To tilt or not to tilt — the decision page
What this is. A standalone page for deciding whether a portfolio should carry a tilt, written so that someone with no investing background can answer it. You do not need to have read anything else, and you do not need to know what a “factor premium” is — §2 explains every term in plain words.
How to use it. Read §1–§5 (five minutes). Then answer the eight questions in §6 on the decision sheet in §8 — they are the whole point of the page. §7 says what happens with each kind of answer. Nobody has to be an expert to do this. Everyone has to be honest, especially about §5 and Q3.
Who decides is not decided yet — but needs to be owned
1. The sixty-second version
- A plain global fund owns every company in the world in proportion to its size. Nobody chose that; it is simply what the market is. It is the default.
- A tilt means deliberately owning something different from that default — more small companies, more “cheap” companies, more non-American companies, or a theme (defence, India, gold).
- Tilting is a real choice with a real cost, and it is a risk decision before it is a return decision.
- The idea behind tilts can be sound and still fail for you, because it can take a decade or more to show up — and most people sell somewhere in that decade.
- Duration is the first question, not the idea. If the money might be needed before ten years are out, a tilt is not eligible — the plain fund is the right home for that money.
- A tilt’s crash is barely worse than the plain fund’s crash (§5). Its real cost is the quiet years of falling behind, which is invisible and much harder to sit through.
- The default if nobody decides is: no tilt. That is not a rejection; it is what happens when a decision has no owner. It should still be written down and reviewed (Q8, §7).
- The one hard rule: no written answers, no position. A tilt bought without a reason, a test and a date is how portfolios drift into bets nobody chose.
2. Plain English
| Word | What it means here |
|---|---|
| The market / the index | A list of companies. A plain global fund owns all of them, in proportion to their size (called market-cap weighted). |
| Tilt | Deliberately doing something different from that default, because you think it will do better or because you want less of what the default gives you. |
| Core | The plain global fund that holds most of the money. Everything else is measured against it. |
| Value / small-cap / ex-US | The three classic tilts: “cheap” companies (low price versus profits), smaller companies, and non-American companies. |
| Factor premium | The theory that some groups pay extra because they are riskier or more uncomfortable to hold — you are paid to wait, not paid on demand. |
| Forecast | A view that something specific will happen. Unlike a premium, a forecast can simply be wrong forever. |
| Drawdown | How far something falls from its high before it recovers. |
| Falsifier | The thing that would prove you wrong. No falsifier means you can never be wrong — and a decision that can’t be wrong can’t be reviewed. |
| Counterfactual | What the same money would have done in the fund you did not buy. The only honest way to score a tilt. |
| Duration / horizon | How long the money can stay invested before you need it — and how long you are committing to the bet. |
| Wrapper | The tax pot the money sits in (ISA, SIPP, or a plain taxable account). |
| Concentration | Having a lot of money in a few things, so a few things going wrong hurts a lot. |
3. What a tilt is — and three traps
A tilt is a deliberate departure from the plain index, made for a stated reason, at a stated size, with a way of being proved wrong and a date to look again. Test any idea against these three traps:
- A fund you happen to like is not a tilt. If you can’t say why you own it, it isn’t a decision — it’s a leftover. (That is the most common way a portfolio becomes tilted: nobody chose it.)
- A tilt is not a way to make money faster. It is a way to own different risks and hope to be paid for them. The extra return is not promised; the extra discomfort is.
- A tilt is not “safer”. It changes which risks you carry; it doesn’t remove them. And you may already own the thing you’re about to buy — a plain global fund already holds emerging markets, and many already hold small companies. Check before you add.
4. The case for and against, in ten lines
| The case for a tilt | The case against a tilt |
|---|---|
| The plain index is top-heavy: the world’s ten biggest companies are roughly a quarter of it, mostly American technology. A tilt spreads that. | The price already reflects what everyone knows. A tilt is a claim that you know better, or that the discomfort will be paid for. |
| Some groups have paid extra over very long periods — if you can hold them through the bad decade. | Many tilts lag for a decade at a time. Value lagged growth for thirteen years. Small caps lagged the plain world index in eight of the last eight years. |
| It is cheap now (roughly 0.07–0.22% a year) and takes one click. | Cheap to buy is not the same as cheap to be wrong in: the cost is measured in years, not fees. |
| A tilt you can actually hold through a downturn beats a “perfect” portfolio you abandon. | Most people can’t hold it. The damage is done by selling at the bottom, not by the idea. |
| A long horizon is the best case a tilt ever gets — you can wait out a drought. | The drought is the point: falling quietly behind for years feels worse than a crash, and nobody rings a bell to tell you it’s over. |
5. The two facts that decide most tilts
Fact one: duration. A tilt is not a one-year or five-year bet. Value lagged for thirteen years (2007–2020, J.P. Morgan). Small caps have lagged the plain world index in every one of the last eight calendar years (MSCI, US dollars, gross returns: 2018 −13.5% vs −8.2%; 2024 +8.7% vs +19.2%; and so on). The premium, if it exists, is paid over decades and only to the people still there when it arrives. So the first question is not “is this a good idea?” — it is “can this money be left alone for ten years?”
Fact two: drawdown — and the surprise in it. Worst fall and the recovery it required (MSCI indices, US dollars, measured peak to trough, 2007–2009):
| Plain global index | Value tilt | Small-cap tilt | |
|---|---|---|---|
| Worst fall | −57% | −61% | −61% |
| Gain needed to get back to level | +135% | +155% | +157% |
| On the £10,000 tilt sleeve of a £100,000 portfolio | −£5,746 | −£6,070 | −£6,108 |
Read that last row honestly. On a £100,000 portfolio with a 10% tilt (£10,000), the tilt sleeve would have fallen to about £3,900 in 2008–09, against about £4,250 if the same £10,000 had been in the plain fund: the tilt’s crash was only about £350 worse. Any page that tells you a tilt doubles your risk is exaggerating, and any tilt sold on “it protects you” is selling you something else.
So the real risk is not the crash — it is the wait. Four points of extra crash is survivable and visible. Losing 1–2% a year to the plain fund for a decade, with no drama and no headlines, is what actually breaks people — and it is when most tilts get sold, at the worst moment. That is what §6 is built to test. (Index figures are in US dollars; a sterling investor’s own numbers differ, but the shape is the same.)
6. The killer questions
Answer all eight, in writing, before any money moves. “I don’t know yet” is a valid answer — it means not yet, not “no”.
A. The basis
Q1. What do you believe that the price does not already reflect — and is it a premium or a forecast?
- Why it decides it: prices already contain what everyone knows. A tilt is a claim that the market is wrong, or that you will be paid for bearing discomfort. A premium is judged over decades; a forecast can just be wrong, forever. They need different rules.
- ✅ Good answer: “I own smaller companies because they carry more risk and have historically paid for it — I expect long droughts and I’ll wait them out.” Or: “A quarter of my money sits in ten American technology companies and I don’t want that.”
- ❌ Fails: “it has done well lately”, “everyone is buying it”, “I like the sector”, or naming a fund instead of a reason.
B. Duration
Q2. How long can this money be left alone — and how long are you committing this tilt for, with a review date and an end date?
- Why it decides it: duration is the fact that decides eligibility. If the money is needed before ten years, a tilt is not eligible — that money belongs in the core, not in a bet.
- ✅ Good answer: “This money is for 10+ years. I’m committing for at least five years before judging it, I’ll review it every year, and I’ll re-decide in 2036. If the reason has failed before then, I sell — no waiting for the price to come back.”
- ❌ Fails: “a couple of years”, “until it does better”, or no end date at all. A position with no end date is a mood, not a decision.
C. Drawdown
Q3. What is the worst realistic fall in pounds — and what will you actually do when it happens?
- Why it decides it: size is decided by the loss you can tolerate, not the return you hope for. Say the number before you own it, not after.
- ✅ Good answer: “The tilt is £10,000, so a 2008-style crash takes it to about £3,900 and the portfolio total falls about X%. I’d hold, because the reason hasn’t failed and I am not selling my spending money. If I’d panic at that number, I’ll halve the size now.”
- ❌ Fails: “it won’t fall that much”; “I’d buy more” without naming the money; or a size whose bad-case number you’d actually sell at.
D. Proof
Q4. What does it have to beat — and how will you measure it?
- Why it decides it: a tilt is only meaningful against the fund you would otherwise have bought.
- ✅ Good answer: “It has to beat the plain global tracker, measured same money, same window, same currency, in a one-line table each year. I’ll score my decision separately from the outcome — a good decision can lose money.”
- ❌ Fails: judging it by its own percentage gain, or moving the starting date when it suits.
Q5. What would prove you wrong — and when would you know?
- Why it decides it: without this, every fall becomes “it’s cheaper now” and the decision can never be reviewed.
- ✅ Good answer: a checkable statement with a date: “If the tilt has not beaten the core over ten years, or if the index I’m buying stops doing what I bought it for, I fold it back into the core.”
- ❌ Fails: “it works in the long run”, “it’ll come back”.
Q6. What evidence do you have that you will hold through a bad decade?
- Why it decides it: this is the single biggest reason tilts fail — not the idea, the owner.
- ✅ Good answer: a real example of something you held through a loss without selling, plus a written rule for the drought: “I hold unless the falsifier fires. I do not sell because it is behind.”
- ❌ Fails: “I’m sure I’ll be fine”, with nothing behind it. Nobody is judged for being honest here.
E. Money and ownership
Q7. How much — and whose money is it?
- Why it decides it: size decides whether this is a decision or a leak, and only some money is allowed to take this risk.
- ✅ Good answer: a number in £ and % (start at 10% or less), funded from money that is not needed for a decade, with the bad-case loss stated out loud. Never the spending reserve.
- ❌ Fails: “a small amount” with no number, or dipping into money you’ll need for rent, fees or a house deposit.
Q8. Who owns this decision — and if that isn’t the account holder, has he or she explicitly handed it over?
- Why it decides it: a decision with no owner becomes a tilt nobody chose. And a position owned by someone who didn’t want it is the most likely to be dumped in the first bad year.
- ✅ Good answer: one named person who answers these questions, owns the exit rule and does the review —
with, if the account holder is not that person, an explicit “yes, you decide until
”, plus the account holder’s right to take it back at a review. - ❌ Fails: “we’ll decide together / we’ll see how it goes”. That is how the default — no tilt — ends up being reversed without anyone deciding it.
7. The decision rule
| What the sheet says | What happens |
|---|---|
| All eight answered, and they support a tilt | The tilt goes in, at the size written (never more than 10% of the portfolio to start), inside the tax wrapper, with the exit rule and review date recorded in the same sheet. |
| All eight answered, and they say keep it plain | That is a full decision, not a failure to decide. It is still written down and dated, with what would change the answer and when to look again. |
| Some answered “I don’t know yet” | Not yet. The core goes in as planned and the question comes back on the review date. Not-yet is not no. |
| Nobody is willing to own it (Q8) | The money stays plain, and that is recorded with a review date. An unowned decision is not a licence to tilt. |
| A tilt without the sheet | Not placed. |
Two rules either way: (1) the tilt is scored against the core from day one, same money and window; (2) the review date goes in the calendar and the sheet is the record — not anyone’s memory.
8. The decision sheet
| # | Question | Answer |
|---|---|---|
| Q1 | What do I believe the price does not reflect — premium or forecast? | |
| Q2 | How long can this money be left alone; how long am I committing, and reviewed when? | |
| Q3 | Worst realistic fall in £, and what I will do then | |
| Q4 | What it must beat, and how I will measure it | |
| Q5 | What would prove me wrong, and by when | |
| Q6 | Evidence I can hold through a long drought | |
| Q7 | Size (£ and %), funding source, and the loss if it halves | |
| Q8 | Who owns this decision, and the delegation if the owner isn’t the account holder |
Decision: ☐ Tilt, as specified ☐ Keep it plain ☐ Not yet — revisit on ________ Amount (£ and %): ____________ Exit rule: ____________ Review date: __________ Owner: ____________ Signed / dated: __________
10. Independent corroboration — PensionCraft live, 20 August 2026
Two practitioners with no connection to this household said much the same thing on a live stream, and it is worth recording here because external agreement is evidence and my own reasoning is not. (Full review, with timestamps.)
- “Anything that deviates from market-cap weight is an active decision” (Toby Newbatt, 27:25–28:10) — the same claim this page makes in §3 and §5, from someone who has decided it for his own pension: his single holding is a global all-cap fund.
- The worst-case test before tilting someone else’s money (28:10): “think about the worst case. Are you going to feel really bad if in 20 years’ time it’s underperformed?” That is Q3 and Q6 on the decision sheet asked in one sentence, and it is the question I would put to any owner before a tilt goes in.
- The psychology, evidenced rather than asserted (26:28–27:25): Ramin Nakisa ran the Vanguard factor funds and held value while it underperformed — “every day was like torture watching this thing underperform” — and still knew that switching would be the wrong response. This is the failure mode §6’s duration test is trying to catch before it happens.
- A tilt away from the US is two decisions, not one (14:08): you must decide when to go out and when to come back — “is there a magic signal you’re waiting for?” That sharpens the exit-rule line on the sheet: a regional tilt needs an entry and an exit written down at the same time, or it is not a decision, it is a mood.
- Sizing comparator (14:22): Ramin runs his tilts inside a side portfolio that is 10% of his investments — the same ceiling §7 sets for a first tilt here.
What it does not change. Nothing structural: the eight questions, the 10% cap and the “no sheet, no tilt” rule all stand. It does add one sentence to the honest answer set — if the owner cannot say how a tilt ends as well as how it starts, the answer to §7 is “not yet”.
My assessment (Alf)
The test is duration, not cleverness. Nearly every tilt failure I can evidence was not a bad idea being wrong; it was an owner who was not ready to wait — and the waiting is the whole premium. So I would not start with “value or growth?” I would start with Q2: can this money be left alone for ten years, and will whoever owns it still be there at the end? If the answer to that is shaky, the honest conclusion is the plain core, and the tilt question returns when the horizon is genuinely long.
My prior, stated plainly: for a novice owner I’d default to plain, and I’d allow a small tilt only where the sheet is answered without hedging. That is an opinion, not a rule — a deliberately owned, properly sized tilt at a 40-year horizon is defensible, and it is the best case a tilt ever gets.
On ownership, since it is undecided: a tilt bought for someone who didn’t ask for it is the position most likely to be sold in its first bad year, which is the worst of both worlds. So either the person who wants it owns it (and answers these questions as the owner of the losses), or the money stays plain and the question comes back when the account holder wants it. Both are respectable. What I would avoid is a tilt that nobody has actually signed for.
What I will do either way: implement whichever decision the sheet records, at the size written, with the exit and review dates set; score it honestly against the core; and judge the decision rather than the outcome. The only answer I will push back on is a position with no written reason.
Analysis for the vault, not financial advice. No money moves until the sheet is filled in.
Sources
- MSCI index factsheets (index owner’s own data, checked 15 September 2026, US dollars, gross returns): MSCI World — maximum drawdown 57.46%, 31 Oct 2007 to 9 Mar 2009; MSCI World Value — 60.70%, 13 Jul 2007 to 9 Mar 2009; MSCI World Small Cap — 61.08%, 13 Jul 2007 to 9 Mar 2009; and the annual returns used for the eight-of-eight comparison (2018–2025).
- J.P. Morgan Asset Management, Value vs growth investing: a historical overview — value’s underperformance from mid-2007 to late 2020, the longest since the Second World War.
- Invesco factsheet, IE000716YHJ7 (the core fund) — FTSE All-World covers large and mid companies in developed and emerging markets: emerging markets are already inside the core, and small companies are not.
- Vault: To Tilt or not to Tilt (the long-form arguments) and To Tilt or not to Tilt (the same questions answered for a private portfolio — useful as an example of what a filled-in sheet looks like, including where the honest answer was “I don’t know”).