Creating a Retirement Projection: A Step-by-Step Approach
Retirement planning can feel abstract until you try to translate it into months, dollars, and trade-offs you can actually live with. A solid retirement projection is not a spreadsheet exercise for its own sake. It is a decision tool. It helps you test whether your savings rate and investment mix can support the life you want, then shows you what to adjust when reality and assumptions collide. In practice, the best projections are built in layers. You start with what you know, you model the uncertain pieces conservatively, and you stress-test the results so you understand the range of outcomes. The goal is not a single magic number. It is a clear picture of “likely,” “tight,” and “not enough,” along with the levers that move the outcome. Step 1: Gather the numbers that do not move much Most people start with retirement accounts because that is where the numbers are easy to see. That said, the first inputs that matter are often the ones outside your brokerage app: your current spending, your income stability, and your obligations. Begin by collecting your last 12 months of household spending, ideally from bank and credit card statements. You are trying to separate spending into categories that behave differently over time. Some costs are sticky, like housing and insurance. Others are discretionary, like travel and dining. There are also costs that tend to rise, like healthcare and some utilities. When you later run scenarios, the categories help you decide what is flexible. Then gather your income details. Include wages, bonuses if they are regular, and any predictable income sources like Social Security estimates (if you have them), pension amounts, or expected annuity income. For planning, it helps to also note how likely your income is to change. If you are in a role that could be replaced by automation, or you expect a career shift, that affects the savings side of the projection as much as retirement assumptions affect the distribution side. Finally, take inventory of assets and accounts, not just balances. Different account types have different withdrawal behaviors due to taxes and required distributions. A retirement projection that treats everything as identical tends to mislead you, especially for households near tax thresholds. What you want at this stage is a baseline snapshot: Account balances, by type (taxable brokerage, traditional IRA or 401(k), Roth IRA or Roth 401(k), HSA if applicable) Current employment income and expected retirement date Current spending, with a realistic view of which categories can change There is value in doing this with a single person in charge of the story for the household. In one family I worked with informally, both partners tracked expenses, but each used a different definition of “housing” and “utilities.” The projection was accurate mathematically, but misleading conceptually. Once they aligned categories, their modeled spending matched reality much better, and the stress test stopped showing unnecessary “shortfalls” caused by accounting confusion. Step 2: Convert spending into a retirement budget that can survive reality A projection will only be as credible as its retirement spending assumptions. Many plans use “current spending minus a little” logic, but retirees do not usually spend less in a smooth, predictable way. Medical costs can jump, taxes can change, and some big expenses land right when you stop earning. Start by separating spending into buckets: Essential spending you would keep even in a lean year Lifestyle spending that is adjustable One-time or occasional expenses (vehicle replacement, home repairs, gifts, weddings, major travel) Healthcare and insurance, with a special line item for premiums and out-of-pocket costs To turn this into retirement dollars, ask a simple question: if you stop working, what changes first? In many households, spending shifts gradually in the first years after retirement. Some people keep a similar lifestyle while they travel more, then later slow down. Others cut travel immediately and use the early retirement years to reduce debt or renovate. The retirement budget should reflect your likely pattern, not an idealized one. If you want a practical rule of thumb, use your current spending as the base and then adjust with judgment: Reduce payroll tax withholding effects by modeling after-tax income needs instead of pretending the pre-tax number stays the same. Consider whether commuting-related costs disappear. Expect healthcare costs to behave differently than general inflation. Be careful with “inflation” as a single assumption. Most projections use a blended inflation rate, but retirees experience specific inflation pressures. Healthcare and insurance can move differently from food or housing. You do not need medical actuarial tables to be thoughtful. You do need at least one healthcare inflation assumption that can be higher than general inflation. Step 3: Decide on the timeline and withdrawal start date, not just a retirement age “Retire at 65” sounds straightforward, but the projection will change depending on when withdrawals actually begin, whether you bridge with part-time work, and how you handle early retirement. A projection should include at least three timing decisions: Retirement start year (when you stop earning active income) Withdrawal start year (when assets begin funding spending) Any major life events that affect cash flow, such as paying off a mortgage, selling a house, moving, or starting a business In real life, some households retire but still earn income for a while through consulting. Others stop work but delay withdrawals to let accounts grow. That timing can make a dramatic difference because compounding depends on how long money stays invested. When I see people get blindsided, it is often because they model withdrawal at the exact year they stop working, without accounting for severance, unemployment benefits, spousal income changes, or a reasonable delay before tapping retirement savings. It is worth building a timeline calendar-style. You do not need a complex schedule, but you should be explicit. Step 4: Model growth with reasonable ranges, not one neat assumption Investment returns are the most discussed assumption, but they are also the one people treat with the most false precision. Instead of choosing a single expected return and pretending uncertainty does not exist, use ranges. You can do this in a simple scenario approach. A robust projection typically uses: An expected return for each asset class or a blended portfolio assumption A conservative case, a baseline case, and an optimistic case A realistic sequence-of-returns risk assumption, meaning returns early in retirement matter a lot when withdrawals start Sequence risk is the reason two people with the same ending balances can experience very different outcomes. If markets drop right when withdrawals begin, the portfolio has to sell more shares at lower prices, and the remaining assets have less time to recover. A projection that only tests returns in the average sense can miss that. A practical way to incorporate sequence-of-returns risk without building a full Monte Carlo model is to run multiple scenarios that vary return paths rather than just average returns. Many planning tools allow “first year return” or early-period return shocks. If you are building your own spreadsheet, it is still possible to simulate a few “bad early years” by applying lower returns in the early withdrawal years and higher returns later, then comparing outcomes. The key is not whether you picked the perfect return path. The key is that you see how fragile your plan becomes under stress. Step 5: Add taxes and account behavior, or your “safe” plan may not be safe Taxes are where many retirement projections quietly fail. The math of portfolio growth might look fine, but the net cash available to spend can be lower than expected because taxes differ based on withdrawal sources and income levels. Account behavior matters: Traditional pre-tax accounts (traditional IRA, 401(k)) typically generate taxable income when withdrawn. Roth accounts provide tax-free withdrawals if qualified conditions are met, but the “don’t tax Roth” assumption is only true if you follow the rules. Taxable brokerage accounts include capital gains and possibly dividends, and taxes depend on holding periods and your realized gains. Required minimum distributions (RMDs) can force taxable withdrawals later in life, changing the tax picture even if your spending stays flat. A projection that ignores taxes can be off by a meaningful amount, especially for households with moderate-to-high income in retirement or those with large pre-tax account balances. You do not need to become a tax attorney, but you do need consistent rules: Estimate tax brackets using projected taxable income. Include federal and state assumptions if you know your likely state of residence. Model withdrawal ordering, because drawing from taxable accounts can realize capital gains in a way that drawing from pre-tax accounts does not, and vice versa. Plan for tax-smart withdrawal sequencing. Many families can lower lifetime taxes by pulling taxable accounts with losses strategically, delaying taxable gains realization, or using Roth conversions in a controlled way. Here is where judgment becomes important. If you are very close to a bracket edge, small changes in withdrawals, capital gains, or Social Security can move you into a higher tax bracket. A projection with a single tax assumption may overstate safety. In those cases, build a “tax sensitivity” scenario where you increase taxable income slightly and see how quickly the plan changes. Step 6: Include inflation assumptions, then validate them against your budget reality Inflation is not one thing, and retirees do not experience it uniformly. Your projection should include: General inflation for most spending categories Potentially different inflation for healthcare and insurance A plan for discretionary spending behavior, since people often cut “optional” categories first when finances tighten If you use a single inflation rate, you are simplifying. That is acceptable if you do not pretend it is precise. I often suggest choosing an inflation assumption that is slightly conservative relative to your personal long-term comfort. If you believe healthcare costs will outpace general inflation, reflect that. Validation helps. Before you lock the model, compare projected retirement spending growth against your lived history of spending inflation. If your household spending has been unusually stable due to long fixed-rate expenses, you may need to adjust down the inflation pressure for a few categories, but you still need a healthcare buffer. Step 7: Account for retirement income streams you might forget to model correctly Retirement income is more than “my 401(k).” Many households have additional sources that influence the withdrawal strategy and tax outcomes. Common examples include: Social Security (spousal benefits, delayed credits, survivor benefits) Pension income (if applicable, and whether it adjusts with inflation) Annuities (if you have them, or are considering them) Part-time or consulting income Rental income (and its expenses, like property tax and maintenance) Social Security is a particularly important variable because it can reduce the amount you need to withdraw from taxable accounts in earlier retirement years. That affects both taxes finance loans comparison and investment risk. If you model Social Security, be explicit about timing. Claiming at 62 versus 70 can change household cash flow for decades. The projection should show how your portfolio behaves under the different claiming ages, not just your total expected benefit. One thing to watch: some income streams have built-in inflation adjustments, some do not. Social Security benefits are generally adjusted, while many pensions vary by plan. Model what you actually have, not what you remember. Step 8: Add the “friction” costs that rarely make it into spreadsheets Retirement projections often assume a smooth conversion from working life to retirement spending. Real life has friction: Home repairs Insurance premium changes Car replacement or unexpected medical bills Family support that becomes more common with age Travel and hobby costs that rise as you have more time Market volatility that forces emotion-driven decisions Friction is not an “expense category” you can ignore. You can include it as a realistic buffer. For example, if your spending budget already includes home maintenance and healthcare deductibles, you might still add a “miscellaneous volatility” amount as a percentage of discretionary spending to avoid modeling a plan that only works on a perfect month. A buffer is also helpful because it protects against a very human behavioral risk: stopping the plan too early or overreacting during a market decline. If you have ever watched a friend try to cut spending aggressively after a portfolio drop, you have seen how friction can turn into a self-fulfilling problem. The projection should show you how much buffer you need to keep your withdrawal behavior stable when markets are bad. Step 9: Stress test the plan in plain language, then decide what you would change Once you have a baseline projection, stress it. The goal is to answer questions like: What happens if markets are weak for the first five years? What happens if inflation runs higher and spending does not adjust quickly? What if healthcare costs land higher than expected? What if you retire one year earlier or live longer than planned? This is where your projection becomes a conversation with your own future self. You are not hunting for catastrophe. You are identifying which assumptions you can afford to be wrong about. At this point, you should also set your “action rules.” When a projection shows shortfall, what do you do first? Reduce withdrawals by delaying retirement or spending cuts. Increase savings through part-time work or continued contributions. Adjust the portfolio risk level. Convert some pre-tax assets to Roth strategically if it is tax-efficient. Consider purchasing or using guaranteed income if you are already near the finish line. Different households will choose different levers based on temperament and constraints. If you hate market risk, you may prefer strategies that reduce volatility even if expected returns are lower. If you are willing to tolerate volatility, you can plan for a more aggressive allocation and longer time horizons. A practical projection workflow you can reuse Below is a simple workflow that mirrors how many competent planners actually work, adjusted for the reality that you may only have time for a few scenarios. Collect last 12 months of spending, categorize it, and estimate which parts are flexible. Build a timeline for retirement start and withdrawal start, including major life events. Create 3 scenarios for portfolio growth using baseline, conservative, and optimistic assumptions. Add taxes and withdrawal sequencing by account type, and model RMD timing if relevant. Stress test early retirement returns and a higher inflation or healthcare case. If you do only one thing with this workflow, make sure your spending assumptions are the anchor. It is easier to adjust withdrawals and saving rates than it is to “fix” a bad spending estimate later. Where projections often go wrong, and how to catch it early You can run a model with sophisticated growth and still end up with a wrong result if the inputs conflict with each other. Here are the most common failure points I have seen, along with how to spot them before you waste weeks. Mismatched “real” and “nominal” assumptions If your portfolio returns are nominal, your spending should also grow with nominal inflation. If your returns are real, spending should grow with real inflation assumptions. Mixing them creates a quiet error that compounds over time. A quick sanity check helps. If your model shows spending that barely increases over 30 years while returns assume meaningful inflation, something is off. Align the basis of your assumptions early. Forgetting required distributions Many people model withdrawals and end the projection at “my retirement years,” then do not extend far enough for RMDs to matter. In practice, RMDs can reshape your tax situation in your later years. If you plan to keep money in traditional accounts, model the RMD effect. Overlooking withdrawal ordering The order in which you tap accounts can matter as much as the allocation itself. Drawing from taxable accounts can generate capital gains tax, while drawing from pre-tax accounts can push income into a higher bracket. A projection that assumes withdrawals are proportionate across accounts may misstate taxes and net spend. Treating “expected return” like a guarantee A single expected return can create false confidence. You want to see the distribution of outcomes. If you cannot run Monte Carlo, use scenario analysis with early-year stress. Turning projection results into decisions, not just numbers A projection becomes useful when it produces choices. If your plan shows “enough,” you still have decisions to make. If it shows “tight,” you want to know what changes would close the gap without breaking your quality of life. Consider these decision patterns that show up repeatedly: If the plan is tight only in a conservative scenario, you might keep the same retirement age but adjust portfolio risk or reduce spending flexibility risk. In other words, you can plan to protect the downside without locking yourself into overly conservative returns that reduce long-term potential. If the plan is tight across both baseline and conservative cases, delaying retirement may be the cleanest lever. One additional year of work adds savings, reduces withdrawal years, and can reduce the need to sell assets during a market decline at the start of withdrawals. It is not always your favorite option, but it is often the most mathematically effective. If the plan is tight due to taxes, you may not need more money, you may need better sequencing. Roth conversions, tax-loss harvesting, or an optimized withdrawal order can reduce drag. These are not guarantees either, but they can make the same pre-tax portfolio produce more spending. The “trade-offs” matter. Increasing savings might reduce your lifestyle now. Reducing portfolio risk might reduce the probability of success in a long life scenario if returns are too conservative. A projection should surface those trade-offs so you can decide based on what you value, not based on spreadsheet convenience. A final check: does the projection reflect how you actually plan to live? The most accurate retirement projection still fails if it assumes you will behave differently than you will. That is why your projection needs at least one realistic behavioral assumption. Ask yourself: When markets drop early in retirement, will you keep withdrawals steady, or will you cut aggressively? Do you plan to downsize housing, or are you staying put? Will you travel more in early retirement, then slow down, or is your lifestyle preference stable? Are you likely to take on part-time work, or would that be stressful? How comfortable are you with tax complexity, like Roth conversions? This is not about predicting your future emotions perfectly. It is about making the model resilient to the way you respond under stress. If you are the type who panics during market declines, build your conservative scenario with more weight. If you are comfortable staying invested through volatility, you can justify a more balanced stance, but still keep a buffer in spending so you can avoid forced selling. Make it your own system, not a one-time spreadsheet Retirement projections should be updated as your life changes. A change in income, marriage status, health status, housing costs, or even your tolerance for risk can shift the plan materially. The best approach is to treat the projection like a living financial document. When you update it, focus on what changed: New account balances and contribution rates Updated spending based on the last few months, not just last year New retirement age decisions Changes in taxes, state residence, or expected income streams Then rerun the same baseline, conservative, and optimistic scenarios. Consistency makes it easier to see what matters and avoid being distracted by noise. A well-built projection gives you a map, not a promise. It shows you where you are likely headed, where you might get stuck, and which decisions can move you from “hope” to “prepared.” That is the real value of retirement planning in finance terms: it turns uncertainty into something you can manage with judgment. If you want, tell me your rough retirement age target, current savings balances by account type, and whether you expect pensions or rely mostly on Social Security and portfolio withdrawals. I can outline a scenario framework and the key assumptions to plug in, without pretending there is a single correct answer.