The Opportunity Set

Capital at Risk — Paper 16

I. What Infrastructure Does

Investment advice tends to be described from the client downwards. The adviser establishes the client's circumstances and objectives. An investment proposition is selected. The assets sit within an appropriate tax wrapper and somebody provides custody and administration.

There is an obvious logic to that order because each layer performs a different function. The client's circumstances determine the purpose of the capital, investment management determines how it is deployed, and the wrapper provides its legal and tax structure. Custody sits underneath this, providing safekeeping, dealing and administration.

In practice, influence also travels in the opposite direction. An investment can be suitable for the client, permitted within the wrapper and wanted by the investment manager, yet unavailable because the infrastructure cannot accommodate it. Sometimes the restriction is explicit. More often the investment is simply awkward to administer, cannot participate in an automated process or sits outside an approved universe, and the usual response is substitution. If an investment manager selects one holding but the platform determines which alternatives can actually reach the client, the final portfolio reflects both decisions.

II. How We Got Here

Modern platforms solved real problems that had accumulated across different parts of the market. Pension investment was once dominated by insured funds, while ISAs and taxable portfolios developed through stockbrokers and fund supermarkets. Wrap platforms brought those structures, investment administration and reporting together.

Aggregation improved dealing and administration, and made consolidated reporting easier. Much of the architecture developed around daily-dealt collective funds, which were easy to aggregate and value and had become increasingly central to advised portfolios. Scale then became part of the proposition, with platforms competing on price, administration and the breadth of their investment universes. Standardisation helped make that possible.

But an investment universe containing thousands of funds is not the same thing as an unrestricted opportunity set. A platform has to decide what it can value, trade, reconcile, report and support economically. It cannot support everything. Once those limits determine what an investment manager can buy, administration has started to influence investment construction.

III. When the System Says No

Consider dealing frequency. Some legitimate investment funds do not deal every day. Weekly dealing can reflect the nature of the underlying assets or simply the way a manager has chosen to operate the strategy. Daily liquidity is not itself evidence of a better investment.

Yet non-daily dealing does not fit comfortably into every platform process. Aegon's ARC documentation, for example, excluded non-daily-dealt investments from some recurring-investment and automated-rebalancing processes. During my time there, they were not available on the platform at all.

That is subtler than banning the fund, and it is ordinary operational design rather than a failure of due diligence. The investment can become operationally second-class: an adviser or investment manager can retain it and accept the administrative friction, but the platform creates an obvious incentive to substitute something that fits its machinery.

I encountered a more direct version of the problem with a commercial-property fund held within an advisory model portfolio. Two pension providers rejected the fund because its investment powers permitted derivatives. Unlike the dealing-frequency example, this was not simply a question of fitting an investment into an automated process; the rejection followed from the investment power itself rather than what the fund was actually using it to do.

The fund documentation described the position rather differently. The Janus Henderson UK Property PAIF Feeder Fund stated that derivatives could be used to reduce risk or manage the fund more efficiently. Its risk disclosures simultaneously identified the much more obvious liquidity problem with physical commercial property: properties can take considerable time to sell and redemptions may have to be delayed while assets are realised. The document also described the substantial transaction costs associated with buying and selling physical property.

Derivatives carry risks of their own, but in this case the power to use them was sufficient to stop the analysis. A control designed to restrict risk had no need to ask what the derivative was doing.

This illustrates the difference between due diligence and screening. Due diligence asks what an instrument does and what risks follow from using it; screening can stop once a prohibited characteristic is present. The latter is considerably easier to industrialise.

IV. The Same Model, Almost

Model portfolios expose another version of the problem. An MPS is intended to provide a consistent portfolio for clients sharing an appropriate mandate. Yet the same model can encounter different investment universes on different platforms. A preferred fund may not be available. The required share class may be absent. An instrument may not fit the platform's dealing architecture.

DFMs anticipate this. Preferred holdings can have substitutes that can be used where the first choice is unavailable.

There is nothing inherently wrong with maintaining alternatives. A well-selected substitute may provide substantially the same exposure at a similar cost and can itself be evidence of sensible operational planning.

But it produces an odd result. Two clients can nominally hold the same model portfolio while the actual implementation differs because their platforms differ.

The distinction may be immaterial. A different share class of the same fund may change almost nothing except cost. A genuine fund substitution has more potential to alter style, exposures, liquidity and eventually performance.

The investment manager's preferred portfolio has nevertheless met an external constraint and changed.

V. Designing Around the Platform

Sometimes the investment manager redesigns the implementation instead of substituting individual holdings.

Tatton provides a useful public example. Its published material described its model portfolios as using an overlay structure in which part of the portfolio was held directly in underlying funds while another part was implemented through a fund structure. Tatton explained that this could provide access to investments that might not otherwise be tradable through a platform and assist with liquidity, cost and operational control.

That is a rational response to the environment in which the portfolio has to operate, but it also shows that platform capability has become one of the inputs into portfolio construction. The implementation has been redesigned partly around what the infrastructure can support.

The same issue appears in pension administration. A SIPP operator sits within the FCA regulatory perimeter and has regulatory, capital, due-diligence and administrative considerations that do not belong to the investment manager. A fixed-term deposit may make perfect sense as an asset yet create an entirely different problem for the operator if its terms affect how the asset is classified or administered.

A SSAS is generally an occupational pension scheme whose governance sits primarily within pensions legislation and The Pensions Regulator's regime. It can therefore encounter a different answer to the same proposed investment. FCA-regulated advisers, investment managers and other firms may still sit around the scheme, but the pension itself is operating through a different regulatory architecture. The economic characteristics of the asset have not changed; the structure through which it is held has.

VI. Why the Constraint Exists

This is not an argument for maximum investment freedom.

Every additional asset type creates work somewhere. Somebody has to establish ownership, obtain valuations, process income, reconcile transactions, maintain records and demonstrate that the asset remains permissible. More unusual investments can create additional fraud, liquidity, valuation and operational risks.

The development of the SIPP illustrates how those responsibilities can change. Self-investment was fundamental to the proposition: the member, usually with an adviser or investment manager, directed how the pension was invested, while the operator provided the pension structure and administration. The operator was not there to provide investment advice.

That apparently clean division became harder to sustain. Failures involving unregulated introducers and non-standard investments exposed the consequences of treating the operator as a purely passive administrator. Expectations around due diligence and asset acceptance increased, and operators responded by exercising greater control over what they would permit within their schemes.

There is an understandable reason for that development, but it creates an unusual division of responsibility. The member directs the investment, the adviser assesses suitability and the investment manager may construct the portfolio. The SIPP operator can still determine that an asset will not be admitted to the scheme. Due diligence at the administrative layer has therefore become capable of constraining investment decisions made elsewhere.

Custodians are supposed to safeguard assets and reconcile what happens to them. Pension operators have to preserve the tax and regulatory integrity of their schemes. Platforms have to process enormous numbers of transactions reliably and at a price advisers and clients are prepared to pay. A tightly controlled investment universe can therefore be a feature rather than a defect.

There is also a governance advantage. A firm supporting a limited set of instruments can understand them better, automate more of the administration and supervise exceptions more effectively. The alternative can be expensive bespoke processing with greater scope for error.

The restrictions imposed by each layer can be entirely rational within its own function. The relevant question for the client is what those restrictions do to the opportunity set once they are combined.

VII. Control from Below

A dealing system designed around daily liquidity affects which funds are convenient to hold. A platform universe affects which version of an MPS can be implemented. A wrapper's asset policy can remove an investment selected elsewhere. The investment manager then works within what remains.

This helps explain why the development of financial infrastructure can produce an apparently perverse result. The layer furthest from the client's objectives can exercise considerable influence over the investment opportunity set available to the people closest to them.

Custody and platforms are important precisely because their design can have these consequences. The original purpose of infrastructure is to make investment possible: hold the asset safely, execute transactions, reconcile income and capital, preserve the wrapper and report what has happened. As those functions become more standardised, the infrastructure inevitably develops rules about what it can process.

Eventually those rules can cease merely to administer the investment decision and begin to shape it. Constraints are unavoidable; the better question is whether a particular constraint follows from the client and the investment, or from the limitations of the system through which the investment has to pass.