https://retirementscenario.com/ Product updates, engine fidelity changes, and retirement-planning guides from RetirementScenario.com. Every change to the math is dated and described — nothing changes silently. en-us Sat, 19 Sep 2026 12:00:00 +0000 https://retirementscenario.com/changelog#e-2026-09-19-fixed-the-results-verdict-no-longer https://retirementscenario.com/changelog#e-2026-09-19-fixed-the-results-verdict-no-longer Sat, 19 Sep 2026 12:00:00 +0000 Product updates Fixed: the Results verdict no longer tells guardrails users to spend more. For plans with Guyton-Klinger spending guardrails enabled, a success rate of 95% or higher showed the verdict “You have room to spare — consider retiring earlier or spending more.” Under guardrails that advice points the wrong way: the high success rate comes from spending trimming itself in bad markets, so raising the spending baseline makes those trims fire more often and cut deeper while the headline number stays high. Guardrails plans at that level now read “Your guardrails hold — spending bends in rough markets, so the money lasts.” Plans without guardrails are unchanged. No numbers changed — only the verdict sentence. https://retirementscenario.com/changelog#e-2026-09-17-fixed-smart-moves-no-longer-calls https://retirementscenario.com/changelog#e-2026-09-17-fixed-smart-moves-no-longer-calls Thu, 17 Sep 2026 12:00:00 +0000 Product updates Fixed: Smart Moves no longer calls a beaten retirement goal “short of your goal.” In the Smart Moves retire-earlier view, when a plan on its own didn’t support the goal age but the recommended moves reached an age earlier than the goal, the verdict sentence still read “N years short of your goal.” The two ages shown were always computed correctly — only the sentence describing them was wrong, in the discouraging direction. It now says the moves put you ahead of your goal, and a result landing exactly on the goal is described as reaching it. No numbers changed. https://retirementscenario.com/changelog#e-2026-09-12-improved-pension-survivor-percentages-take-any https://retirementscenario.com/changelog#e-2026-09-12-improved-pension-survivor-percentages-take-any Sat, 12 Sep 2026 12:00:00 +0000 Fidelity & methodology Improved: pension survivor percentages take any value, and the yearly-increase field explains itself. The survivor election on each pension was a fixed menu — 0, 50, 75, or 100 percent. It’s now a direct entry: a pension that continues 40% to the survivor is entered as exactly 40. The engine already computed any percentage; only the control limited the choices. The Yearly increase field under the pension COLA checkbox also now states what each state means where you enter it: blank rises with your plan’s inflation rate, a percent compounds, a dollar amount steps without compounding, and unchecking the COLA box means no increase at all. Nobody’s saved numbers change — existing survivor elections keep their value. https://retirementscenario.com/changelog#e-2026-09-11-new-play-your-plans-story-your https://retirementscenario.com/changelog#e-2026-09-11-new-play-your-plans-story-your Fri, 11 Sep 2026 12:00:00 +0000 Fidelity & methodology New: Play your plan’s story — your whole plan, told in about a minute. A new button on Results plays the plan as a short film: the build, then your milestones one date at a time — the day you retire, the day your spouse retires, Medicare, each first Social Security check, the day required withdrawals begin — each frame carrying the money picture at that moment, read straight from the projection. Life events appear on their dates, every plan gets its own reel, and the film ends the only way we know how: the fan of 1,000 simulated futures opens around the path you just watched, and your success rate arrives as the verdict. The AI Advisor narrates the film for your plan — the connective lines are written to your milestones while every date and dollar stays computed, never AI-generated. Free for everyone. https://retirementscenario.com/changelog#e-2026-09-10-improved-a-pensions-yearly-increase-can https://retirementscenario.com/changelog#e-2026-09-10-improved-a-pensions-yearly-increase-can Thu, 10 Sep 2026 12:00:00 +0000 Fidelity & methodology Improved: a pension’s yearly increase can also be a fixed dollar amount. A reader of the morning’s entry pointed out a third pension shape: plans that add the same dollar amount every year, set once as a percentage of the first year’s check and never recomputed. That’s different from a compounding percentage — at 2%, compounding overstates such a pension by about 6% after twenty years — so the Yearly increase control now takes either form: a percent that compounds, or a fixed dollars-per-month bump added each year. Saved plans are untouched unless the new field is used. https://retirementscenario.com/changelog#e-2026-09-10-new-a-pension-can-increase-at https://retirementscenario.com/changelog#e-2026-09-10-new-a-pension-can-increase-at Thu, 10 Sep 2026 12:00:00 +0000 Fidelity & methodology New: a pension can increase at its own fixed rate, not just CPI. Plenty of pensions raise the cheque by a rate written into the plan document — a fixed 2% a year is the CalPERS-style classic — and until now the engine could only model “tracks inflation” or “never rises.” With the COLA checkbox on, a Yearly increase field now takes your pension’s own rate, and every projection compounds the payment at exactly that contractual rate regardless of what inflation does — including the honest part, where a fixed 2% raise slowly loses ground in a 3% world. Existing plans are untouched unless the field is used; both spouses’ pensions have the option. https://retirementscenario.com/changelog#e-2026-09-09-improved-the-temporary-6000-senior-tax https://retirementscenario.com/changelog#e-2026-09-09-improved-the-temporary-6000-senior-tax Wed, 09 Sep 2026 12:00:00 +0000 Fidelity & methodology Improved: the temporary $6,000 senior tax deduction (2025–2028) is now in the math. Federal law added a temporary deduction for people 65 and older — $6,000 per person for tax years 2025 through 2028, shrinking for incomes above $75,000 single / $150,000 married and gone entirely above $175,000 / $250,000. It stacks on top of the existing 65-plus standard deduction, which was already modeled. The projection now applies it in exactly those calendar years, per qualifying person. Who’s affected: anyone 65 or older during 2026–2028 with income under those caps — projected federal tax for those years was overstated, typically $900–$1,950 a year for a household in or below the phase-out range; those tax rows drop and long-run balances rise slightly. Plans under 65 through 2028 are unchanged to the dollar. The deduction expires after 2028 and the projection drops it then automatically. One deliberate boundary, disclosed in the app: Roth-conversion room sizing still excludes it — slightly conservative for a 65-plus converter through 2028. https://retirementscenario.com/changelog#e-2026-09-09-new-restore-your-plan-from-a https://retirementscenario.com/changelog#e-2026-09-09-new-restore-your-plan-from-a Wed, 09 Sep 2026 12:00:00 +0000 Product updates New: restore your plan from a backup file, right on the opening screen. Your plan lives in your browser — that’s the privacy promise — so a cleared cache or a different browser means starting at the welcome screen. Until now the only way back to your own backup file was at the bottom of the Inputs tab, behind loading an example plan first. The opening screens now offer it directly: pick your exported file and you land straight in your plan. It accepts a full backup or a single exported scenario, same as the Inputs tab. Suggested by a user who hit exactly this every time they switched browsers — thank you; this one’s yours. https://retirementscenario.com/changelog#e-2026-09-07-new-tell-us-your-birthday-and https://retirementscenario.com/changelog#e-2026-09-07-new-tell-us-your-birthday-and Mon, 07 Sep 2026 12:00:00 +0000 Fidelity & methodology New: tell us your birthday and the calendar snaps into place — exact RMD timing, an honest first Social Security year, and an age that updates itself. The app used to know only your age, and an age alone can place your birth year — and the calendar with it — one year off if your birthday hasn’t happened yet this year. At exactly the wrong boundary: someone born in December 1959 read as 1960 and had required withdrawals scheduled two years late. You can now add your birthday (month and year — no day needed) beside your age, for you and your spouse. Required-withdrawal timing lands on the true IRS schedule, every projection year carries the right calendar label, and the first Social Security year is credited honestly — checks start the month after your birthday, so a November birthday collects one check that year, not twelve. With a full birthday saved, your age computes itself: a plan reopened next year simulates the age you actually are. Everything stays optional — give nothing and your plan behaves exactly as before, to the dollar. Reported by an independent reviewer, down to the worked examples now in our test suite. One boundary is deliberately unchanged for now and disclosed in the app: the 59½ and 70½ half-year rules still work in whole years. https://retirementscenario.com/changelog#e-2026-09-07-improved-inflation-now-varies-inside-the https://retirementscenario.com/changelog#e-2026-09-07-improved-inflation-now-varies-inside-the Mon, 07 Sep 2026 12:00:00 +0000 Fidelity & methodology Improved: inflation now varies inside the Monte Carlo — your rate becomes the center of a realistic range, not a constant. Every one of the 1,000 simulated futures now carries its own inflation path: persistent year to year — a spike is followed by elevated years, not an instant return to normal — centered on the rate you entered as the typical year, and leaning against the same year’s market return, so the future where prices rise into a falling market is genuinely in the sample. Previously inflation was held at your stated rate in every simulation, which quietly flattered plans whose income can’t follow prices. The model is fitted from 75 years of actual inflation history — the same series the Historical Back-Test replays — and the constants and their derivation are published in the methodology. Who’s affected: success rates move — across our 48 test plans, about half moved down (typically one to three points, up to about six for strong plans holding lots of cash or income without cost-of-living adjustments), a third were unchanged, and a few thin-margin plans ticked slightly up. Nothing you entered changed, and the year-by-year projection table still runs at your stated rate — one legible future you can check by hand. https://retirementscenario.com/changelog#e-2026-09-07-improved-every-control-in-the-app https://retirementscenario.com/changelog#e-2026-09-07-improved-every-control-in-the-app Mon, 07 Sep 2026 12:00:00 +0000 Product updates Improved: every control in the app now announces its name to assistive technology. Screen-reader and voice-control users previously met unnamed sliders, dropdowns, and inputs — most fields on the Inputs tab among them. Every control now carries the same name its visible label shows. Fixed in the same pass: asking the AI Advisor to change your plan goal or a contribution escalation setting looked accepted but silently didn’t apply — those changes now take effect. https://retirementscenario.com/changelog#e-2026-09-06-fixed-the-ai-advisor-setup-conversation https://retirementscenario.com/changelog#e-2026-09-06-fixed-the-ai-advisor-setup-conversation Sun, 06 Sep 2026 12:00:00 +0000 Fidelity & methodology Fixed: the AI Advisor setup conversation could write a different retirement age than the one you stated. A user who opened the AI Advisor–guided setup by stating their retirement age — “retiring in two years at 58” — could get a final plan carrying a different age (62 in the reported case), and every downstream analysis then read the wrong date. The conversation’s instructions now require every figure you state to be copied into your plan exactly as you said it, and a permanent test opens with that user’s exact phrasing and fails if the stated age doesn’t survive to the finished plan. Reported and fixed the same day. If a plan of yours came out of the AI Advisor conversation with a number you didn’t say, correcting that field on the Inputs tab recalculates everything — nothing else in the plan is affected. https://retirementscenario.com/changelog#e-2026-09-06-fixed-a-social-security-portal-value https://retirementscenario.com/changelog#e-2026-09-06-fixed-a-social-security-portal-value Sun, 06 Sep 2026 12:00:00 +0000 Fidelity & methodology Fixed: a Social Security portal value alone now counts — entering only your age-70 or age-62 benefit no longer models $0. The Social Security fields take your benefit at full retirement age plus optional age-62 and age-70 values from your SSA account. Entering only a portal value — common if you plan to claim at 70 and that’s the number you trust — previously modeled $0 Social Security for the entire plan: the math anchored on the full-retirement-age field and ignored the other two when it was blank. Any one of the three figures is now enough; the engine derives the rest with the SSA formula, and entering only your age-70 value with a claim age of 70 uses exactly the number you typed. Reported by an independent reviewer. Affects plans that entered a portal value but left the full-retirement-age field blank — those plans showed $0 Social Security and a warning. Plans with the full-retirement-age benefit entered are unchanged to the dollar. Two projection labels also now say what the engine was already doing correctly: the year a household projection models a surviving spouse, the Social Security line is labeled as the survivor benefit, and the “RMD active” marker names whose schedule triggered it when an older spouse’s required withdrawals begin first. https://retirementscenario.com/changelog#e-2026-08-30-fixed-required-minimum-distributions-now-run https://retirementscenario.com/changelog#e-2026-08-30-fixed-required-minimum-distributions-now-run Sun, 30 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: required minimum distributions now run on each spouse’s own schedule. Household plans previously started RMDs for all pre-tax money on the primary plan-holder’s birth-year schedule (73 or 75). For couples with an age gap that was wrong in both directions: an older spouse’s real-world RMDs begin on their own schedule — potentially years earlier than the projection showed — and a younger spouse’s share was forced early. Reported by a user whose spouse, born 1958, starts at 73 while the projection showed 75. Each spouse’s share of the household pre-tax balance now follows that spouse’s own start age and their own IRS divisor each year; the Roth conversion window ends before the household’s first RMD year; a full-RMD charitable QCD covers your own required amount. Who’s affected: households where spouses differ in age or birth cohort — success rates barely move, but forced taxable income timing shifts ending balances either way. Same-age couples and single plans are unchanged. Details in the methodology’s RMD section. https://retirementscenario.com/changelog#e-2026-08-29-new-a-partner-program-and-the https://retirementscenario.com/changelog#e-2026-08-29-new-a-partner-program-and-the Sat, 29 Aug 2026 12:00:00 +0000 Fidelity & methodology New: a partner program — and the referral cookie that credits it, disclosed. Bloggers and reviewers who send readers our way now earn a commission on purchases. To credit them, both sites include a small script from Rewardful, our affiliate platform. Arrive through a partner’s link and a cookie stores a random referral ID for 90 days so that partner is credited if you buy; it identifies the link you clicked, never you, and carries nothing you’ve entered. Arrive any other way and nothing is stored. The privacy policy on both sites now lists Rewardful alongside our other processors. https://retirementscenario.com/changelog#e-2026-08-27-new-repeating-life-events-can-skip https://retirementscenario.com/changelog#e-2026-08-27-new-repeating-life-events-can-skip Thu, 27 Aug 2026 12:00:00 +0000 Fidelity & methodology New: repeating life events can skip years — a new car every 7 years is one entry. This one came straight from a user: their scenario needed a new car every 7 years, and the Life Events cards could only repeat a cost every single year — so it took five separate entries, or got left out. The cards now ask the natural first question — is this a lump sum or a payment? A lump sum is a cost at an age, once or repeating every few years, and the card lists the actual purchase years, each priced in the year it happens. A payment is an amount per year over a stretch — a loan, tuition — where “through age” is the last payment. Works for windfalls too, and every entry you already have lands in the right kind automatically with identical numbers. The AI Advisor can now set up repeating entries for you, it preserves an entry’s repeat and fixed-amount settings exactly when it edits your lists, and both lists grew from 10 entries to 15. The math is identical to entering the same years by hand — locked as an automated test, to the dollar — and every-year repeats and one-time entries are unchanged. https://retirementscenario.com/changelog#e-2026-08-26-new-divorce-your-retirement-a-private https://retirementscenario.com/changelog#e-2026-08-26-new-divorce-your-retirement-a-private Wed, 26 Aug 2026 12:00:00 +0000 Fidelity & methodology New: Divorce & Your Retirement — a private workshop for modeling a gray divorce. Some retirement questions are easier to explore than to say out loud, and this one was built for exactly that. From the Gray Divorce card on the Stress Test tab, the workshop models a split at any age against your real plan, side by side with staying: how each account divides, what happens to the house (four options, priced from your own home card), support paid or received, and your single-life spending — with single tax filing, health-coverage loss, and divorced-spouse Social Security (marriages of ten years or more) modeled on the SSA’s actual rules. Nothing you enter there is ever saved: no browser storage, no analytics, no trace in your plan — closing the workshop erases it. Three free searches show what would restore the household plan’s security; the Insulate search (part of Navigator) finds the change today that makes both futures sustainable at the app’s 85% bar. Full math in the app’s How It Works §37. https://retirementscenario.com/changelog#e-2026-08-26-fixed-the-rule-of-55-72t https://retirementscenario.com/changelog#e-2026-08-26-fixed-the-rule-of-55-72t Wed, 26 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: the Rule of 55 / 72(t) penalty switch and the QCD switch were unreachable for most plans. The switch that removes the 10% early-withdrawal penalty when a plan will use a legal early-access route (Rule of 55, a 72(t)/SEPP schedule, or a Roth conversion ladder) — and the switch that sends an RMD to charity as a QCD — rendered only inside a panel that appears for households above the Roth IRA income limit. For every other plan, neither control existed anywhere in the app. Both now live in their own “Withdrawals & Giving” group in the app’s Tax section: the penalty switch for any plan retiring before 60, the QCD switch for any plan with pre-tax savings that reaches age 70½. No number changed — the penalty was always modeled correctly, and QCDs computed correctly where enabled; the controls were simply unreachable. The AI Advisor also now sees whether a plan asserts a penalty-free strategy. https://retirementscenario.com/changelog#e-2026-08-24-fixed-the-stress-test-tab-could https://retirementscenario.com/changelog#e-2026-08-24-fixed-the-stress-test-tab-could Mon, 24 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: the Stress Test tab could crash when opened before its simulations finished. Opening Stress Test in the first seconds after loading the app — while the plan’s simulations were still computing — could crash the page to an error screen instead of the scenario grid. It depended on load timing, so it appeared random, and a reload usually recovered. The cause was in how the page was assembled, not in the math — no number was ever wrong. The tab now waits out the computation and renders normally. https://retirementscenario.com/changelog#e-2026-08-24-changed-the-compare-radar-now-measures https://retirementscenario.com/changelog#e-2026-08-24-changed-the-compare-radar-now-measures Mon, 24 Aug 2026 12:00:00 +0000 Fidelity & methodology Changed: the Compare radar now measures every scenario against one ruler. The “spend more” axis on the Compare tab’s radar used to be measured against each scenario’s own monthly goal, so comparing scenarios with different goals plotted them on different rulers — a small real difference in affordable spending could look like a large one purely because the goals differed. Every scenario is now measured against the first scenario’s goal, so the shapes are directly comparable; the legend says so, and the table below the chart still shows the absolute dollars. Same-goal comparisons are unchanged. Also in this release: when income sits within a few percent of the ACA subsidy cliff, the app now notes the year convention it prices the cliff on and points to healthcare.gov that close to the line; and the “retire earlier” goal color changed from pink to cyan. https://retirementscenario.com/changelog#e-2026-08-24-fixed-sharing-a-plan-failed-silently https://retirementscenario.com/changelog#e-2026-08-24-fixed-sharing-a-plan-failed-silently Mon, 24 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: sharing a plan failed silently if it contained a curly apostrophe or other special text. Share links are built from your plan’s own text — future expense and windfall names, and the note you can attach to the link. If any of that contained a character like the curly apostrophe an iPhone types automatically (“Daughter’s wedding”), generating the link failed silently: no error, no link, and the Share button could stick in its working state. Links now carry any text exactly as typed, and links created before this fix keep working unchanged. Affects anyone who tried to share a plan containing such text — most commonly text typed on a phone. Also in this release: when the step-by-step setup estimates your Social Security from an assumed salary because you left the field blank, it now says so — and the contributions step states what is assumed for any field you skip. Nothing about how those assumptions are used changed; they are simply disclosed. https://retirementscenario.com/changelog#e-2026-08-23-fixed-plans-built-by-describing-your https://retirementscenario.com/changelog#e-2026-08-23-fixed-plans-built-by-describing-your Sun, 23 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: plans built by describing your situation to the AI Advisor could carry the wrong spending, filing status and goal. If the conversation never captured a monthly spending figure, the plan was saved with a spending goal of zero. The projection then modelled almost no spending at all and reported a success rate near 100%. Anyone who built a plan this way and did not state what they expect to spend saw a result that was far too optimistic. The spending goal now falls back to the app’s default, and the figure is shown on the summary screen before you continue. Three related fixes to the same hand-off: if you are married but only wanted your own accounts modelled, that choice could not be recorded, so the plan was built either as a single filer — using single tax brackets on a jointly-filed return — or as a household with a spouse the app had no figures for, understating the plan by roughly 9 and 14 percentage points of success rate on a representative case; the retirement goal picked during the conversation was collected and then dropped; and the summary could describe your savings as split by a default 80/20 rule when the plan had actually used the balances you gave, or the reverse. Plans built through the step-by-step questions were not affected by any of these. https://retirementscenario.com/changelog#e-2026-08-23-fixed-two-figures-the-ai-advisor https://retirementscenario.com/changelog#e-2026-08-23-fixed-two-figures-the-ai-advisor Sun, 23 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: two figures the AI Advisor quoted were wrong — the Social Security break-even age and the projected RMD. The Advisor was given a Social Security break-even age calculated with an arithmetic error that made it about twelve times too soon. On a $3,000 full-retirement-age benefit claimed at 67, it was told the break-even against claiming at 62 was age 68; the correct answer is about 79. Delaying Social Security therefore looked as though it paid for itself almost immediately, when the break-even in fact sits near life expectancy — which is the substance of the claiming decision. The projected required minimum distribution grew the pre-tax balance from retirement to RMD age without subtracting anything spent along the way, so it returned the same number regardless of withdrawals: on one plan it reported $122,867 where the projection’s own figure was $69,130. Both now come from the projection itself. Neither figure appeared anywhere on screen, and no projection, success rate or balance in your plan changed — they were only ever used in conversation with the AI Advisor. https://retirementscenario.com/changelog#e-2026-08-23-fixed-compare-now-names-every-input https://retirementscenario.com/changelog#e-2026-08-23-fixed-compare-now-names-every-input Sun, 23 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: Compare now names every input that differs between two scenarios. The “Δ What Changed Between Scenarios” section on the Compare tab tracked a short list of inputs. If the only difference between two scenarios was an input outside that list — a spouse’s retirement age, for example — the section did not appear at all, so a comparison could show a large difference in success rate with nothing named as its cause. It now covers every input the Progress tab already tracked. Affects anyone comparing saved scenarios. On the same tab, the Couples Social Security map now states that its shading and its ★ are measured at different resolutions: the ★ marks the winner scored at full resolution, while the shading is a faster read across all 81 combinations, so a near-tie can look level while the ★ still separates them. The figures themselves are unchanged. https://retirementscenario.com/changelog#e-2026-08-23-fixed-progress-no-longer-describes-a https://retirementscenario.com/changelog#e-2026-08-23-fixed-progress-no-longer-describes-a Sun, 23 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: Progress no longer describes a plan that runs out of money as more predictable. When a plan depletes, every projected path ends at zero, which collapses the range of outcomes. The Progress tab read that collapse as good news and reported that outcomes had become more predictable. It now names it as depletion. Affects anyone whose plan reached zero between check-ins. Separately, the lens radar on Compare and Progress described all five of its axes as absolute. Four are. The spending axis is measured against your own spending goal, so raising your goal can move that axis down even as the spending you can afford goes up. Both legends now say so and point to the figures in dollars. No scores changed. https://retirementscenario.com/changelog#e-2026-08-22-fixed-the-roth-conversion-figure-now https://retirementscenario.com/changelog#e-2026-08-22-fixed-the-roth-conversion-figure-now Sat, 22 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: the Roth conversion figure now reads your real bridge-year income, and says it is a ceiling. The Roth conversion figure — shown on the Inputs tab and the Roth Strategies card, and used by the AI Advisor — estimated your taxable income during the bridge years instead of reading it from your projection. The estimate assumed your spending was withdrawn from your pre-tax accounts in proportion to your account balances. For a plan funded from an earmarked bridge reserve, or one with a still-working spouse or interest on cash set aside, that estimate could be off by tens of thousands of dollars a year, in either direction. It now reads the ordinary income the engine actually realises in each bridge year, and reports the largest amount you could convert in every year of the window rather than in an average year. Two limits were added: the recommendation now stops at the ACA subsidy cliff and says so when that is what constrains it, and the Medicare income-related surcharge now applies only from age 63, the first year whose income can reach a Medicare year under the two-year lookback. The wording changed too. Every surface used to call this number a recommendation; it is a ceiling — the most you could convert before crossing a limit — and being able to convert that much is not a reason to. On some plans converting nothing leaves the most behind, which is the question the Bridge Optimizer answers. The screens now say which is which. The amount changes on most plans — up where the old estimate overstated your bridge income and suppressed the suggestion, down where it missed the subsidy cliff. Nothing recalculates on its own: if you applied the earlier figure, that conversion is still in your plan and still drives your projection. The Inputs tab now shows how far your entered amount sits from the ceiling, and which year sets it. https://retirementscenario.com/changelog#e-2026-08-22-fixed-the-ss-income-gap-card https://retirementscenario.com/changelog#e-2026-08-22-fixed-the-ss-income-gap-card Sat, 22 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: the SS Income Gap card now includes the tax in “total portfolio draws”. The SS Income Gap card reports what your portfolio has to carry between retiring and claiming Social Security. Its headline counted the spending you need to cover, but not the tax you withdraw in order to cover it — and withdrawing from a pre-tax account means pulling both. On a plan funded mainly from a 401(k) or traditional IRA the figure was understated by roughly 16–19% from the second bridge year onward. It now shows the withdrawal the projection actually makes, so the card, the Projection tab and the year-by-year table below the card all report the same number. The headline also states that it is in today’s dollars, which the expanded panel already did and the collapsed card did not. The figure on this card goes up — your plan has not got worse, the card was reporting less than the plan always withdrew. No projection, success rate or balance changes. Reported by a subscriber who could not reconcile the card against the projection chart. https://retirementscenario.com/changelog#e-2026-08-22-clearer-the-moneyflow-diagram-names-the https://retirementscenario.com/changelog#e-2026-08-22-clearer-the-moneyflow-diagram-names-the Sat, 22 Aug 2026 12:00:00 +0000 Fidelity & methodology Clearer: the money-flow diagram names the Roth conversion tax inside a 401(k) withdrawal. On the Projection tab, “Where your money flows” shows what left each account in a given year. In a year you convert to a Roth, the tax on that conversion can be withheld from the converted money — so the 401(k) figure legitimately runs well above the spending draw shown elsewhere, with no explanation of the difference. The diagram now names it, the same way it already names a required minimum distribution: “401(k) — incl. $43k Roth conversion tax.” Labelling only — no figure in the diagram moves, and the flows balanced correctly before and after. https://retirementscenario.com/changelog#e-2026-08-22-fixed-six-more-places-a-screen https://retirementscenario.com/changelog#e-2026-08-22-fixed-six-more-places-a-screen Sat, 22 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: six more places a screen disagreed with the plan behind it. A second sweep covered ground the first missed — the conversational setup, the Compare and Progress tabs, and the paid Deep Search tools. As before, the underlying plan was correct in every case; what was wrong was what was shown or stored on top of it. Two affect anyone who used the conversational setup: a described home sale always had its proceeds earmarked as a bridge reserve whether or not you said so, and clicking “Edit Details” then pressing Continue through the questions quietly rebuilt a household plan as a single-person one, dropping a spouse’s pension, retirement accounts and Social Security. On Compare, a saved scenario whose money never runs out was recorded as lasting zero years, so the “longevity edge” card could name the plan that runs out of money as the one lasting longer. On Progress, a plan moving from a monthly surplus to a shortfall was described as “Up” by that amount rather than down, and the bridge-funding target counted years already past for anyone already retired. In Deep Search, Solve for Goal measured most of its search at a lower simulation count than its footnote claimed and interpolated across the mixture, which on the plan that surfaced it recommended cutting monthly spending about $500/mo further than the goal required. Your success rate, projection tables and saved plans are unchanged: none of these altered a calculation. https://retirementscenario.com/changelog#e-2026-08-22-fixed-eight-places-where-a-number https://retirementscenario.com/changelog#e-2026-08-22-fixed-eight-places-where-a-number Sat, 22 Aug 2026 12:00:00 +0000 Fidelity & methodology Fixed: eight places where a number on screen disagreed with the plan behind it. A sweep of the app found eight surfaces that recalculated a figure the projection had already worked out, and got it wrong. The underlying plan was correct in every case — the arithmetic running your simulations, projection tables and success rate was never affected. What was wrong was what some cards and tables displayed on top of it. The largest: the “Total nest egg” breakdown inside the Projected Nest Egg card could differ from that card’s own headline by as much as $1.15M on plans involving a home sale or purchase. Also corrected: the withdrawal rate shown against the 4% rule was understated for anyone already retired, and increasingly so the longer you had been; the Roth Conversion Window stress card had its condition reversed, so it was locked for the people who do have a gap between retiring and claiming Social Security and offered to those who don’t; the bridge-year ACA table left Other Income out of its estimate and could claim premium savings you would not receive; a pre-Medicare healthcare figure subtracted the ACA credit twice and could display as a negative amount; and for reverse-mortgage plans the banked part of a cheque now appears as its own column so the projection rows add up on screen. Two of these ran optimistic — the withdrawal rate and the ACA subsidy table — so a few plans looked safer on those specific readouts than they were. Your success rate, projection tables and saved plans are unchanged: nothing here altered a calculation, only what was printed from it.