The score is a single number from 0 to 7 that summarises how technically healthy a fund looks right now. It's recomputed every time the dashboard refreshes (typically daily).

Five independent technical signals (A–E) each contribute points if they pass. A, B and C are 1 point each; D and E are 2 points each (they read sustained trend, not a single moment). Add them up: 7 = every signal firing including the sustained-trend ones, 0 = nothing working.

The score measures recent technical health — price behaviour over the past 1–6 months. It does NOT predict future returns. It says whether the fund's relative strength, MACD, RSI mean-reversion, channel breakout and trend consistency all currently look like a fund in a healthy phase.

Score is absolute (0–7). It measures internal signal health regardless of how peers are doing.

Rank is relative (1 = best). It places the fund against every other PruLink fund by 12-month total return.

A fund can have a high score but middling rank (technically healthy but not actually outperforming peers — wait for it to break out), or a low score but high rank (trailing performance is strong but the underlying signals are mixed — often a fund whose momentum is fading). Entry rules use BOTH because each filters out a different kind of false positive.

Fund size measures structural risk ("will this fund still exist in 12 months?") — a fundamentally different question from price action. Mixing them into one number would blur two different signals and lose information.

Instead, fund size is surfaced separately as the SMALL badge plus the Monthly fund-size audit card on the dashboard. Visual flags, not part of the algorithmic score.

Scores fluctuate month-to-month due to normal market noise. A score-1 fund might recover to 3 next month — and switching at every wobble would lock in losses right before recovery.

The "exit at score ≤ 1" rule waits until four of five signals fail before action. That's strict enough to filter most noise without being so strict that you ride a fund all the way to the bottom. Backtest evidence: across 10 archetype × start-date scenarios, exit-at-≤1 outperforms exit-at-0 (too patient) in every cell, by an average +22 percentage points of total return vs buy-and-hold.

DIST = distribution share class. These funds pay regular cash distributions to investors (monthly, quarterly, or yearly). The fund name usually ends in (Dis), (Distribution), or (Dist).

The scoring engine reads bid price only. Each time a distribution is paid, the NAV drops by exactly that amount — the investor's wealth is unchanged, they just got cash in their bank instead of NAV in their units. So distribution funds look like persistent underperformers on a price-only chart even when total return is on target.

The score is not diagnostic for DIST funds. They are excluded from the Exit triggers list. The right question for an income holding is "are distributions still being paid at the expected rate?" — not "what's the trend score?"

CLOSED = soft-closed to new investments. Prudential no longer accepts new money into this fund, including switches IN. Existing holders can keep their units but can't add.

Closed funds get special treatment because the standard methodology breaks down: a fund that can't grow its asset base will eventually be wound down, manager attention typically drops, and "wait for recovery" carries asymmetric downside.

Rule: exit a closed fund at any score when an eligible replacement exists. If no same-sleeve replacement qualifies, the rebalance engine permits a sleeve change into defensive (bonds/cash) — unique to closed-fund exits.

The list of closed funds is maintained manually in the project's code. As of the last audit, only PruLink India Equity Fund is closed.

SMALL = low assets under management. The fund's feeder AUM is below S$10 million OR its underlying fund's AUM is below S$200 million.

Small AUM is a leading indicator of soft-closure risk, typically 6–12 months ahead of the announcement. Prudential phases out commercially unviable share classes when AUM falls too low.

BUT small AUM alone is not a confirmed problem. A fund can be small because it's newly launched, niche, or low-volume. The methodology treats SMALL as informational:

• The score-based exit rule continues to apply normally
• Small funds are excluded from being recommended as switch TARGETS (don't put new money into a fund that might be closed soon)
• Held positions in small funds are NOT proactively exited — the advisor reviews the audit card and decides case-by-case

Fund-size data is refreshed monthly from the factsheets on Prudential's ILP product page.

Each of the five technical signals shows up as a symbol, reading left-to-right in this fixed order: A · Relative Strength, B · MACD, C · RSI bounce, D · Donchian breakout, E · Trend consistency.

• ● filled circle = a 1-point signal passes (signals A, B, C)
• ○ hollow circle = a 1-point signal fails
• ◆ filled diamond = a 2-point signal passes (signals D, E). The diamond shape is a visual reminder that this signal counts double.
• ◇ hollow diamond = a 2-point signal fails

A signal string like ●●○◆◇ reads as: A pass, B pass, C fail, D pass, E fail → score 1 + 1 + 0 + 2 + 0 = 4 of 7.

• Red — exit signal fired (score 0, or closed fund being held). Appears on the dashboard's Today's exit triggers card and on negative returns in tables.
• Amber / yellow — informational warning. Used on the Monthly fund-size audit card and on the AUM column for small funds.
• Orange — used specifically for the SMALL badge.
• Teal / green — eligible, strong, or positive (passing signals, positive returns).
• Grey — neutral, no data, or insufficient history.

Red ⚠️ Today's exit triggers — algorithmic action required. Lists funds that meet the formal exit rule (score 0, or closed fund). A red card means at least one client conversation is needed.

Amber 📋 Monthly fund-size audit — reference only. Lists funds with small AUM or large month-on-month AUM drops. Informational, not auto-action.

If both cards are hidden, the dashboard is "clean" — nothing requires immediate attention.

Test (1 point): Is the fund's 1-month return rank at least 10 places better than its 6-month return rank?

This catches funds that are climbing through the league table — recently outperforming peers by more than they were over the medium term. The idea is to spot accelerating relative momentum before it shows up in the headline trailing return.

A fund could have a great 6-month return but a deteriorating 1-month rank (slipping back) — that fails. Conversely a fund coming off a soft 6-month period can pass if its last month is markedly stronger than its peers'.

Test (1 point): Has the 12/26-day MACD histogram crossed from negative to positive within the last 30 trading days?

The MACD is the classic medium-term momentum indicator: difference between a 12-day and 26-day exponential moving average, smoothed by a 9-day signal line. A bullish histogram crossover means the recent uptrend has just turned positive on the smoothed-momentum view — the formal "uptrend start" handshake.

The 30-day lookback window means the signal stays on for about six weeks after the crossover. After that it fades unless a new crossover prints — by design, so a stale crossover doesn't carry a fund forever.

Test (1 point): Did Wilder's 14-day RSI dip below 30 within the last 30 days, AND is current RSI now above 35?

Below 30 is technically "oversold" — the fund has been falling fast. The bounce condition (RSI now > 35) confirms that the selling pressure has eased and price is recovering. Together they describe a reversal from an oversold extreme, which historically marks high-reward entry points (and occasionally falling knives — hence why it's only one of five signals, not a sole trigger).

A fund that has been quietly trending upward will normally have RSI in the 45–65 zone and will fail this signal — that's intentional. Signal C is specifically the "reversal off the floor" pattern, not the "healthy trend" one (E does the latter).

Test (2 points): Is the current bid in the top 10% of its 50-day high–low range AND is the 10-day price change positive?

The 50-day Donchian channel is the highest high and lowest low of the past 50 trading days. "Channel position ≥ 90%" means today's price is near the top of that range — i.e., we're at or near a 10-week high. The 10-day-positive-change check filters out a price that's drifting near the top without actual upward momentum.

This is a confirmed range breakout — the most decisive of the five signals and one of the two that count double. Funds that pass it are usually in clean uptrends with no overhanging resistance.

Test (2 points): Of the last 40 trading days, has the bid price closed above its 20-day moving average on at least 60% of them?

This is a "persistence of uptrend" check. A fund that spikes for one day above the 20-day MA and then sinks back will fail; a fund that's been quietly running above its short-term moving average for most of the last two months will pass — that's the kind of steady-trending fund the methodology favours.

The MA20 is responsive enough that the 40-day window captures a real regime, not random noise. Worth 2 points because sustained trend is empirically a stronger predictor of forward returns than any single-event signal.

The five signals were chosen to cover orthogonal dimensions of fund health: improving peer-relative momentum (A), just-starting medium-term momentum (B), recovering-from-oversold mean reversion (C), range breakout trend confirmation (D), and sustained trend persistence (E). Each one catches a different mode of "fund is currently working."

Weighting D and E double reflects the empirical finding that sustained-trend signals are more reliable forward indicators than reversal or crossover signals. A fund passing D and E alone (score 4) is in cleaner shape than one passing A, B, and C alone (score 3) — and the scale captures that.

The 0–7 weighted variant replaced an earlier 0–5 equal-weighted engine (Trend/Momentum/Volatility/Drawdown/RSI) after backtests showed the trend-confirmation signals outperformed the older indicators across most archetypes and starting regimes.

Three triggers, in priority order:

1. Score drops to 1 or below (out of 7) — at most one 1-point signal firing and neither of the two 2-point sustained-trend signals (D, E). The exit threshold was calibrated against backtests; exit-at-≤1 outperforms both exit-at-0 (too patient) and exit-at-≤2 (too aggressive) in every tested scenario.
2. The held fund becomes CLOSED to new investments. Exit at any score when an eligible replacement exists.
3. Advisor judgment override — the system is decision support, not a mandate. An advisor can decide to exit for client-specific reasons (rebalancing, life events, tax) regardless of what the algorithm says.

A replacement candidate must satisfy BOTH conditions:

1. Score ≥ 3 (out of 7) — enough signal-strength to be considered a clean setup
2. Rank ≤ 30 (in top 30 by 12-month total return)

Either filter alone isn't enough. A score-5 fund with rank 45 might be technically pristine but not actually outperforming. A rank-1 fund with score 2 might be on a momentum spike with weak underlying signals.

The replacement also can't be: closed (CLOSED badge), small (SMALL badge), the fund being exited, another fund already in the portfolio, or in the same "group" (sector/region) as a fund already held.

For score-based exits: HOLD the failed fund. The rule is "do nothing rather than force a switch into a weak alternative." Holding a broken fund is bad; switching into another weak fund is worse.

For closed-fund exits: a defensive-sleeve fallback is permitted — switch into the best available bond or cash fund, even if that crosses sleeves. The reasoning is that holding a closed fund accrues structural risk regardless of alternatives.

The fallback hierarchy in the rebalance engine:

• Tier 1 (preferred): same-sleeve fund, score ≥ 3, rank ≤ 30
• Tier 2 (best available): same-sleeve fund, score ≥ 1, rank ≤ 30
• Tier 3 (last resort): highest-scoring same-sleeve fund, any rank
• Tier 4 (closed-fund only): defensive sleeve fund

The threshold was tested two ways. First, a historical replay across 10 scenarios (5 archetypes × 2 start dates). Then a Monte Carlo simulation — 200 synthetic 8-year histories generated by block-bootstrap resampling of actual fund returns, run on all 5 archetypes — for 1,000 additional backtests. The robust finding from this combined evidence:

• Drawdown reduction is consistent and reliable. Across all 5 archetypes in the Monte Carlo, exit-at-≤1 produced median max drawdowns 0.9–2.0 percentage points shallower than exit-at-0. This shows up in every archetype, every time.
• Return outperformance is roughly tied. Median portfolio totals between exit-at-≤1 and exit-at-0 differ by less than 0.05% across the Monte Carlo. The historical replay favored exit-at-≤1 by a wide margin, but that path-specific advantage doesn't survive synthetic-history testing.
• Exit-at-≤2 (more aggressive) under-performed both alternatives on returns and didn't add to drawdown protection. More churn, more false exits.

The methodology was built around drawdown protection, not return enhancement. The Monte Carlo confirms exit-at-≤1 delivers that protection structurally — not just on the particular slice of 2018-2026 history we lived through. Score-1 funds are weak (at most one 1-point signal firing, both 2-point sustained-trend signals failed) but not yet fully broken. Rotating at that point catches the breakdown one level before complete collapse, which is what produces the shallower drawdown.

Each filter catches a different false positive:

• Score alone can miss a "technically clean but not delivering" fund — score 5 with rank 45 means everything looks healthy but other funds are outperforming. Probably an early-cycle fund that hasn't moved yet, or a defensive fund. Skip.
• Rank alone can pick a "late-cycle momentum trap" — rank 1 with score 2 means recent performance is strong but underlying signals are mixed. Often a fund that just peaked and is about to roll over. Skip.

The intersection — clean signals AND already-delivering performance — is the sweet spot. Both filters together reject ~80% of the universe in any given month.

The portfolio is divided by purpose:

• Core — broad equity, the long-term growth engine. Examples: PruLink Global Equity Fund, PruLink Singapore Growth Fund.
• Satellite — thematic / regional / specialised equity. Examples: PruLink Global Technology Fund, PruLink Greater China Fund.
• Defensive — bonds, cash, income. Examples: PruLink Singapore Cash Fund, PruLink Singapore Dynamic Bond Fund.

Each sleeve gets a target weight per archetype. Switching always happens WITHIN a sleeve — core fund replaces core fund. The only sleeve-crossing exception is the closed-fund defensive fallback.

Archetypes map a client's profile (investment horizon, drawdown tolerance, income need) to a target weight breakdown across sleeves:

• Conservative — high defensive (~70%), low equity. Short horizon or low risk tolerance.
• Moderate-Conservative — defensive-tilted balanced (~50% defensive).
• Balanced — equal-ish across core, satellite, defensive.
• Moderate-Aggressive — equity-tilted balanced (~70% equity).
• Aggressive — high core + satellite (~90% equity), low defensive. Long horizon, high risk tolerance.

Each archetype prescribes the number of slots (one slot = one fund holding) and the target weight per slot. The discovery questionnaire on the New Client tab walks through the questions that map a client to an archetype.

Yes. Most archetypes have multiple slots in the same sleeve. For example, an Aggressive archetype might have 2 core slots and 2 satellite slots — four funds total — with the two core slots filled by funds of different geographic or strategy "tilts" (e.g., one US-focused core, one global core).

The point is diversification WITHIN the sleeve. The scoring engine picks the best fund per slot independently, with the constraint that no two slots end up holding funds in the same "group" (same region, same sector).

No. PruLink internal switches are executed bid-to-bid — no bid-offer spread, no commission, no transaction fee. The exit and the entry both happen at the bid price, so there's no friction cost regardless of switching frequency.

This is different from a fresh purchase (which is at the offer price) or a full redemption to cash (at the bid price with potential surrender charges). Internal reallocation between PruLink funds within the same policy carries no cost.

Typically T+2 to T+5 business days from instruction submission to the funds being repriced under the new holding. Bid-bid pricing means the value at the moment of instruction is preserved through the settlement window — no slippage.

The methodology is calibrated for monthly check-ins:

• Biweekly checks add churn without improving outcomes (validated in backtests)
• Quarterly checks miss fast regime changes — a fund can break down between checks and you'd ride the drop for 2–3 months
• Monthly is the sweet spot

For working advisors, a workable hybrid: monthly system check (5 min on the dashboard) + quarterly client meetings + emergency contact when a held fund hits exit between scheduled meetings.

For Singapore tax residents, there are currently no capital gains taxes on PruLink fund switches. Distribution funds may have implications depending on the client's tax residency. The system does not model tax — consult a qualified tax adviser if the client has overseas exposure or unusual tax situations.

Two Prudential sources:

• PruLink Fund Pricing page (pruaccess.prudential.com.sg/prulinkfund/viewFundPricing.do) — daily bid/offer prices for each fund. Scraped each weekday morning.
• PruLink ILP Product page (prudential.com.sg/products/wealth-accumulation/ilp/prulink-funds) — fund factsheets (PDFs) for fund size, currency, manager, and soft-close status. Audited monthly.

All numbers in the dashboard are derived from these two sources. No third-party data is used.

• Prices, scores, ranks: refreshed every weekday morning (9am SGT scheduled scrape, then portal rebuild and republish).
• Fund AUM and soft-close status: refreshed on the 1st of each month (monthly fund audit).
• "As of" date: shown in the top right of every dashboard view. Reflects the latest fund price date, not today's date — Prudential publishes prices with a 1-day lag, so the dashboard typically shows data from the previous business day.

A feeder fund is a Singapore-specific ILP share class that invests primarily into a single master fund managed by an external asset manager. The feeder's AUM is the money Prudential has in this specific share class — it's the "Funds Under Management" figure on Prudential's factsheet.

The underlying fund is the master fund that the feeder feeds into. Its AUM is the master fund's total assets across all distributors globally — typically much larger than the feeder.

Both numbers matter for different risks:

• Small feeder AUM → soft-closure risk (Prudential might phase out the share class)
• Small underlying AUM → wind-down risk (the manager might wind up the strategy entirely)

Non-feeder funds (PruLink directly managed) have only one AUM number — the fund's own assets.

On the 1st of each month, a script:

1. Fetches Prudential's ILP product page and scans for any fund tagged "(soft closed)"
2. Cross-checks the soft-close list against the system's internal CLOSED registry; alerts on any drift
3. Downloads every fund's factsheet PDF
4. Extracts Fund Size (feeder AUM) and Underlying Fund Size from each PDF
5. Compares against the previous month's snapshot; alerts on AUM drops > 30%
6. Saves the snapshot and refreshes the dashboard

Findings appear on the amber Monthly fund-size audit card on the dashboard. No notifications are sent (intentional — keeps the inbox clean).

No. Two-signal weakness is not an exit. The methodology says hold until score hits 0. Scores fluctuate — a 4-to-2 drop in one month often reverts within a quarter.

If the drop continues to score 1, then 0, the exit signal fires and the client conversation happens then. Until then, it's just normal score volatility.

Schedule a client conversation. Use the Diagnose & rebalance tool (Existing Client tab) to enter the client's current holdings — the tool will show the recommended same-sleeve replacement(s) with rationale.

If a replacement is recommended, discuss with the client and execute. If no replacement qualifies, the recommended action is HOLD — don't force a switch into a weak alternative.

Yes — the system is decision support, not a mandate. If you have thesis-based reasons to expect recovery (a specific catalyst, a closing deal, a planned manager change), human judgment overrides the algorithm.

Best practice: document the rationale (in client notes or in your CRM). At the next review, revisit whether the thesis played out. If yes, good call. If no, the algorithm was right and your override cost the client some return — useful feedback.

The system excludes small funds from being algorithmically RECOMMENDED, but it doesn't prevent the advisor from choosing one manually.

Valid reasons to override:

• The client specifically wants exposure to a niche strategy that only this small fund offers
• The fund is small because it's newly launched, not because it's declining
• You've checked the AUM trend (monthly audit card) and it's growing, not shrinking

The SMALL badge is a flag to think twice, not a hard prohibition.

When the Today's Exit Triggers card surfaces a fund that one of your clients is holding. The methodology only flags score-0 exits and closed-fund exits, so these events are rare — when they happen, it's worth a dedicated conversation rather than waiting for the next quarterly review.

Rule of thumb: a held fund at score 0 = call within 7 days. A closed fund being held = call at the next scheduled review (less urgent because there's no acute price action).

Treat it as a score-0 exit (the primary trigger). The score-0 rule applies regardless of fund size. The SMALL flag tells you the structural backdrop is also weak — making it even more important to find a replacement that's not small. The Diagnose tool will automatically exclude small funds from the candidate pool.

113 scenarios spanning:

• Single-fund chains — start with one fund and follow the switching rules over time
• Multi-slot archetypes — start with a portfolio matching one of the five archetypes
• Lump sum entries vs DCA (dollar-cost averaging over 12 months)
• Start dates from 2016 through 2023

Aggregate result across all scenarios: 6PW-IS reduced maximum drawdown in nearly every cell and produced a positive return gap of approximately S$430,000 on equally-sized buy-and-hold benchmarks over the test period.

These are simulations on historical NAVs — not actual managed-account returns.

When a client starts the portfolio right before a major correction. The system's exit signal only fires AFTER the breakdown is confirmed (all five signals failing), by which point the drawdown is already meaningful. Buy-and-hold doesn't pay the exit/re-entry timing cost.

The system is designed for a smoother long-run ride, not for being right every quarter. Clients who optimise for absolute return in every period will be disappointed; clients who optimise for drawdown protection will be satisfied.

The system has no macro view. It reads price action only.

In a fast crash (2020 COVID style), equity funds collectively hit score 0 within weeks. The risk-off auto-detection (triggered when ≥ 60% of equity funds hit score 0 simultaneously) unlocks bond and cash funds as defensive switch candidates — but this is reactive, not predictive. The drawdown has already happened by the time risk-off mode activates.

What this means in practice: the system can't avoid the initial leg down of a fast crash. It can identify when the crash is broad-based (risk-off mode) and shift defensively. It doesn't try to time entries back in — that's an advisor decision.

Explicitly out of scope:

• Predicting macro events (recessions, market crashes, currency moves)
• Accounting for client-specific tax situations
• Accounting for client-specific cashflow needs (planned expenses, retirement timing, education funding)
• Replacing advisor judgment on suitability, KYC, or risk profile
• Overriding regulatory requirements (Financial Advisers Act, KYC, etc.)
• Forecasting returns of any specific fund

The system handles technical fund selection given a chosen archetype. Everything else is the advisor's job.

The 6PW-IS methodology has been stable since 2024. Successor methodologies are in research and tracked in the project repository, but they are not deployed to this portal until backtesting confirms an improvement over the current 6PW-IS rules.

Component-level changes — small additions like the closed-fund rule or the small-fund flag — happen more frequently as gaps are identified in real-world use. These are logged in the methodology page text as they ship.

No. The portal is a decision-support tool for 6PW advisors. The recommendations it surfaces are starting points for advisor-led conversations, not standalone advice. The advisor remains responsible for KYC, suitability assessment, and final recommendations.

Clients should not be given direct access to this portal as their primary source of investment guidance.

All return figures in this portal are results of applying the 6PW-IS rules to historical PruLink fund NAV data. They are simulations, not actual returns of any managed account. Past performance — whether simulated or actual — is not indicative of future results.

The 6PW portal itself is a private analytical tool — not a regulated product. PruLink funds, which the portal references, are issued by Prudential Assurance Company Singapore (Pte) Limited and regulated by the Monetary Authority of Singapore.

Recommendations made by your advisor based on the portal's output are subject to the regulatory framework that applies to your advisor's licensing (typically the Financial Advisers Act, where relevant). Please confirm your advisor's licensing status directly with them.

The system is decision support. Every recommendation can be overridden by advisor judgment. If the disagreement is systematic (i.e., you think a rule is wrong, not just one fund), raise it with the methodology owner — those discussions feed into future versions.

If the override happens often, that's a signal the methodology needs updating. Track override rates over time as a feedback loop.