Use model portfolios as decision templates, not as magic allocations

Research snapshot

Model portfolio shelf

What a model portfolio actually is (and is not)

A model portfolio is a starting structure, not a recommendation. It answers one question well — "what should the skeleton look like for this job?" — and leaves the rest to you. The value is in the shape: a small number of broad, low-cost building blocks doing most of the work (the core), and at most a couple of deliberate tilts on top (satellites). The shelf above gives you that shape so you are editing a sensible default in the builder rather than assembling something from a blank page.

What a model is not: a forecast, a guarantee, or a fixed allocation you should copy to the decimal. Two investors with the same age can correctly hold very different portfolios because their time horizon, income stability and tolerance for a paper loss differ. Treat the weights as a considered default to argue with, not a number to obey. If you only take one thing from this page, take the core-and-satellite habit: decide the boring 80–100% first, and only then ask whether any tilt earns its place.

Match the model to the job, not the other way round

The single biggest input is time horizon — how long until you actually spend the money. The longer the horizon, the more short-term volatility you can sit through in exchange for higher expected long-run return, so the higher the sensible equity weight. A pot you will not touch for decades can hold a very different shape from one you need in a few years, where capital stability matters more than growth. Risk capacity (can you afford a fall without derailing the plan?) usually matters more than risk appetite (how a fall feels), and the two are easy to confuse in a calm market.

Work in this order: define the job and horizon, pick the model whose equity/defensive split fits that horizon, then sanity-check it against how you behaved the last time markets fell. If you are unsure how horizon maps to an equity weight, the time-horizon and drip-feed framing in how to start investing and drip-feed vs lump sum is the right place to calibrate before you load a model into the builder.

Implementing a model in a UK ISA or SIPP

A model only works if it survives contact with the UK wrapper reality. A few structural choices matter more than ticker selection:

• Wrapper first. Holding the model inside a Stocks & Shares ISA or SIPP removes the dividend and capital-gains friction that quietly erodes a General Investment Account. The complete UK ISA guide covers the allowance mechanics; if you are moving existing holdings in, do it without an unnecessary tax event using bed-and-ISA.
• Accumulating vs distributing. For a long-horizon growth model, accumulating share classes reinvest income for you and reduce admin; the trade-offs are set out in accumulating vs distributing ETFs.
• UCITS and domicile. UK investors should generally prefer UCITS, usually Ireland-domiciled, for tax and reporting reasons — see UCITS vs US-domiciled ETFs. Copying a US 60/40 model with US-listed funds is a common own goal.
• Cost compounds. The ongoing charge is one of the few variables you control with certainty. A small annual difference in fund cost compounds against you for as long as the model runs, which is why the core is built from the cheapest broad funds that do the job.
• Rebalancing. A model drifts as markets move. Rebalancing back toward target — on a fixed cadence or a drift band — is what keeps the risk you signed up for from quietly becoming a different portfolio.

Before any holding reaches a model, it should clear the same checks a professional would run: index methodology, replication, size and closure risk, tracking difference and spread. Run candidates through the ETF due-diligence checklist, and if you are still deciding between fund structures see fund vs ETF vs investment trust and the broader UK ETF guide. A pure two- or three-fund version of any model is a legitimate destination, not a compromise — the logic is in how to build a 3-fund portfolio.

Common ways model portfolios go wrong

• Performance chasing. Switching to whichever model topped the last 12 months is buying high and selling low with extra steps.
• Accidental over-diversification. Stacking five "global" funds usually buys the same large companies five times while adding cost and complexity, not real diversification.
• Tinkering. Every discretionary change is a chance to be wrong twice (when you sell and when you buy back). A model's discipline is most of its value.
• Ignoring the wrapper. A great allocation in the wrong account can hand back its edge in tax and admin.
• Copying a foreign template literally. Headline US models rarely map cleanly onto UK wrappers, domicile or currency exposure.

How many funds does a model portfolio need?

For most investors, fewer than instinct suggests. A single broad global equity fund plus one high-quality bond fund already captures the large majority of the diversification on offer; a third building block is usually about a deliberate tilt rather than more diversification. More lines mean more cost, more overlap and more to rebalance, without a matching improvement in outcome. Add a fund only when you can say in one sentence what job it does that the existing holdings do not.

How often should I rebalance a model portfolio?

A simple, durable rule beats market timing: rebalance on a fixed cadence (for example once a year) or whenever an asset class drifts beyond a set band from its target. Inside an ISA or SIPP this is generally frictionless; in a taxable account, prefer directing new contributions toward the underweight asset so you rebalance without triggering a disposal. The aim is to keep the risk level you chose, not to predict the next move.

Should a model portfolio change as I get older?

Usually yes, but driven by shrinking time horizon rather than a birthday. As the date you need the money approaches, the cost of a deep drawdown rises because there is less time to recover, so most plans gradually shift weight from growth assets toward more stable ones. The right glide is gradual and tied to when each pound is actually spent — a pension you will draw over decades in retirement still has a long horizon for much of its balance, so "old" does not automatically mean "defensive".