How we calculate.
The full reference behind every Allocra number: seven cost components with defined formulas and worked examples, and the scoring the ETF Optimiser uses to build and explain its portfolios. Open and unguarded - because if we're going to quote a number to you, you should be able to inspect the working.
1. Purpose
This document defines how Allocra measures and reports the total annual cost drag on a UK self-directed ETF portfolio. It exists because:
- Public claims about cost drag (e.g. "1.5% per year") require a published methodology to be defensible. Without one, the figure is hand-waving.
- The ETF Screener and Tax Optimiser tools require a consistent, internally-coherent cost model. This document is that model.
- Social posts comparing "typical" vs "optimised" portfolios reference this document via the C5 disclosure footnote. Readers who want to inspect the working can find it here.
- As Allocra scales, this is also the document a regulator, journalist, or sophisticated user would interrogate first. We treat it as a public artefact, not an internal scratchpad.
The headline output of this methodology is the total annual cost drag (in % of portfolio AUM per year). It is the amount by which an investor's net returns are reduced below the gross performance of their target index - averaged annually, compounded over the analysis horizon.
2. Scope
2.1 In scope
Six cost components are aggregated to form total drag:
- Platform fees - annual % charged by the broker or platform (e.g. HL fund ISA tier at 0.35%; Trading 212 ISA at 0%)
- Fund-level Ongoing Charges Figure (OCF) - the regulated KIID/KID disclosure
- Fund-level transaction costs - from MIFID II disclosure where available; conservative default otherwise
- FX costs - for funds bought/sold in a currency other than the investor's wrapper currency
- Tax inefficiency - the difference between tax actually paid and tax under an optimal ISA/SIPP/GIA allocation
- Portfolio drift - the cost of not rebalancing relative to a quarterly-rebalanced benchmark
A seventh adjustment partially offsets the above:
- Securities lending revenue - the small income passed back to investors by funds that lend out their portfolio holdings (iShares funds typically share 62.5% with the fund)
2.2 Out of scope
The following costs are real but excluded from the "total drag" figure because they're either too investor-specific, behaviourally-driven, or already covered by other Allocra outputs:
- Realised capital gains tax - timing-dependent; we model wrapper choice (where it's avoided) but not specific realisation events
- Stamp duty - does not apply to UCITS ETFs in the UK
- Behavioural costs - overtrading, performance-chasing, panic-selling. Material but unquantifiable without trade history
- Active-management alpha - positive or negative. This methodology is for index-tracking analysis; active-fund decisions are beyond scope
- Inheritance, estate, or non-UK tax considerations - covered separately in the Tax Optimiser tool with disclaimers
3. Component definitions
Each component has a definition, formula, source-of-truth, and conservative default.
PPlatform fees
Annual fee charged by the broker or platform, expressed as % of portfolio value.
Sources: published rate cards as of fetch date. Allocra maintains a curated list at /platforms/ in the data layer.
| Platform | ISA tier | GIA tier | SIPP tier | Notes |
|---|---|---|---|---|
| Trading 212 | 0% | 0% | 0% | FX charge 0.15% per trade - included in FX line. SIPP launched 2026, zero platform fee. |
| InvestEngine | 0% | 0% | 0% | ETF-only platform |
| Interactive Brokers | 0% | 0% | 0% | FX spread ~0.002% - included in FX line |
| Vanguard Investor | 0.15% | 0.15% | 0.15% | Capped at £375/yr |
| AJ Bell | 0.25% | 0.25% | 0.25% | Capped at £42/yr ISA / £120 SIPP |
| Hargreaves Lansdown | 0.35% | 0.35% | 0.35% | Tiered per account: 0.35% to £250K, 0.25% to £1M, 0.10% to £2M, 0% above; shares and ETFs capped £150/yr (ISA and SIPP). Repriced March 2026. |
| interactive investor | £11.99/m flat | - | - | Annualised varies with portfolio size |
For tiered/capped platforms, calculate at the user's stated portfolio size.
OFund OCF
The fund-level Ongoing Charges Figure as disclosed in the KIID/KID. Includes management fees, administration, audit, custody, and other operating costs of the fund.
Source-of-truth hierarchy:
- KIID/KID PDF (regulated disclosure) - authoritative
- Issuer product page HTML - used when KIID unavailable
- Issuer factsheet PDF - Pattern B parsers (Xtrackers, Invesco) use this as primary
Observation worth recording from VWRL: KIID-disclosed Ongoing Charges (0.19%) can differ from the marketing-page figure (0.22%). We always use the KIID figure as authoritative.
TTransaction costs
Implicit costs incurred by the fund manager when buying/selling underlying securities - bid-ask spreads, market-impact costs, broker commissions on the fund's trades. Disclosed under MIFID II.
- Source: MIFID II transaction-cost disclosure in the fund's KIID (where available - some pre-2018 KIIDs predate MIFID II)
- Conservative default: 0.05% (used when fund provides no figure)
Note: this is separate from the bid-ask spread the investor pays when buying/selling the ETF itself on the secondary market - that's typically negligible for liquid ETFs and is excluded from the drag figure.
FFX costs
The annualised cost of converting between currencies when buying funds denominated in a currency other than the investor's wrapper currency.
Two scenarios:
- Investor buys a GBP-denominated share class of a USD-base fund (e.g. VWRL on LSE in GBP): FX is internal to the fund, partly captured in OCF, partly in tracking error.
F = 0in this case. - Investor buys a USD-denominated share class via a platform that converts at sub-mid rates (e.g. T212 charges 0.15% per FX trade; HL charges up to 1% on the first £5K):
Fis non-zero.
For passive buy-and-hold investors, FX is approximately one-off (purchase + eventual sale) and amortises across the holding period.
Cap: F ≤ 0.5% annualised. Above this, the investor is on the wrong platform or the wrong share class.
XTax inefficiency
The difference between actual tax paid and the tax paid by an optimally-wrapped equivalent portfolio.
Components considered:
- Dividend tax on holdings outside ISA/SIPP - 10.75% basic / 35.75% higher / 39.35% additional rate, after £500 dividend allowance (2026/27, reflecting the 6 April 2026 rise)
- Suboptimal asset location - high-yield assets held in GIA when they could be in ISA/SIPP, paying unnecessary tax
- Wrapper underuse - ISA allowance £20K/yr unused, SIPP allowance £60K/yr unused (for relevant earners)
Optimal is calculated using current HMRC rates and the household's stated income tier.
Cap: X ≤ 0.5%. Above this for an index investor suggests a structural problem (large GIA holdings of high-yield assets that should be migrated).
DPortfolio drift
The cost of not rebalancing a multi-asset portfolio. Drift makes the portfolio progressively more concentrated in winners; the cost is the difference between actual portfolio variance and benchmark variance, plus implementation slippage.
Allocra's default benchmark: quarterly rebalancing to stated target weights. Drift cost is measured against this benchmark.
- Typical drift cost for a 5+ year period on a 60/40 equity/bond portfolio: 0.10 to 0.25%
- Single-ETF portfolios (e.g. VWRL only): D ≈ 0 - no allocation to drift
- Multi-fund portfolios held without rebalancing for >12 months: D can exceed 0.30%
where λ is the investor's stated risk-aversion coefficient (default λ = 1.0).
Cap: D ≤ 0.5%.
−SSecurities lending offset
Many ETFs lend out their underlying holdings to short-sellers, earning a small fee that's partially shared with fund investors. This reduces the effective OCF - so we treat it as negative drag.
- iShares funds: 62.5% of lending income to the fund / 37.5% to BlackRock
- Vanguard funds: ~100% to the fund (BlackRock-style splits are not used)
- Typical magnitude: −0.01% to −0.05%
Source: fund factsheet "Securities Lending Return" line. iShares discloses this directly; Vanguard discloses it less explicitly. Conservative default: S = 0 when undisclosed.
4. Cap-logic rules
Bounds applied to each component (and the total) to prevent absurd or implausible outputs from propagating into user-facing tools. If any value falls outside its range, the ETF Screener flags it for manual review rather than displaying it.
| Component | Symbol | Floor | Ceiling | Default if missing |
|---|---|---|---|---|
| Platform fee | P | 0.00% | 1.00% | 0.25% (HL/AJ Bell midpoint) |
| Fund OCF | O | 0.00% | 5.00% | Reject record (must be sourced) |
| Transaction cost | T | 0.00% | 0.50% | 0.05% (conservative default) |
| FX cost | F | 0.00% | 0.50% | 0% (assume GBP share class) |
| Tax inefficiency | X | 0.00% | 0.50% | Calculated per user inputs |
| Portfolio drift | D | 0.00% | 0.50% | 0% (assume rebalanced) |
| Securities lending | S | −0.10% | 0.00% | 0% (undisclosed = treat as 0) |
4.1 Total cost drag bounds
| Bound | Threshold | ETF Screener behaviour |
|---|---|---|
| Implausibly low | < 0.05% | Flag for review. Even free platforms have some cost (spread, tracking error). |
| Realistic optimised | 0.05% to 0.40% | Display normally. |
| Realistic typical | 0.40% to 2.00% | Display normally. |
| High-cost warning | 2.00% to 3.00% | Display with amber warning + suggest optimisation. |
| Implausibly high | > 3.00% | Flag for review. Above this likely indicates active-fund holdings or input error. |
5. Compounding model
Cost drag is applied annually, end-of-year, to a portfolio that grows at the gross index return rate. Cumulative cost over a horizon is the difference between gross-compounded and net-compounded portfolio values.
Net portfolio value after N years: V_N = V_0 × (1 + r_net)^N
Cumulative cost: V_gross(N) − V_net(N), where V_gross uses r_gross
Convention - drag is compounded multiplicatively, not subtracted at the end. A 1.55% drag on a £500K portfolio over 10 years at 5% gross is NOT "1.55% × 10 × £500K = £77.5K". It's the difference between (1.05)10 and (1.05 − 0.0155)10 applied to £500K - which works out to ~£128K. The compounding nuance matters.
6. Return assumptions
Cost drag in isolation is a fee figure. To project the cumulative cost of that drag over time, an assumed gross return rate is needed.
| Asset class | Allocra working assumption | Historical long-run reference |
|---|---|---|
| Global developed equities | 5.0% | ~7% nominal / ~5.5% real (1900 to 2024, Dimson-Marsh-Staunton; USD with dividends reinvested). We use 5.0% - see §6.1. |
| Emerging market equities | 5.5% | ~8% nominal historically but with higher variance; small EM-premium added to developed baseline. |
| Global aggregate bonds | 2.0% | Long-run real return ~1.5 to 2%; nominal ~3.5% historically but rate-environment dependent. |
| UK Gilts (intermediate) | 2.0% | Conservative; rate environment dependent. |
| Cash (deposit) | 1.5% | Long-run real return ~0.5%; nominal allowance for inflation. |
| 60/40 balanced portfolio | 4.0% | Weighted blend of the above. |
| Concept A default | 5.0% | Used in "What 1.5% costs you over 10 years" chart. Equity-heavy single-fund proxy. |
6.1 Note on the 5% vs 7% choice
Long-run historical nominal returns for global developed equities have averaged approximately 7% per year over the 1900 to 2024 sample (Dimson-Marsh-Staunton; USD-denominated, dividends reinvested). Allocra uses 5.0% as the working assumption in cost-drag projections - including the Concept A "What 1.5% costs you over 10 years" chart and the ETF Screener's default scenario - for four reasons:
- UK-investor FX translation. The 7% nominal historical figure is USD-quoted. UK-based investors experience GBP-translated returns, which historically have been ~0.5 to 1.5% lower per year for unhedged USD assets due to long-run GBP/USD trend.
- Forward-looking compression. Many institutional capital-market-assumption frameworks (BlackRock, Vanguard, JPM Long-Term Capital Market Assumptions, etc.) now project 4.5 to 6.5% nominal for developed equity over the next 10 to 15 years, citing high starting valuations, demographic headwinds, and lower productivity-growth expectations than the 20th-century average.
- Real-terms framing matches investor experience. Inflation has averaged 2.5 to 3% per year in the recent decade; a 5% nominal gross return roughly maps to a 2 to 3% real return - which is what investors actually feel in spending power. The 7% nominal figure feels generous to investors comparing portfolio statements with grocery bills.
- Conservative on the gross makes the cost-drag impact more conservative too. At 5% gross, a 1.55% drag compounds to ~£128K cost over 10 years on a £500K portfolio. At 7% gross, that same drag would compound to ~£186K - even more dramatic. Allocra deliberately uses the less dramatic figure. Our point about cost drag landing harder over time stands at either return assumption, and we'd rather be quoted as understating the value of our tools than overstating it.
The 5% figure is not a forecast of any particular fund or portfolio. It is a working assumption used to translate cost-drag percentages into projected portfolio-value differences for illustrative purposes. The historical 7% (or ~5.5% real) figure is the long-run average over 124 years, which masks substantial year-on-year and decade-on-decade variance.
6.2 Standard disclosure
Any Allocra public output that uses a return-projection figure carries this disclosure, in addition to the C5 cost-methodology footnote in §9:
Projected returns use a 5.0% nominal annual gross return assumption for global developed equity - deliberately conservative versus the long-run historical average of approximately 7% nominal (Dimson-Marsh-Staunton 1900 to 2024, USD with dividends reinvested). The conservative figure reflects (a) UK-investor FX translation, (b) institutional forward-looking projections of 4.5 to 6.5% nominal, and (c) real-terms framing. Actual realised returns can and do diverge materially from any long-run average over 1-10 year periods. Past performance is not a guarantee of future performance. Full methodology at allocra.co/methodology §6.
7. Source-of-truth hierarchy
Per ETL Schema v0.5 §3, the data inputs to this methodology have a defined hierarchy. Reproduced here in cost-relevant form:
| Input | 1st priority | 2nd priority | 3rd priority |
|---|---|---|---|
| Fund OCF | Issuer KIID (PDF) | Issuer product page HTML | Issuer factsheet PDF |
| Fund transaction cost | Issuer KIID (MIFID II line) | Issuer factsheet | Conservative default 0.05% |
| Securities lending return | Issuer factsheet | Issuer annual report | 0 (treat as undisclosed) |
| Platform rate | Platform's published rate card | - | - |
| FX spread | Platform's published FX charge | Pricing reference (BoE mid) | - |
| ISA/SIPP/GIA tax rates | HMRC published rates | - | - |
Tier-1 source date is captured in record metadata (fx_rate_date, cost_source_date, etc. per §2.8 of the ETL schema).
8. Worked examples
Two scenarios that anchor the "typical" vs "optimised" framing used across Allocra's social content and the ETF Screener.
8.1 Typical UK self-directed investor (1.43% total drag)
Profile: £500K portfolio held in HL GIA. Mix of 5 active funds + 2 index trackers. Moderate dividend income. No rebalancing in past 18 months. Investor pays basic-rate income tax.
| Component | Symbol | Value | Rationale |
|---|---|---|---|
| Platform fee | P | 0.30% | HL GIA funds, tiered: 0.35% on first £250K + 0.25% on next £250K, effective 0.30% on £500K (March 2026 schedule) |
| Fund OCF | O | 0.30% | Blended OCF of active + index mix |
| Transaction cost | T | 0.10% | Active funds have higher TCs than passive |
| FX cost | F | 0.20% | Some USD-denominated funds; HL FX charge applies |
| Tax inefficiency | X | 0.40% | GIA-heavy; ISA allowance only partially used |
| Portfolio drift | D | 0.15% | No rebalancing in 18 months on multi-fund portfolio |
| Securities lending | S | −0.02% | Partial offset on iShares holdings |
| TOTAL | 1.43% | ||
Compounded cost over 10 years on £500K @ 5% gross: gross final = £814K, net final (drag = 1.43%) ≈ £710K. Difference: ~£104K.
8.2 Optimised UK self-directed investor (0.30% total drag)
Profile: £500K portfolio held in T212 ISA + SIPP. 2 low-cost iShares ETFs. Quarterly rebalancing. Maximum wrapper utilisation.
| Component | Symbol | Value | Rationale |
|---|---|---|---|
| Platform fee | P | 0.00% | T212 ISA + SIPP both at 0% |
| Fund OCF | O | 0.10% | Blended CSPX (0.07%) + IGLT (0.07%) + EIMI (0.18%) - weighted ~0.10% |
| Transaction cost | T | 0.05% | Passive ETFs only |
| FX cost | F | 0.00% | GBP share classes only |
| Tax inefficiency | X | 0.00% | ISA + SIPP only - no taxable income |
| Portfolio drift | D | 0.10% | Quarterly rebalancing keeps drift minimal |
| Securities lending | S | −0.02% | Partial offset on iShares holdings |
| TOTAL | 0.23% | Rounded to 0.30% in Concept A for conservatism | |
Compounded cost over 10 years on £500K @ 5% gross: gross final = £814K, net final (drag = 0.30%) ≈ £792K. Difference: ~£22K.
8.3 The gap
Difference in final portfolio value between Typical (£710K) and Optimised (£791K) on the same £500K starting portfolio over 10 years at 5% gross: £81K.
Concept A's chart quotes ~£93K from a flat modelled pair (1.55% typical against 0.30% optimised, built from the homepage decomposition of fund fees, tax drag and drift rather than from this platform-specific example) at the same conservative 5.0% gross return. Following the March 2026 platform repricings, the platform-based worked example above lands below that modelled pair (1.43% and an £81K gap). The modelled pair is the marketing illustration; this section is the reproducible arithmetic, and both are linked wherever the headline figures appear.
9. C5 disclosure language
Standard footnote to accompany any Allocra public claim involving cost drag. Used on Concept A, B, C and any future fee-comparison content; also displayed in the ETF Screener output.
Standard form (short)
Cost drag estimates use the Allocra Cost Methodology v1.1 (allocra.co/methodology). Platform fees from published rate cards. Fund OCFs from KIID documents. Tax friction modelled per current UK ISA/SIPP/GIA rules. Returns shown are nominal long-run averages; past performance is not a guarantee of future performance.
Standard form (long, for posts that include specific provider/fund names)
Cost figures: platform fees from each provider's published rate card as of [DATE]. Fund OCFs from the KIID document published by the issuer. Transaction costs from MIFID II disclosure where available, otherwise a 0.05% conservative default. FX costs from platform-published FX charges. Tax inefficiency modelled per HMRC rates for the 2026/27 tax year. Portfolio drift modelled against a quarterly-rebalanced benchmark. Compounded returns use 5.0% nominal gross annual for global developed equity (long-run historical average; not a forecast). Full methodology at allocra.co/methodology. Allocra is not a regulated investment adviser; this is for general information only.
Rules for use:
- Any post or visual asset that includes a specific cost figure (e.g. "1.5%", "£93K gap", "fee comparison") MUST link to /methodology.
- Any fund-specific OCF figure quoted publicly MUST be from the KIID source, not the marketing page (per the VWRL 0.19% vs 0.22% observation).
- Any return projection over 5+ years MUST include the disclosure that past performance is not a guarantee.
- Any wrapper-specific claim (ISA/SIPP allowance utilisation, etc.) MUST cite the relevant HMRC tax year and rate.
10. Optimiser scoring
The ETF Optimiser builds portfolios by scoring every eligible fund and sizing the highest scorers within your constraints. From 22 July 2026 the score behind every selected fund is shown inside the tool itself (tap any holding in an optimised portfolio). This section is the published definition of that score. The engine and the display call the same function; this page describes it.
10.1 Eligibility, before any scoring
A fund must first pass hard gates: it must match your listing region and platform; non-HMRC-Reporting funds are excluded by default (they are taxed as income outside an ISA or SIPP), with reporting status verified against HMRC's published register of approved offshore reporting funds rather than inferred; funds whose analytics our data pipeline cannot compute honestly are excluded rather than scored on missing numbers; and under a Dividend goal, equity funds must carry a real distribution yield. The home-market cap constrains the equity sleeve only: gilt, domestic-bond and property ballast is supposed to be home-market and is never excluded for it. Your own exclusions always win. Your forced inclusions always enter, consciously, past every gate.
10.2 The score
Each eligible fund is scored as a weighted sum. Rewards: distribution yield, three-year NAV return, three-year total return, and dividend growth, weighted by your goal (a Dividend goal weights yield eight times more heavily than a Growth goal does; a Growth goal weights returns six times more heavily than a Dividend goal does). At Low risk, a fund's Sharpe ratio earns a bonus; at High risk, recent growth earns a bonus (halved under a Dividend goal so income books cannot be colonised by zero-yield growth funds). Penalties: volatility (weighted more heavily at Low risk), maximum drawdown (deeper is always worse), and ongoing charges. A monthly-payer bonus applies only if you asked for monthly income.
10.3 From scores to a portfolio
The bond and property sleeve is set by your goal and risk (higher at Low risk, lower for Growth), bounded by your own floor and ceiling. The highest-scoring funds fill each sleeve, sized equally within your per-fund weight cap (a lower cap applies in full-eligible scope so the whole universe can genuinely compete). Forced inclusions are sized first so caps cannot crowd them out.
One fund per index. Where several funds track the same underlying index, only the strongest scorer enters the portfolio: a platform whose X-Ray exposes overlap should not generate it. The boundary is the index itself (an MSCI World fund and an FTSE All-World fund are different indices and may coexist), strategy variants such as covered-call funds count as distinct instruments, and a fund you force-include claims its index so a duplicate is never auto-selected on top of it.
10.4 What the scoring does not know
The engine does not know which funds your platform offers in your specific account type (availability varies by platform and by Trading, ISA or SIPP account; the tool says so under every optimised portfolio). It does not know your tax position. It does not predict returns. Its incentive directions are locked by an automated test suite: a change that made any penalty act as a reward would fail the build.
11. Weekly data-feed quality guards
Every analytic on the platform (returns, volatility, Sharpe, drawdown, yields, prices) is computed from weekly market observations refreshed each Saturday. Market data arrives from third-party sources, and third-party sources contain errors: series that switch quote currency mid-history, single bad quotes that appear and vanish, identifiers that do not resolve. The pipeline's stance is that a dash is always better than a wrong number. These are the guards every fund passes before its figures are published, and they run as automated tests before any weekly refresh can ship.
Identifier integrity. Every fund must carry a structurally valid ISIN (ISO 6166 checksum) and a sane ticker, with no duplicates. Reporting-fund status is not inferred from listings or domicile: it is checked against HMRC's published register of approved offshore reporting funds, refreshed monthly, with the register's date recorded.
Plausibility gates. A computed figure that is not plausible (an annualised return beyond ±150%, volatility beyond 100%) voids the entire return-and-risk family for that fund rather than shipping alongside it. Minimum-history floors apply: three-year figures require at least two and a half years of weekly observations; a young fund shows dashes, not guesses.
Regime truncation. If a source price series changes quote currency or scale mid-history, no figure is computed across the break. Metrics use only the consistent segment after the most recent break, and the truncation date is recorded against the fund. If the trusted segment is too short to support a figure honestly, the figure is a dash until the segment grows.
Single-point despiking. A price observation is discarded only when it moves more than 15% in one week and fully reverses (within 6%) the next: the signature of a bad quote, not a market. Genuine crashes never match this pattern, because real recoveries are partial and slower, and that distinction is pinned by the automated tests. Discarded dates are recorded against the fund.
Prices. The latest observed closing price ships with its observation date and currency, with pence-quoted London lines normalised to pounds at source. A fund whose series fails the guards above ships no price at all.
12. Change log
| Version | Date | Changes |
|---|---|---|
| v1.4 | 2 Sept 2026 | Facts refresh following the site-wide audit: HL platform fees updated to the March 2026 schedule throughout (fund tier 0.35%, ETF cap £150/yr, tiering per account); worked example 8.1 recomputed to 1.43% total drag with the compounding corrected to the stated convention (£710K net, £104K difference, £81K gap in 8.3); Concept A reconciliation restated; Interactive Investor plans updated to the Core/Plus/Premium lineup; Trading 212 SIPP added; dividend tax updated to 2026/27 rates (10.75%/35.75% after the 6 April 2026 rise); tax-inefficiency modelling year moved to 2026/27. |
| v1.3 | 28 Jul 2026 | §11 NEW - Weekly data-feed quality guards published: identifier integrity with HMRC-register-verified reporting status, plausibility gates and minimum-history floors, regime truncation on quote-currency breaks, single-point despiking with the crash-shape distinction, and dated prices with pence normalisation. §10.1 updated - reporting status verified against HMRC's register, not inferred; home-market cap constrains equities only, never ballast. §10.3 updated - one fund per underlying index, best scorer wins, with the strict index boundary and force-include contract stated. |
| v1.0 | 19 May 2026 (am) | Initial draft. Six cost components + securities-lending offset defined. Cap-logic rules established. Two worked examples (typical 1.58% / optimised 0.23%) anchor the Concept A "£93K gap over 10 years" headline. Standard short + long C5 disclosure language drafted. |
| v1.2 | 22 Jul 2026 | §10 NEW - Optimiser scoring published: eligibility gates, score terms and goal weights, sleeve and sizing rules, and stated limitations. Matches the in-tool workings panel shipped the same day. Drawdown term formalised as a strict penalty on drawdown magnitude, with incentive directions locked by an automated test suite. |
| v1.1 | 19 May 2026 (pm) | Widened §6 return-assumption disclosure to acknowledge the ~7% nominal historical figure and explicitly justify the 5% conservatism choice. §6.1 NEW - "Note on the 5% vs 7% choice" with four explicit reasons (UK-investor FX translation, institutional forward-looking projections, real-terms framing, deliberately less-dramatic cost-drag projection). §6.2 NEW - standard return-assumption disclosure box for use alongside the C5 cost-methodology footnote. |
Allocra Ltd is not authorised or regulated by the Financial Conduct Authority. This methodology document is informational and educational only. Questions about a specific figure, source, or assumption? Email hello@allocra.co.