Portfolio Projection

Engine

• Paths per scenario (Monte Carlo overlay): 2,000, through the full tax-aware engine (Session B) — not the legacy single-bucket sampler
• Step: monthly (12 steps per year)
• Return distribution: normal within each holding's own regime, sampled via Box-Muller
• Per-step return: r = μ/12 + (σ/√12) × z, where z ~ N(0,1)
• Cross-holding correlation: a single shared market factor at ρ = 0.85 (z = √ρ × z_market + √(1−ρ) × z_idiosyncratic), loaded identically onto every included financial asset and super account, combined with each holding's own profile-based regime switching. A genuine per-asset-class correlation matrix (Australian vs international equity, property, fixed interest, cash each with their own factor loading) is a considered future refinement, not built: it would need per-class σ calibrated to reproduce each profile's firm-set total σ, and an incorrect calibration would silently contradict the firm's stated volatility assumptions.
• Balance update: B(t+1) = B(t) × (1+r) + contribution
• Stored snapshots: yearly only

Asset class assumptions

Click a profile (curve or legend) to isolate it — the others fade. Click again to restore. Hover along a curve to see the chance of a single year landing at or below that return.

Total returns (nominal) are the firm's proposed assumption set; income and growth splits are shown for reference — the simulation engine consumes total return only and does not model distributions or tax. Real returns are derived at simulation time from total nominal and the CPI parameter above via the Fisher relation, so changing CPI flows through every projection without a reload.

σ values and regime-switching parameters are model calibrations, not firm-published figures. Defensive-end σs (Cash 1.5%, Defensive 3.0%) are intentionally higher than nominal-terms intuition would suggest: real cash and short-duration variance are dominated by inflation variance rather than nominal price fluctuations.

Residential Property's σ is set higher than typical published direct-property volatility, which is understated by appraisal-based valuation smoothing; the model σ is adjusted upward to reflect underlying transaction-price variance.

Assets with a custom allocation use their entered income + growth return as the expected return, and borrow σ and regime-switching parameters from their selected volatility basis profile.

Franking percentages are derived, not stored — each is computed from the profile's own asset class weights (Australian equity share) and its stated income return, at a 4%-fully-franked Australian-equity yield assumption, so it can never disagree with the weights it comes from. The two Accelerated Growth profiles' income/growth return split looks inconsistent with their own class weights on this basis and is flagged in profiles.js for firm CMA review; the class weights themselves are not adjusted to fit. Consumed by the tax phase.

In couple households the projection timeline is anchored to the client's age; the partner ages alongside from their own current age. The projection ends at the client's end age with both alive throughout — no mortality modelling.

Real-terms convention

All amounts in this tool are expressed in real (today's dollars) terms. Profile expected returns are net of inflation — a 5% real return means the portfolio's purchasing power grows by 5% per year, not its nominal value.

Contributions and withdrawals marked as indexed maintain their real purchasing power throughout the projection — a $1,000/month contribution today represents the same purchasing power across all projection years, which is equivalent to that contribution growing with CPI in nominal dollars. A cashflow with indexing switched off is fixed in nominal dollars, so its real value declines by CPI each year.

Inflation assumption

The future-dollars display converts real (today's dollars) values to nominal future values using a deterministic inflation rate. The default rate is 2.5% per annum, consistent with the RBA's target band midpoint. Users can adjust this rate; the change flows through future-dollars display values, derived real returns, and non-indexed cashflow treatment, but the simulation itself stays in real terms.

Cashflow rows carry an index basis (None, CPI, or the wage index) plus an additional percentage. A CPI-indexed row is constant in real terms; a row with no indexation is fixed in nominal dollars and its real value declines by CPI each year; a wage-indexed row grows in real terms by the gap between AWOTE and CPI. Dates of birth are captured for the life-expectancy lookup — ages still tick over each 1 July per the plan-year convention, so DOB precision does not shift cashflow timing.

Inflation in this tool is deterministic — every projection assumes the same inflation rate. Real-world inflation varies stochastically; modelling stochastic inflation is out of scope for this version.

Today's dollars vs future dollars

The projection runs in real (today's dollars) terms. All inputs — contributions, withdrawals, starting balances, profile returns — are denominated in today's purchasing power.

The display toggle lets you view the same projection in either:

• Today's dollars. Values are shown in current purchasing power throughout the projection. This is the default. Easier to compare to current financial goals and spending levels.
• Future dollars. Values are shown in inflated nominal dollars at each future year, using the fixed inflation assumption above. Useful for understanding what the projection looks like in the dollars that will actually be circulating at that future date.

The underlying projection is identical regardless of which display mode is active. Only the displayed units change.

Return distribution and regime switching

The Monte Carlo engine models market returns with a two-state regime-switching variance model. At any given month, the market is in one of two states:

• Normal. Typical conditions with lower volatility.
• Stress. Occasional periods of elevated volatility.

The market transitions between states with small monthly probabilities, set so that stress periods occur every 7–10 years on average and last about 10 months when they occur. This captures the empirical observation that volatile periods in real markets cluster — the 2008 financial crisis, the 1970s stagflation, COVID-19, and similar episodes were not isolated bad months but coherent stressed regimes.

The mean return (μ) is the same in both states — the forward-looking expected return from the profile assumptions is preserved exactly. Only the volatility (σ) differs between states. The weighted average of normal-state and stress-state volatility reproduces the profile's long-run σ assumption.

This approach is "forward-looking on what returns are on average, backward-looking on how returns evolve through time." Expected returns and long-run volatilities come from forward-looking capital market assumptions; the time structure (clustering, stress duration, stress amplification) is calibrated to historical market behaviour.

Tax

The projection models resident individual income tax on each person's attributed income: income rows they own, distribution income from assets they own (joint assets split 50/50), less each asset's ICR as a deduction. Marginal brackets, the Medicare levy with shading-in (single thresholds per person), and the low income tax offset are applied per person per financial year. Franked distribution income is grossed up and the franking credit applied as a refundable offset — a person's net tax can be negative (a refund, which increases household cashflow).

For a person with employment income, PAYG withholding is estimated on that employment income alone (the tax-free threshold, Medicare levy, and low income tax offset only — ignoring deductions, other income, and franking credits, mirroring what an employer actually withholds) and debited from household cashflow in the months salary is actually paid. The gap between that estimate and the person's full actual tax liability (every income source, every deduction and offset) is assessed each financial year and settles as a single household outflow — a refund (increasing cashflow) or a balancing payment (decreasing it) — in July of the following financial year, the same convention as CGT and Division 293/296. A person with no employment income keeps the smooth accrual instead: their full tax is spread evenly across the months their income arises, within the same year, with net refunds credited in the final month.

Deductions (a Working Expense, Vehicle Deductions, Deductible Insurance Premiums, Social Club, Novated Lease, or Salary Packaging row, or an investment property's Depreciation field) reduce the owning person's assessable income in the same assessment as everything else. None of them are a household cashflow event by themselves — there is no linked income row to reduce at source the way salary sacrifice is, so entering a deduction here does not itself move cash. If the underlying spend also needs to leave household cash, enter a matching Expense row too.

Capital gains tax uses one pooled cost base per asset (no parcels). Contributions, one-off inflows, invested surplus, and reinvested distributions add to the pool; any sale — an explicit withdrawal, an outbound one-off, or a deficit-funding draw — consumes a proportional slice, and the difference is a realised gain or loss. Sales before 1 July 2027 apply the 50% discount (holdings are assumed to satisfy the 12-month rule except amounts added in the same financial year, which are treated as coming proportionally from new money). At 1 July 2027 every asset's pool resets to market value under the enacted deemed-reacquisition transition — gains accrued before that date are not taxed at that point at the level this tool models. From that date the pool is CPI-indexed (constant in real terms) and tax on gains is the greater of the marginal calculation and the 30% minimum, with capital losses carried forward per person against future gains only. CGT on a year's gains is paid as a single household outflow in July of the following year; the final year's assessment cannot be paid inside the projection and is reported as an accrued liability.

Each person's HELP/HECS balance (Setup) accrues a compulsory repayment each financial year on their repayment income — taxable income plus reportable super contributions (salary sacrifice and personal deductible only; employer Superannuation Guarantee is never included) plus net investment losses (property negative gearing; a leveraged share portfolio's loss is not separately tracked and so isn't added back — a disclosed simplification). Rates are marginal within each bracket except the top: at and above $186,052 (FY2026–27, indexed), the repayment is 10% of the WHOLE repayment income, not just the excess — a genuine cliff, not a smoothing artefact. The repayment is withheld through PAYG alongside income tax and settles the same way — a refund or balancing payment in July of the following year — and reduces the balance; the balance is held constant in real terms until then, and the loan ends once it reaches zero. Reportable fringe benefits and exempt foreign employment income are not modelled and so are not added to repayment income — a disclosed simplification. Indexation of the repayment-income thresholds is legislated to wage growth (WPI); this tool uses its existing AWOTE wage-index assumption as the closest available proxy, pending confirmation against the firm's reference figures.

The Medicare Levy Surcharge applies to each person without private hospital cover (Setup, defaulting to cover held) whose income for surcharge purposes — the same repayment-income figure used for HELP above — crosses a threshold: singles compare their own income against the single thresholds; couples compare their COMBINED income against the family thresholds (which shift up by $1,500, indexed, for each dependent child after the first — Setup's household-level Dependent children field), then each uncovered partner is charged on their OWN income at the resulting rate. Like HELP, this is a step function: crossing a threshold charges the WHOLE income at the new rate, not just the amount above it. It is withheld through PAYG and settles the same way as income tax and HELP. Reportable fringe benefits and exempt foreign employment income are not modelled and so are not added to the income tested — a disclosed simplification, matching HELP's.

Not modelled, by design: SAPTO, family Medicare thresholds beyond the surcharge's own family bands, and trusts or companies. Division 293 and superannuation's own tax treatment are modelled separately — see the Superannuation section. Assets marked as not CGT assets never generate capital gains tax.

Lifestyle assets (contents, vehicles, jewellery and similar) are treated as exempt personal-use assets: they grow at their simple rate and never generate income, deductions, or capital gains tax. The collectables rules (which can tax gains on items acquired above $500) are not separately modelled — a disclosed simplification.

A person set to non-resident in the Setup section is taxed on the non-resident bracket table (no tax-free threshold), with no Medicare levy and no LITO. Two disclosed simplifications: franking credits remain refundable offsets for non-residents, and non-resident CGT treatment (discount and indexation eligibility differences under both regimes) is not differentiated in this version.

Property: rent is the owner's assessable income; deductible property expenses and deductible loan interest are the owner's deductions. Net rental losses offset other income when the loss year is before FY2027–28, the property is a new build, or the property was acquired before 12 May 2026 (grandfathered); otherwise the loss is quarantined per owner — carried forward against future net rental profits first, then capital gains, per the enacted restriction. A main residence is CGT-exempt and generates no income or deductions; holiday-home expenses are non-deductible and not modelled. Main-residence complexities (such as the six-year absence rule) are out of scope until property sales are modelled; properties are held to the projection end and are never drawn on to fund deficits (illiquid). Stamp duty schedules for all eight jurisdictions are embedded at their as-at values (nominal-dollar brackets, not indexed) — the per-property duty override exists for precision. First-home-buyer phase-outs are linearised. A planned main-residence purchase can release First Home Super Saver Scheme savings at settlement (see the Superannuation section) — the toggle is only available for a planned main residence, never an investment or holiday property or one already owned.

Lenders Mortgage Insurance (LMI) applies above 80% LVR, calculated from an embedded premium table indexed by LVR band and loan size (indicative tiering, not a specific insurer's published rate card — confirm before client use; a per-purchase override exists for precision, entered in nominal dollars of the purchase year like the duty override). By default the premium is capitalised — added to the loan's drawn balance rather than paid as cash — with an option to pay it at settlement instead, which adds it to the cash required. A per-purchase First Home Guarantee toggle (available once First home buyer is ticked) waives LMI entirely; the scheme's property price cap is modelled per state (indicative figures, one cap per state since no postcode/region is captured — confirm against Housing Australia). A purchase priced over the cap is flagged, not blocked, since caps are reviewed periodically.

Indexed holds brackets, Medicare thresholds, and LITO constant in real terms from FY2027–28 onward — the model of CPI-indexed tax settings, exact in today's dollars. No indexation freezes the FY2027–28 nominal settings, so in real terms every threshold shrinks by CPI each year and tax quietly rises.

Superannuation

Accumulation-phase super is modelled as a distinct asset class with its own tax treatment. Accounts hold a balance and a tax-free component, grow monthly at their allocation's expected return net of a 15% earnings tax on the income component and an effective 10% tax on the growth component (the CGT one-third discount assumed inside the fund — a smoothing simplification, not a parcel-level calculation). Employer Superannuation Guarantee contributions are derived from employment income rows at 12% up to the per-FY maximum contributions base, and can be switched off per row. Salary sacrifice, personal deductible, personal non-deductible, and spouse contributions can also be entered, either as a fixed amount, a percentage of a nominated income row, or set to automatically fill the remaining concessional cap each year.

Concessional contributions (SG, salary sacrifice, personal deductible) share a single annual cap plus any unused cap carried forward from the preceding five years — carry-forward usable only when the person's total super balance at the prior 30 June was under the carry-forward gate. Contributions over the cap are included in assessable income at marginal rates with a 15% non-refundable offset for the contributions tax already paid; the excess itself is left in the fund (no release-from-fund election is modelled). Non-concessional contributions draw on a separate annual cap with an automatic multi-year bring-forward when a contribution exceeds it and the person's total super balance permits; amounts beyond the available bring-forward total are rejected outright and flagged, not taxed as excess. Contributions after age 75, and personal deductible contributions between age 67 and 74 without the work test met, are rejected and flagged.

The First Home Super Saver Scheme (FHSSS) lets a voluntary contribution flagged eligible (Investment cashflows' Super contributions section; salary sacrifice, personal deductible, or personal non-deductible only — Superannuation Guarantee and spouse contributions are never eligible) build toward release at a planned main-residence purchase (Property section's "Release FHSSS at purchase" toggle). Eligible contributions accrue associated earnings at a configurable deemed rate (Inflation assumption section) and are capped at $15,000 per financial year and $50,000 over a lifetime, combined across both concessional and non-concessional amounts — a request that would exceed either cap is accepted only up to it, split proportionally between the two types, and the excess is flagged, not credited. On release: 85% of eligible concessional contributions plus 100% of eligible non-concessional contributions plus all associated earnings arrive as household cash in the purchase's settlement month, reducing the cash required at settlement dollar for dollar; the concessional 15% notionally already taxed in the fund is never released. The taxable component (the released 85% plus all associated earnings) is assessed at the member's marginal rate with a 30% non-refundable offset, as part of that financial year's ordinary tax assessment — it settles through the same PAYG/refund mechanism as every other tax component in this tool, not netted at the settlement month itself, since the true marginal rate depends on the whole year's income. Not modelled, by design: the 12-month recontribution requirement if a home is ultimately not purchased, first-home-buyer eligibility conditions, and the maximum release request (this tool always requests the full eligible balance). Every projection starts FHSSS tracking fresh at zero — a real client's already-accrued contributions or prior release aren't carried in from outside the tool, the same disclosed simplification as the non-concessional bring-forward window above. Only one lifetime release per person is modelled.

Division 293 tax (an additional 15% on low-tax contributions for high-income individuals) is assessed per person each financial year and paid in July of the following year, the same convention as CGT.

Division 296 tax (an additional tax on high superannuation balances, from 1 July 2026) is assessed per person each financial year on total superannuation balances above $3m: 15% of the proportion of that year's superannuation earnings attributable to the balance above $3m, plus a further 10% on the proportion attributable to the balance above $10m (30%/40% effective, combined with the fund's own 15% earnings tax) — the total superannuation balance used is whichever of the start-of-year or end-of-year figure is higher. Earnings for this purpose are modelled as REALISED earnings only — the same income and growth components already used for the fund's ordinary earnings tax, which this tool treats as realised smoothly across the year rather than tracking individual asset sales inside the fund. This smoothing is the largest approximation in this feature; a real fund's Division 296 earnings depend on exactly which gains it actually crystallised that year. Both thresholds are indexed to CPI ($150,000 steps for $3m, $500,000 steps for $10m). Paid in July of the following year, the same convention as CGT and Division 293. Not modelled: the SMSF cost-base reset election for the FY2026–27 transition.

By default, both Division 293 and Division 296 are paid via a release authority against the person's superannuation — the ATO releases the amount directly from a nominated super account (defaulting to whichever account has the largest balance at the time), reducing that balance rather than household cash. This is the common real-world election, and it materially changes high-income scenarios: the cost lands on super instead of cash available for other strategy. A release authority is not a benefit payment, so it applies regardless of preservation status — it bypasses the condition-of-release test below entirely — and is neither assessable income nor subject to withdrawal tax. Each person can instead elect, in Setup, to pay from personal cash, which reproduces the household-outflow treatment described above exactly. If the nominated (or largest-balance) super account can't cover the full amount, the shortfall falls back to household cash, then to unfunded cashflow if cash can't cover it either.

Withdrawals are available once a condition of release is met — reaching age 65, or retiring at or after the age-60 preservation age (approximated by the client's Setup retirement age, since no separate retirement event is modelled). Before release, a super account is completely invisible to the deficit-funding order; once released, it is drawn only after every other funding-order asset is exhausted. Every lump sum is paid in the account's current tax-free/taxable proportion, recalculated at each payment as contributions and growth alter the mix — this is the accumulation-phase rule; pension accounts (not modelled here) instead fix the proportion once at commencement. From age 60, both components of a benefit paid from a taxed source are tax-free, which is the only cohort this tool reaches, so no low-rate-cap withdrawal tax is modelled.

Not modelled, by design: any pension phase (account-based pensions, transition-to-retirement, minimum drawdowns, the transfer balance cap), SMSFs, defined benefit funds, insurance premiums inside super, the downsizer contribution, the CGT small business concessions, death benefits, contribution splitting between spouses, the government co-contribution, the low income super tax offset, excess non-concessional contributions tax (excess amounts are rejected up front instead), release elections for excess concessional contributions, the work-test exemption, and reportable fringe benefits or net investment losses within Division 293 income. Where a rule interacts with one of these, only the accumulation-side treatment is implemented.

Deterministic projection

The projection steps month by month. Each month, in order: assets grow at their expected return; contributions, withdrawals, and one-off amounts are applied to their target assets; household income and expenses are netted; a surplus is spent or invested per your setting; a deficit is drawn from your assets in the funding order, draining each to zero before the next.

Plan years are Australian financial years (1 July – 30 June). Annual cashflows and one-off amounts are applied in July, the first month of each financial year — in a partial first year that starts after July, that year's annual amounts are assumed to have already been made and are skipped. Monthly cashflows simply run for the months in the projection.

An explicit withdrawal larger than its asset's balance takes what is there; the remainder is recorded as unfunded rather than drawn from other assets — a withdrawal is an instruction about a specific asset. A household deficit that outlives every asset in the funding order is also unfunded. Unfunded amounts mean the plan's outflows exceed available funds; the projection reports when that first happens and by how much in total.

Asset returns are expected returns net of ICR, converted to real terms at the CPI setting and compounded monthly. Because the deterministic projection uses expected returns, it will sit above the Monte Carlo median when both are on screen — see "Why a deterministic projection sits above the Monte Carlo median" below.

Liabilities amortise monthly in nominal dollars: interest accrues on the loan balance net of any offset account, repayments follow the standard level-payment formula (interest-only periods pay interest as charged, then principal-and-interest over the remaining term), and payments enter household cashflow like any outflow — because they are nominal-fixed, their real burden falls at CPI over time. The offset portion of an offset account earns nothing nominally (its real value declines by CPI); only the excess above the loan earns the asset's return. Deductible interest joins the owner's deductions (joint loans split 50/50). A liability can be Variable or Fixed: a fixed loan holds its own contracted rate until its own rollover date, then reverts to a configurable revert rate (defaulting to the mortgage rate assumption above) — the level payment recomputes once, exactly at rollover, over the remaining balance and term ("the step change in repayment is the point of the feature — it is not smoothed"). v1 limitation, by design: no redraw.

Under Monte Carlo (below), interest rates are NOT modelled as an independent random process — they are driven off the SAME simulated inflation path each path already generates: marketRate(path, year) = neutralRealRate + cpi(path, year) + margin, applied as a DEVIATION from each loan's own deterministic rate (so a household's actual entered rate is always the anchor, and a path at the assumed average CPI reproduces the deterministic projection exactly, regardless of what rate that loan was entered at). Defaults: neutralRealRate 1.0%, margin 2.5% — together, at the 2.5% CPI assumption, consistent with the 6.0% mortgage rate default above (1.0% + 2.5% + 2.5% = 6.0%), though neither figure needs to be precise: what matters for the model is that rates move dollar-for-dollar with CPI surprises, not their absolute level. A variable loan's rate — and a fixed loan's rate from its own rollover onward — moves with that path's own simulated inflation, and its level payment recomputes at the start of every simulated year to reflect it; a fixed loan's contracted rate stays untouched until its own rollover. This is deliberate, not an oversight: central banks respond to inflation, so this produces the correlation between rates and inflation for free, without an independent process that would otherwise permit economically implausible paths (high inflation with low rates, which is not a world that exists). Verified result — a fixed-rate client's fan chart is WIDER during the fixed period, not narrower (measured: identical client, $700k loan, zero return volatility, 1.5% CPI σ, 400 paths — 10–90 net-asset spread at age 50: variable $7,588, fixed $84,383). A variable borrower's rate moves with inflation, so a high-inflation path's higher nominal repayment is largely offset by faster real erosion of the nominal balance — their real cost is close to invariant to the inflation draw. A fixed borrower's rate does not move, so they take the full real-erosion benefit unhedged in a high-inflation path, and the full penalty in a low-inflation one. Fixing a rate is an inflation bet; a variable rate is substantially hedged in real terms. (An earlier draft of this feature's spec predicted the opposite; see the errata note at the end of docs/specs/14-what-if.md.)

Each liability can carry a repeatable list of extra repayments (a fixed amount, monthly or annual, over a DateRef window) and one-off lump-sum repayments (a fixed amount at a single DateRef) — both reduce the scheduled payment's principal further each time they apply, shortening the loan's actual payoff; the loan closes at zero and stops accruing interest or accepting further extra repayments from that point. Every extra or one-off repayment is a household cashflow outflow through the Working Cash Account exactly like the scheduled repayment — an amount the household can't actually afford still reduces the loan balance (the repayment itself happens, the same convention a planned property purchase's settlement already follows), while its cash consequence runs through the ordinary deficit-funding-then-unfunded cascade, surfacing an unaffordable repayment plan the same way any other unaffordable outflow does. Once a loan with extra or one-off repayments is fully repaid within the projection, the Liabilities table reports the interest and time saved against the scheduled (no-extras) path, in today's dollars — a loan still open at the end of the projection doesn't show this comparison, since the true lifetime interest of either path can't be observed past the projection's edge.

Goals

A goal tracks a named savings target (a car, a wedding, a deposit) separately from ordinary living expenses, funded either from a nominated financial asset or from household surplus. It accrues straight-line from the start of the projection to its (indexed) target date — the (indexed) target amount divided evenly across every month from now until then — modelling "spent at the target date" as the accrual itself: the money progressively leaves its funding source exactly as it's earmarked, so by the target date the target amount has already left the model. There is no separate goal-balance ledger holding money in limbo between accrual and spend, and a goal balance earns no return of its own (it is a pure sinking-fund accumulator, not an investment). An asset-funded goal draws down that asset like a scheduled withdrawal, naturally capped at its balance — if the asset runs low, the goal simply accrues more slowly. A surplus-funded goal is capped at whatever's actually left over in household cashflow each month; unlike an instructed transaction (a loan repayment, a property purchase), a discretionary savings contribution can't manufacture cash that isn't there, so a shortfall here reduces the goal's own accrual rather than becoming unfunded household cashflow. When several surplus-funded goals compete for the same month's surplus, they draw from it in the order they're listed — a disclosed simplification, not a priority system. A goal that falls short by its target date reports the shortfall and an alternative date it would be reached instead, assuming its own average funding rate so far continues unchanged. Goals appear in their own group in the Cashflow table (kept apart from ordinary Expenses, matching the workbook's separate Goals block) and as dated markers on the composite chart (green once reached, red if short).

Snapshot

The Snapshot table (Tables) reproduces the firm's Cash Flow SOA document for up to six chosen plan years, one column per year with Client / Partner / Total sub-columns for a couple. Every figure reuses the SAME per-owner breakdown the Cashflow table's own row vocabulary already computes for the household total, so a Snapshot column always reconciles to the Cashflow table for that year, and Client + Partner always reconciles to Total. A jointly-owned property or liability splits 50/50 between Client and Partner; income, expense, and deduction rows are never joint, so they attribute exactly to their own owner. Two figures have no per-person attribution anywhere in the underlying ledger at all — Working Cash Account interest (a joint cash buffer) and pooled asset cash distributions (summed across every asset regardless of owner) — both also split 50/50 between Client and Partner, a disclosed simplification specific to this per-person view. Year selections default to the current year, retirement, and four more spread across the projection, and persist per scenario. "Copy for Word" copies a plain HTML table to the clipboard — pasting into Word keeps the row/column structure, though this tool does not generate a native .docx file or align to the firm's own document template; Export CSV produces the same rows as a spreadsheet file. Both exports contain exactly the rows visible on screen (Hide empty rows applies to the export too).

Why monthly, not daily

Under i.i.d. normal sampling, returns drawn at any sub-monthly frequency converge to an indistinguishable terminal distribution once correctly scaled. Daily steps multiply compute by ~21× with no visible change in the bands. Sub-monthly granularity only matters when path-dependent features (rebalancing rules, withdrawal triggers, stop-losses) require finer resolution. None apply here.

Annual steps would distort the early years, as monthly contributions do not compress cleanly into yearly buckets.

Why a deterministic projection sits above the Monte Carlo median

A deterministic projection compounds at the expected return μ every year — it represents the arithmetic expected outcome. In a log-normal accumulation, the median outcome sits below that mean, roughly by μ − σ²/2 in continuous time. When both are on screen, the deterministic line will visibly sit above the Monte Carlo median, and the gap widens at higher σ. This is volatility drag — correct behaviour, not a discrepancy to reconcile.

Overlay paths

The 30 grey paths drawn over the Monte Carlo bands are sampled from the full simulation. This is a presentation choice. The fan bands convey the true range; unconstrained overlay sampling occasionally produces outlier paths that compress the chart's y-axis and obscure the central story.

Model limitations

i.i.d. normal sampling within each regime is the dominant simplification. Real returns exhibit:

• Fat tails (more extreme outcomes than the normal distribution predicts)
• Volatility clustering beyond the two-state approximation
• Mild autocorrelation at short lags
• Occasional jumps and regime shifts

Fan widths shown here are likely understated at the tails relative to historical experience. Tools incorporating jump-diffusion, GARCH, or empirical bootstrapping will show wider extremes at the cost of model complexity.

No fees beyond the explicit per-asset ICR inputs. Personal income tax, franking, and CGT are modelled per the Tax section above; the Monte Carlo overlay remains pre-tax until it is rewired. Contribution and withdrawal amounts you enter are in today's purchasing power.

Time diversification: what this tool doesn't capture

The variance ratio measures how dispersion scales with horizon: it's the variance of N-year cumulative returns divided by N times the variance of 1-year returns. Under i.i.d. returns it equals 1 at every horizon. Under mean-reverting returns (the classical case for equities) it falls below 1 — time diversifies, and per-year dispersion narrows with longer horizons. Under persistent returns it rises above 1 — time amplifies dispersion.

Historical US data (Asness, Cliff, Ostdiek and others; 1872–present) is roughly:

The counter-intuitive result is the second row. Bond and bill variance ratios rise with horizon because their real returns are driven by inflation persistence — multi-year inflation surprises compound rather than wash out. Time diversifies equities but anti-diversifies bonds and cash in real terms.

This tool does not reproduce that. Its engine assumes i.i.d. returns within each regime state, with no mean reversion and no inflation persistence. The implied variance ratio is approximately 1 for every asset at every horizon, including the Cash and Defensive profiles. The historical numbers above come from outside the tool's model; treat them as a known limitation of the projection — long-horizon fan widths for bond-heavy and cash-heavy profiles are likely overstated, and for equity profiles likely understated relative to a model that captured time diversification properly.

Scope

This tool projects investment assets — balances, contributions, withdrawals, and costs — under the firm's capital market assumptions, in real terms, with an optional Monte Carlo view of outcome dispersion.

It is analytical, not prescriptive: it presents projections and tradeoffs without recommending one option over another. It models personal income tax, franking, and CGT (see the Tax section), and accumulation-phase superannuation (see the Superannuation section); it does not model age pension entitlement, non-portfolio assets, or product selection.