of each year rather than spread throughout it, so a year’s deposits start earning in the year after they are made. That understates things slightly, which is the right direction to be wrong in. Additions can rise with inflation, and they can change partway through, because what you can set aside at thirty is rarely what you can set aside at fifty.

Withdrawals come in three shapes, and they behave differently on purpose. A fixed amount is the one that can empty the account, and the tab tells you in which year. The same amount, rising with inflation, empties it sooner and preserves its purchasing power while it lasts. A percentage of the balance each year cannot fully empty it, because you are always taking a share of what is left rather than a set sum, and the income moves with the balance in both directions. Withdrawals can start in a later year and can run in up to three stages, counted from the year they begin, so pushing the start date back slides the whole plan rather than eating the first stage.

The rate shock is not part of the projection. It draws a second line showing what the holding would fetch if you sold it before maturity after rates moved, scaled by how many years are left to run: a long bond falls much further than one about to mature. Held to the end, none of that touches you, and you receive the yield you agreed to. GICs and CDs cannot be sold at all, so the line for those is hypothetical.

What this does not model is the rest of what makes a bond a bond. There is no credit risk here: every payment arrives, and issuers do occasionally default. Bonds that can be called back early are not modelled, though a falling-rate world is exactly when that happens and exactly when it hurts. There is no ladder, no rate path, and no bond fund — one yield is assumed to hold for the whole term, so the risk of renewing maturing money at a worse rate does not appear anywhere. And interest is normally taxed as ordinary income rather than at capital-gains rates, while the tax setting here applies the capital-gains treatment, so it flatters this tab more than any other.

Metals produce nothing. No dividend, no interest, no rent — the entire return is whether someone later pays more than you did. So, this tab is one thing compounded: a price assumption, less a yearly charge for storage and insurance. The charge is taken off the growth rate before anything compounds, which is why it costs more than it would otherwise. A dividend compounds for you year after year; a vault bill compounds against you the same way, and the results panel shows what that difference came to over your horizon.

The starting growth rate is 5.5% per year, and where you begin measuring determines that number more than anything else. Gold since 1928 works out near 5.1%, but that stretch includes decades when the price was set by law rather than by buyers. Measured from 1975, from 1985, or from 1990, it lands at 5.4 to 5.5% every time, which is why that is the default. Counting from 1971 gives 8.2%, though most of the gap is one repricing when the gold standard ended — gold rose 127% in 1979 alone.

The comparison against dividend stocks is enabled by default and uses the Stocks tab’s own assumptions, so the two agree. What it shows is rarely a contest about which price rose faster. It is that one asset paid you while you held it and the other charged you.

What the arithmetic leaves out starts at the moment you buy. Physical metal sells above the spot price and buys back below it, and on coins that spread can run several percent in each direction, so a holding must gain before it breaks even. An exchange-traded fund avoids the spread and the vault bill but charges a management fee instead, and the storage field is the place to put it. Nothing here distinguishes gold from silver or platinum, though silver behaves quite differently because half its demand is industrial.

One smooth rate is also the least realistic thing on this page. Metals do not drift upward; they sit still for years and then move violently, and an average conceals exactly the part that decides whether you were glad you owned any. Gold bought at the 1980 peak did not recover its purchasing power until 2008 — twenty-eight years, with the average over the whole period still looking respectable. There is no volatility fan here to show that, unlike the Stocks tab.

Two more things are missing. Metal is priced in US dollars, so a Canadian holder’s return includes the currency move, which this does not model. And in the United States, physical precious metals are taxed as collectibles at a higher rate than long-term capital gains, while the tax setting here applies the ordinary capital-gains treatment — so like the Bonds tab, it flatters the result.

The most straightforward engine on the site, and the one whose answer is least about the arithmetic. Your balance earns the rate you set, credited monthly or once a year — monthly compounding gets you slightly more from the same stated rate, and the results panel shows the effective rate the two work out to. Deposits arrive throughout the year rather than all in January, so a year’s contributions earn roughly half a year’s interest.

The number worth reading is not the final balance. It is the line underneath showing what that balance buys in today’s money. Cash is the one tab where the nominal figure and the real figure tell opposite stories: a balance that grew every year can still be worth less than what you put in, and if your rate sits below inflation the tab says so outright, with the yearly loss of purchasing power spelled out. That is not a flaw in savings accounts. It is the price of the two things they are for — getting your money back, and getting it today.

Withdrawals are a flat monthly amount, and the tab reports whether the fund runs dry or lasts the whole horizon. There are no phases or percentage modes here, because a savings account is usually the thing you are drawing from between one event and the next, not the thing you retire on.

The rate is the point at which this projection is least believable, and it is the only input that really matters. High-yield savings rates float with central bank rates and are not locked in the way a GIC or a CD is. A 4% account can be a 1% account within a year, with no notice and nothing you can do about it, and this tab holds whatever you type for the entire horizon. Promotional rates make it worse: many of the best-advertised numbers apply for three or six months to new money only, then quietly revert to something ordinary.

What the arithmetic leaves out sits around the edges. Deposit insurance is not modelled, and it has limits — CDIC covers $100,000 per category per institution in Canada, FDIC $250,000 per depositor per bank in the United States. Nor are monthly fees, minimum balance requirements, or the transfer delays that decide whether money is actually available the day you need it. And interest is taxed as ordinary income rather than at capital-gains rates, while the tax setting here applies the capital-gains treatment, so like Bonds, this tab flatters the outcome.

None of which is an argument against holding cash. It is an argument for knowing what the balance is for, and for not mistaking a safe number for a growing one.

One pot split four ways, with each share compounding at its own rate. Contributions are divided by your target split rather than dropped into whichever sleeve is largest, and withdrawals come out of all four in proportion to what each is worth at the time, so drawing money down does not quietly change your allocation. If the sliders do not add up to 100%, nothing is left uninvested — the split is scaled proportionally, and the tab tells you so.

The blended return shown in the results is a weighted average of the four rates you set, and it is the least useful number on the page. It cannot tell you anything you did not already put in. Everything worth knowing here is about what happens to that average on the way.

Rebalancing is the real subject. Once a year, the tab sells whatever grew and buys whatever lagged, back to your target. That usually ends up slightly lower than leaving it alone, because in a long stock run you are trimming the winner every year — this is the cost, and the tab shows it rather than presenting rebalancing as free. What you get back is a narrower range of outcomes. Switch rebalancing off, and the tab reports what your mix has drifted into by the final year, which is often a far more aggressive portfolio than the one you chose, arrived at without ever deciding.

The blend is tested two ways, and neither is flattering by design. The thousand-run fan reports where this mix lands in the worst tenth of futures against 100% stocks in its worst tenth — and it will sometimes report that the blend’s floor is lower, meaning the return you gave up cost more than the volatility you shed. The historical overlay runs your allocation through real years starting from a date you choose, against all-stocks over the same period. Start it in 2000 or 1972, and the blend usually looks wise; start it in a calm decade, and it looks expensive. Both are true, which is the point of being able to move the date.

What this does not do is model how the four move together. In the simulated fan, each is randomized independently, so no correlation exists at all — and whether that flatters the blend or undersells it depends on the decade you imagine. The historical overlay is the honest test because real years reflect whatever correlation occurred. Beyond that: no fund fees on any sleeve, no trading costs, and no tax on rebalancing, which in a taxable account means selling winners and realizing gains every single year. Real estate is left out on purpose — a mortgage makes it a leveraged position that a percentage slider cannot honestly represent. It has its own tab.

The whole thing is a month-by-month simulation rather than a formula. Each month, every live balance grows by its own interest, then every minimum payment is made, then whatever money is left attacks one debt. Minimums are always paid. The only question this tab asks is where the money above the minimums should go, which is the question people face.

The monthly budget is every minimum you’ve entered plus whatever extra you add, and it never shrinks. When a debt is cleared, its minimum stays in the pool and rolls onto the next target, so the second debt dies faster than the first, and the last one goes quickest of all. That rollover is most of why paying extra works as well as it does.

Avalanche attacks the highest-interest rate first and is mathematically optimal. It always settles the debt for less money. Snowball clears the smallest balance first, costs more, and finishes individual debts sooner, which is worth something the arithmetic cannot see. Both are shown so you can decide what the difference is worth to you. In many realistic scenarios, the two clear in the same month and differ only in the interest paid, and the calculator reflects that rather than manufacturing a winner.

One result is worth more attention than the rest. If a debt’s minimum payment doesn’t cover that debt’s own monthly interest, the balance grows no matter how long you pay it, and the tab flags that debt by name. At 19.9%, a $3,000 balance generates about $50 a month in interest on its own.

The invest-instead view compares paying down debt to putting the same money into the market. Paying off an 8% loan is a guaranteed 8% return; investing is a hoped-for return with real chances of being negative for years. The comparison treats the two as if they carried the same certainty. They do not, and that difference should weigh in favour of clearing the debt.

The minimum payment you enter stays fixed for the whole run, while a real card recalculates it as a percentage of a falling balance, which stretches a minimum-only payoff out much further than this shows. Rates are held constant, so promotional rates that expire and variable rates that move are not modelled. Nothing here accounts for late fees, penalty rates, balance transfers, consolidation, or what any of it does to a credit score. The fourth debt and any after it start at $3,000 at 19.9% with a $90 minimum, which is a placeholder for a typical card rate rather than a researched figure and is meant to be replaced with yours.

This compares buying a home with renting the same home and investing the difference. The owning side carries the mortgage, property tax and insurance, ongoing maintenance, and the cost of selling at the end. The renting side invests the down payment and the closing costs on day one, adds whatever renting saves each month, and pays its own renter’s insurance and a security deposit that sits in a landlord’s account earning nothing.

Major repairs are optional and kept separate from the maintenance percentage, because nobody replaces a roof in smooth annual instalments. A water heater, a furnace, a roof and a catch‑all bucket each carry their own cost and their own replacement cycle, so they land as single large bills in specific years, each inflated to the year it actually happens.

Tax is switched off by default. Turned on, it treats the two sides as the law actually treats them: a home you live in gets the primary residence exclusion — the first $250,000 of gain, or $500,000 filing jointly in the US, and no cap at all in Canada — while the renter’s portfolio gets nothing of the kind. That asymmetry is real, it is large, and it favours buying.

Some of what decides this is not arithmetic at all. Buying is a savings plan you cannot skip: the payment leaves your account whether or not you felt like saving that month, and each one buys a little more of the house. The payment is fixed while rent is not, so owning tends to look better the longer you stay and the higher inflation runs, and in most places the gain on a home you live in is taxed lightly or not at all. You decide what happens to the place, you can stay as long as you like, and you get to belong somewhere. Against that: a house is slow and expensive to sell, every repair is yours, it holds you in one city, and it is a single asset on a single street bought largely with borrowed money — which multiplies the gain and the loss in equal measure.

Renting buys freedom instead. You can leave at the end of a lease, it costs far less to start, the repairs are somebody else’s problem, and your money stays liquid and spread across hundreds of companies rather than sunk into one building. Against that: you build no equity, the rent keeps climbing for as long as you pay it, and somebody else decides whether you stay. And the whole case rests on one condition the numbers quietly assume and most people never meet — that the difference actually gets invested, every month, for decades. The “surplus actually invested” slider exists to test that. It reduces the amount invested on whichever side is investing, and which side that hurts depends on which one has more left over each month.

Defaults are roughly 2026 US medians: a $410,000 price, a 6.5% mortgage and $2,000 rent. Your settings save in your browser, so changing a default here never disturbs numbers you have already entered — only “Reset this tab” brings in new ones.