A Model Portfolio Is Only the Starting Point
Model portfolios are often presented as though they define what an investor will hold.
In practice, they do not.
A model portfolio sets out an intended asset allocation or investment structure. But the portfolio ultimately experienced by an investor depends just as much on the rules, tolerances, judgements and controls used to implement that model over time.
Markets move. Cash enters and leaves portfolios. Tax consequences arise. Clients have different circumstances. Investments become unavailable. Advisers exercise judgement. Platforms have different capabilities.
So the real investment proposition is not simply the model portfolio.
It is: the prescribed (model) portfolio + the rules used to approximate it in practice.
That distinction matters because two investors notionally invested in the same model portfolio can experience quite different outcomes depending on how that model is implemented.
It also means that advisers comparing platforms, portfolio managers or implementation services need to look beyond the model itself.
Three questions become particularly important.
1. What drives the need to review a portfolio?
The obvious question is often:
How frequently are portfolios reviewed?
Daily? Monthly? Quarterly? Annually?
But that may be the wrong place to start.
The more fundamental question is:
What do we believe can happen that makes a portfolio require review?
A portfolio may become meaningfully different from its prescribed position because of:
- market movements;
- asset allocation drift;
- changes to the underlying model;
- changes in investment views;
- contributions or withdrawal
- corporate actions;
- cash accumulation;
- client-specific circumstances;
- changes to restrictions or preferences;
- tax considerations; or
- a series of small differences accumulating into something material.
Different investment propositions will reasonably attach different importance to these events.
A manager seeking very close alignment to a prescribed model may regard relatively small deviations as significant.
Another may believe short-term market movements are largely noise and tolerate wider variation.
Another may prioritise tax efficiency and deliberately allow portfolios to remain away from their model where rebalancing would create unnecessary capital gains.
The right review frequency should therefore be an output of the investment philosophy, not simply an operational convention.
This also highlights the importance of separating three activities that are often conflated.
Monitoring means detecting that something has changed.
Review means deciding whether that change matters.
Intervention means actually changing the portfolio.
These need not happen at the same frequency.
A portfolio could be monitored continuously, reviewed when a meaningful event occurs and traded only where intervention is justified.
That is very different from simply saying that a portfolio is “reviewed quarterly”.
The important questions are:
- What do we believe can make the portfolio materially inappropriate?
- How quickly would we want to know?
- How significant must the difference become before it requires judgement?
And when does that judgement turn into action?
2. What variation from the model is permitted?
The second question concerns the extent to which the investor’s actual portfolio is expected to replicate the prescribed portfolio.
In theory, a model might specify exact asset class and investment weights.
In practice, exact replication may be neither possible nor desirable.
An existing investment may be retained to avoid crystallising a capital gain.
A small holding may not justify the transaction cost of selling it.
A client may request that a particular asset be retained.
Cash may temporarily sit above its target.
A platform may not offer a preferred investment.
A minimum trade size may make exact implementation impractical.
An approved substitute may be used.
In other circumstances, an adviser or portfolio manager may simply conclude that an otherwise technically appropriate trade is not in the client’s interests.
None of these differences are necessarily undesirable.
The key issue is the framework within which they occur.
Firms therefore need to decide things such as:
- permitted asset allocation tolerance;
- permitted deviation from individual model weights;
- minimum trade sizes;
- treatment of cash;
- treatment of capital gains;
- treatment of client-directed holdings;
- use of substitute investments;
- handling of restrictions and exclusions; and
- circumstances in which professional judgement may override the model.
These parameters may be less visible than the model portfolio itself, but they can have just as much impact on the investor’s actual experience.
Two firms can therefore offer what appears to be the same investment strategy while delivering materially different portfolios.
The key question is not:
Does the portfolio exactly match the model?
It is:
Are the differences intentional, permitted and appropriately controlled?
3. What should investors and advisers be told?
The third question concerns disclosure.
An investor does not need to understand every operational rule or algorithm used behind the scenes.
But where implementation assumptions materially affect the portfolio they are likely to hold, those assumptions form part of the service they are buying.
Useful disclosure should therefore help an investor and their adviser understand:
- what causes portfolios to be reviewed;
- how quickly meaningful changes are identified;
- whether monitoring is continuous, periodic or event-driven;
- whether portfolios are intended to replicate or approximate the model;
- the tolerances permitted around target allocations;
- what triggers a rebalance;
- how tax influences trading;
- how transaction costs are considered;
- how cash is managed;
- how client-specific restrictions are treated;
- whether substitute investments may be used;
- where professional judgement may override the model; and
- what happens when portfolios move outside agreed parameters.
This matters for transparency.
But it also matters for something arguably just as important:
comparability.
An investor and adviser should be able to compare two different platforms or implementers of model portfolios and understand how their approaches differ.
One may monitor continuously and operate to tight tolerances.
Another may tolerate considerably more drift.
One may prioritise tax efficiency.
Another may prioritise close replication.
One may accommodate extensive client-specific tailoring.
Another may deliver a highly standardised implementation.
These differences may affect:
- investment outcomes;
- tax outcomes;
- transaction costs;
- portfolio turnover;
- responsiveness to model changes;
- personalisation; and
- how closely the investor actually experiences the prescribed investment strategy.
Yet they are often far less visible than fund fees, model performance or asset allocation.
That makes it difficult for advisers to answer a fundamental suitability question:
Which implementation approach is best for this investor?
For one investor, close alignment to the model may be the priority.
For another, minimising unnecessary tax may matter more.
For another, accommodating existing holdings or personal restrictions may be critical.
Good disclosure therefore serves two purposes.
It provides transparency about how the service actually works.
And it provides comparability, allowing advisers and investors to distinguish between different implementation approaches.
Importantly, this does not require everyone to implement models in the same way.
A useful principle is:
Standardisation of disclosure does not require standardisation of implementation.
In fact, better disclosure can support greater diversity in implementation because investors and advisers can understand the trade-offs and choose deliberately between them.
What is the investor really buying?
When an investor is told that their portfolio is being managed according to a prescribed asset allocation or model portfolio, they are buying much more than the model itself.
They are buying:
1. the prescribed portfolio;
2. the firm’s beliefs about what can cause meaningful divergence;
3. the monitoring framework used to detect it;
4. the tolerances permitted around the model;
5. the judgement used in deciding whether intervention is appropriate;
6. the treatment of tax, costs and individual circumstances;
7. the controls used to identify unacceptable differences; and
8. the disclosure that allows the service to be understood and compared.
This is the real-world investment proposition.
The model is simply its most visible component.
From model portfolio management to portfolio assurance
This leads to a broader way of thinking about oversight.
Instead of simply asking:
“Is this investor in Model A?”
we should be asking:
What was prescribed?
What could cause the portfolio to become meaningfully different from that prescription?
Are we monitoring for those things appropriately?
What variation is permitted?
What was actually delivered?
Why are there differences?
Are those differences within the intended framework?
And could the investor and adviser reasonably understand and compare the approach being used?
The objective is not to eliminate judgement.
Nor is it to require every client portfolio to mirror its model exactly.
The objective is to ensure that the gap between what was prescribed, what was permitted and what was delivered is visible, explainable and controlled.
That is the essence of portfolio assurance.
