Most ten-year net worth projections fall apart the moment they meet a real decade. The spreadsheet says you should have $640,000 by 2036. The decade actually delivers two corrections, one job change, a roof replacement, and a stretch of inflation that quietly shaves a third off the purchasing power of the number on the screen. The forecast was not wrong because the math was wrong. It was wrong because the assumptions were optimistic in five different places at once.
Projecting net worth a decade ahead is a useful exercise. The version of it that helps you is the version that bakes in the things that will probably go sideways, not the version that compounds an 8 percent annual return on every dollar you own and calls it planning. This article walks through how to build a projection that is honest about returns, contributions, debt, taxes, and inflation, so the line you draw is one you can actually steer toward.
Photo by MART PRODUCTION on Pexels
Why most ten-year projections lie by accident
The default mistake is not lying on purpose. It is using a single optimistic number to stand in for five different uncertain numbers. A typical kitchen-table projection looks something like this: take today's net worth, assume 7 percent compound annual growth on everything for 10 years, add expected savings, and read off a number.
That model has four quiet flaws. The 7 percent assumes the entire balance sheet behaves like the S&P 500, which it does not. It ignores the fact that contributions are themselves uncertain (a job change, a year of high medical costs, a college tuition shock can all flatten contribution for a stretch). It treats debt as if it disappears in the background, when in fact mortgage principal and student loan amortization schedules have specific shapes that matter. And it does not adjust for inflation, so the future number is in dollars that buy less.
A more honest projection keeps each of those moving pieces visible, lets you stress-test them, and reports the answer in both nominal and real dollars.
Start from a clean current snapshot
You cannot project off a baseline that is fuzzy. Before any forward math, the present-day net worth needs to be a single defensible number, with assets and liabilities itemized in categories you can actually update each quarter.
A clean snapshot has, at minimum:
- Liquid assets. Checking, high-yield savings, money market, certificates within 12 months of maturity.
- Brokerage and taxable investments. After-tax positions you could realistically sell.
- Tax-deferred retirement. 401(k), traditional IRA, 403(b). Note the unrealized tax sitting on these accounts so you do not double-count the dollars.
- Roth retirement. Roth 401(k), Roth IRA, contributions and growth (no future tax owed).
- Real estate. Primary residence at a defensible market value (not Zillow, not the purchase price). Rentals or land at conservative current values.
- Other illiquid assets. Vehicles at honest resale value, private business equity at a defensible figure, collectibles only if you would actually liquidate them.
- Short-term liabilities. Credit cards, personal loans, lines of credit, anything under 24 months remaining term.
- Long-term liabilities. Mortgage balance, student loans, car loans by remaining term and rate.
The free net worth tracker by EvvyTools lets you punch all of these into the right buckets and produces the percentile, liquidity ratio, and debt-to-asset numbers automatically. If you would rather work in a spreadsheet, that is fine, as long as the categories above are kept distinct. The projection only works if the buckets are clean.
The current Federal Reserve Survey of Consumer Finances data is the simplest sanity check on whether your snapshot looks reasonable against your peer cohort. It does not change your number, but it gives you a defensible comparison point.
Pick a return assumption per bucket, not one number for the whole balance sheet
Here is where most projections quietly go wrong. The 7 percent annual return assumption is roughly defensible as a long-run average for diversified equities, but the average household balance sheet is not 100 percent diversified equities. It is some mix of cash, bonds, equities, real estate, and vehicles, each with its own expected return profile.
A more honest approach uses a different growth rate per bucket:
- Cash and short-term savings. Use the prevailing high-yield savings rate or short-term treasury yield. As of mid-2026, somewhere in the 4 to 5 percent range is realistic. In a different rate environment, use something different.
- Brokerage and taxable equities. A 6 to 7 percent long-run nominal expected return is defensible for a diversified portfolio, with the caveat that 10-year stretches can produce significantly higher or lower realized returns.
- Tax-deferred and Roth retirement. Same rate as the underlying portfolio composition. Most of these accounts are also equity-heavy, but if a 401(k) is in a target-date fund near retirement, the bond allocation pulls the expected return down.
- Real estate. Long-run residential real estate appreciation, net of capital improvements, has historically run a percentage point or two above inflation. Three to four percent nominal is a defensible base case for a primary residence in a non-supply-constrained market.
- Vehicles. Depreciate, do not appreciate. A car you bought new typically loses 15 to 25 percent of its value in year one and trails off from there.
- Other illiquid assets. Be skeptical. Collectibles and private business stakes are easy to overstate in a forward model.
The simplest version of the model multiplies each starting bucket by (1 + bucket_rate)^years and sums the results.
The Vanguard Principles for Investing Success library has a clear explanation of why a single return number for a mixed portfolio overstates the certainty of the forecast. The Bogleheads wiki carries a community-maintained version of the same idea with explicit math.
Photo by RDNE Stock project on Pexels
Project contributions with realistic gaps
Contributions are the lever you actually control, and they almost never run flat for 10 years. Real-world contribution patterns have shape: a few years of strong savings, a year or two of a major life expense that flattens contributions, often a return to normal afterward. Pretending the contribution number is constant overstates the projection.
A defensible contribution model has two parts:
- A base contribution rate. Whatever percentage of income you are saving today, by account. Express it as a dollar figure based on current income.
- An honest haircut for at least one disruption. In a 10-year window, the probability of a job change, a medical event, a family expense, or a housing transition that meaningfully cuts contributions for at least one calendar year is high. The model should bake in at least one such year at zero contribution.
This is not pessimism. It is acknowledging that 10 years is long enough for at least one disruption to occur and short enough that the disruption matters to the final number. The IRS contribution limits in IRS Publication 590 are the ceiling on tax-advantaged accounts, and they typically move up roughly with inflation. Build that into the model if you are planning to max out, but do not assume the ceiling rises faster than it has historically.
Amortize the debts on their actual schedules
Mortgage principal, student loan principal, and car loan principal all decline on known schedules. A 30-year mortgage with 22 years remaining will pay down principal slowly in years 1 through 5 of your projection and faster in years 6 through 10, because amortization is back-loaded toward principal. A 5-year auto loan finishes inside the projection window, freeing cash flow for the back half of the decade.
The mistake is treating the mortgage balance like a fixed liability that drifts down by some average percentage. It does not. It follows a schedule that you can pull from your lender's amortization table or generate from a standard formula. The shape matters for the projection because debt payoff frees cash that compounds at your bucket return rate for the remaining years.
The Consumer Financial Protection Bureau publishes example amortization tables and standard formulas. Plug your actual balances in. The output is more honest than guessing.
Adjust for inflation, in both directions
A 10-year nominal projection in 2036 dollars is not directly comparable to today's lifestyle. Inflation eats purchasing power on the asset side and reduces the real burden of fixed-rate debt on the liability side. Both effects need to show up.
A defensible model reports two numbers at year 10:
- Nominal net worth. The number the spreadsheet produces with no inflation adjustment.
- Real net worth in today's dollars. The nominal number divided by
(1 + inflation_rate)^10. A 2.5 to 3.5 percent annual inflation assumption is defensible based on long-run Bureau of Labor Statistics CPI data, with a wider band acknowledged for high-inflation or low-inflation outcomes.
The real number is the one to plan against. A nominal projection of $940,000 in 2036 dollars at 3 percent annual inflation is about $700,000 in 2026 dollars. That is still meaningful, but it is meaningfully less than the headline figure suggests.
Photo by Beyzaa Yurtkuran on Pexels
Stress test in three places
Once the base projection is built, three stress tests reveal whether the plan is robust or fragile.
- A flat decade. Set portfolio returns to 0 percent for years 1 through 5 of the projection. This is not a recession scenario, it is a sideways decade. Stocks have delivered roughly flat 10-year stretches before (2000 to 2009 in real terms is a frequently cited example). If the projection still produces a workable number, the plan is robust to return uncertainty.
- A contribution interruption. Set contributions to zero for two consecutive years somewhere in the middle of the projection. This simulates a job change, a family event, or a major expense. Look at the gap between this scenario and the base case. If the gap is small, the savings rate is doing the heavy lifting. If it is large, the plan is fragile to disruption.
- A real estate haircut. Reduce primary residence value by 15 percent at year 5 and let it recover linearly to the base case by year 10. Real estate is illiquid and locally cyclical. A national projection that depends on uninterrupted residential appreciation is not robust.
If the projection survives all three stress tests with the final number still in a livable range, the plan is in good shape. If any one of the three breaks it, that is where the additional savings, additional diversification, or additional debt paydown should be directed.
Update the projection every quarter, not every January
Annual projections drift. A year is long enough for the contribution assumption to be quietly stale, for the asset mix to have shifted from a market move, and for new debt or paid-off debt to have changed the liability side. By the time you re-run the model in January, you have been steering off an old map for months.
A quarterly cadence is enough to catch material drift without becoming a hobby. The tools directory carries the rest of the everyday calculators in this series if you want to pair the net worth tracker with a debt payoff or savings projection tool, and the EvvyTools blog walks through how each of these snapshots fits together.
A short checklist before you trust the line on the chart
- Snapshot is itemized by bucket, with real estate at a defensible value, not optimism.
- Each bucket has its own return assumption, not a blended 7 percent.
- Contribution model includes at least one zero year somewhere in the decade.
- Debts are amortized on actual schedules, not a flat percentage drift.
- Final number is reported in both nominal and inflation-adjusted dollars.
- Three stress tests have been run and the answer still holds.
A projection built this way will still be wrong in some specific detail. The point is not to be right. The point is to be wrong by a small enough margin that the plan still works, and to know which assumptions you would have to defend if the next decade does not cooperate.
