WHEN
COMPLEXITY
EARNS ITS PLACE.

There is a particular kind of business diagram that should make every entrepreneur nervous.

It contains boxes. Lots of boxes.

Companies, partnerships, trusts, holding entities and arrows march across the page in impressive formation. Jurisdictions appear in several corners. Somewhere, inevitably, there is a dotted line.

It can look wonderfully sophisticated. That does not mean it is good.

In structuring, sophistication should not be measured by how much architecture can be created. It should be measured by how much unnecessary architecture can be avoided.

SIMPLE DESERVES TO WIN FIRST.

If one company can sensibly perform one job, there is little virtue in using three.

Every additional entity, agreement, jurisdiction, bank account, governance process and professional relationship introduces cost, administration and another point at which reality must continue to match design.

So simplicity should begin with an advantage. But it should not receive a permanent exemption from scrutiny.

Because a business can eventually become too complicated for a structure that was designed principally to be simple.

COMPLEX BUSINESSES CAN REQUIRE MORE THAN ONE ANSWER.

A founder may begin with one activity, one market, one shareholder and one principal objective.

Years later, the same enterprise may trade internationally, own valuable intellectual property, hold accumulated capital, make investments, employ people in different places, own property, involve several shareholders and support a family whose future objectives differ from those of the trading business.

Trying to make one legal entity perform every one of those functions may look simple on an organisation chart. Economically, it may be anything but.

The question is therefore not whether complexity is good or bad. It is whether a particular degree of complexity is proportionate to the problems being solved.

EVERY LAYER SHOULD HAVE A JOB.

Additional architecture becomes easier to judge when each component can be described in ordinary language.

These are not reasons automatically to create additional entities. They are examples of functions that may justify considering them.

If the role cannot be explained without reaching for jargon, the design deserves another look.

THE STRUCTURE SHOULD BE COMPLICATED ONLY WHERE THE BUSINESS IS.

A straightforward domestic business with modest retained profits, no meaningful international activity, limited assets outside its trade and no unusual ownership issues may gain very little from elaborate architecture.

Its complexity budget should probably remain unspent.

At the other end of the spectrum, a substantial enterprise with several activities, valuable assets, accumulated capital, international operations, multiple owners and long-term family considerations may be poorly served by insisting that everything remain in one place merely because one place feels simpler.

COMPLEXITY HAS A PRICE.

Any additional layer should be judged against its full cost rather than merely its theoretical benefit.

There may be incorporation and implementation costs. Ongoing accounting, legal, fiduciary, banking, governance, reporting and administration may be required. International arrangements can create additional tax and regulatory considerations. Management time has a value too.

A structure that produces £1 of genuine benefit at a recurring cost of £2 has certainly created complexity. It has not created value.

The same principle applies to flexibility. An arrangement that is elegant on day one but becomes painfully rigid when circumstances change may have solved today's problem by creating tomorrow's.

AND COMPLEXITY HAS TO BE REAL.

Paper architecture is particularly dangerous when it describes a commercial world that does not actually exist.

If an entity is said to perform a function, it should genuinely perform that function. If governance is supposed to occur somewhere, the governance should be genuine. Agreements should reflect what the parties actually do. Appropriate substance, decision-making, records and professional oversight should support the arrangement where required.

This matters especially in cross-border structures, where legal, tax and regulatory outcomes can depend upon detailed facts and conduct rather than attractive diagrams.

A structure should survive contact with reality.

GOOD STRUCTURING IS OFTEN AN EXERCISE IN SUBTRACTION.

Advisers naturally demonstrate value by proposing things. Yet some of the best structural work can involve deciding what not to do.

Do not create an international entity without a genuine reason. Do not separate an asset merely because separation sounds sophisticated. Do not introduce a trust, company, partnership or other vehicle until its role is understood. Do not preserve yesterday's complexity merely because somebody once paid to create it.

And do not assume that a structure must remain complicated forever.

Businesses evolve. Architecture should be capable of evolving with them — sometimes by adding a component, sometimes by changing one and sometimes by removing one entirely.

THE BENEFIT MUST EXCEED THE BURDEN.

What commercial, protective, investment, ownership, succession or other legitimate objective does the additional architecture address? How material is that objective? What does implementation cost? What does ongoing operation require? What new risks or obligations arise? What flexibility is gained or lost?

Tax consequences may form part of that analysis, as may legal and regulatory consequences. But no single number should be allowed to disguise an arrangement that makes little sense as a whole.

The structure has earned its place only when the combined answer remains persuasive.

COMPLEXITY SHOULD BE THE CONSEQUENCE — NEVER THE PRODUCT.

This distinction matters.

A client should not arrive seeking “a structure” any more than a patient should arrive seeking a particular operation before diagnosis.

The starting point is the business, the people around it, the assets, the risks, the geography, the capital and the future objectives.

Perhaps the answer requires additional architecture. Perhaps it requires less. Perhaps nothing needs to change at all.

The quality of the advice lies partly in being willing to discover any of those three answers.