The client with one W-2 gets a plan in an afternoon: retirement contributions, withholding, maybe a Roth conversion. The client with a consulting S-Corp, two rental LLCs, a partnership interest and a holding company gets a plan that is only as good as the numbers it starts from, and those numbers live in five sets of books that were last closed for the prior year's return. Most multi-entity tax planning is done on stale data, which is why the recommendations are broad ("elect S-Corp", "max the Solo 401(k)") and the quantified savings are guesses.
This guide is for tax planners and the CPAs who do planning for multi-entity owners. It covers the levers that only exist when a client has several entities, the order to pull them in, the data each one needs, and how to run planning as a continuous process on current books instead of an annual project on last year's.
With one entity, the plan is about timing and elections inside a single set of rules. With several, the plan is about where income sits. The same dollar earned in a disregarded LLC, in an S-Corp, in a partnership or in a C-Corp faces different self-employment tax, different deduction limits, different state treatment, and different retirement plan capacity. A multi-entity client has a set of levers that a single-entity client cannot reach, and a set of ways to get it wrong that a single-entity client cannot reach either.
The other difference is data. A single-entity plan can be run from the prior-year return plus a conversation. A multi-entity plan needs the current-year result of every entity, the intercompany flows between them, and the owner's personal picture, because the levers move income between all three.
Before any lever, draw the map: every entity, its tax treatment, its state, who owns it and in what percentage, and every flow between entities (rent, management fees, loans, distributions). Half of the planning opportunities are visible on the map before a single number is run: an operating business paying rent to the owner personally instead of to an asset LLC; a rental LLC inside an S-Corp; three disregarded LLCs that could be one; a loan between siblings with no interest.
The order matters because the levers interact. The sequence that avoids rework:
| Lever | Data required | Where it usually is | Where it should be |
|---|---|---|---|
| Entity selection | Year-to-date profit per entity, owner hours per entity | Prior-year return; a guess | Each entity's closed books, monthly |
| Reasonable compensation | Roles, hours, market rates, entity profit | A ratio | A dated memo with sources, tied to the payroll run |
| Intercompany rent and fees | Agreements, rates, both sides booked | Transfers | Dedicated intercompany accounts that agree on both sides |
| Retirement plans | W-2 compensation per entity, employees across the controlled group | Payroll provider, sometimes | Payroll linked to the entity's books |
| PTET | Entity-level income by state, owner's SALT position | Discovered at filing | Elected in the first quarter, estimates scheduled |
| Loss placement | Depreciation schedules, passive vs active hours | Fixed asset register, if any | Asset register per entity with method and business-use percentage |
| Fourth-quarter cure | Projected total tax vs paid to date | Computed in April | A live forecast across all entities |
Annual planning fails multi-entity clients for a structural reason: the decisions have deadlines spread through the year (S-Corp election by March 15, PTET elections that vary by state, retirement plan adoption before year end, withholding adjustments in December), and the numbers that inform them change monthly. The alternative is a loop: every entity closes monthly, the owner's consolidated picture updates, the projected tax is recomputed, and the planner looks at the deltas rather than rebuilding the model.
The loop needs one thing the traditional workflow does not provide: a place where every entity's current books, the intercompany flows and the personal side exist together and produce a live projection. That is what HaraPro is for the client, and what the planner reads from. Each entity keeps its own books in one login, the ownership map and the intercompany balances come from the books, the tax forecast applies each entity's treatment and updates when a transaction is categorized, and the planner sees the same numbers the client sees. It does not replace the planner's tools for scenarios and elections; it replaces the stale inputs those tools have been running on. Firms can see the arrangement on the CPA page; the Partner tier is a flat fee per client with no revenue share.
The levers move income between entities with different treatments: entity selection per business, compensation across S-Corps, intercompany rent and fees, retirement plan capacity by entity, and state pass-through elections. All of them need current numbers from every entity, not last year's return.
No. The election is made per entity. It helps active businesses with an owner working in them and enough profit to cover the costs; it usually hurts rentals and low-profit or passive entities.
Yes, with a written lease at a market rate, booked on both sides. It moves income between entities and can be a planning tool; without the agreement and the rate, it is reclassified.
The entity pays state income tax and deducts it federally, which sidesteps the individual SALT cap. It is elected per entity, per state and per year, and the entity then makes its own state estimated payments.
Monthly, on closed books. The decisions have deadlines spread across the year, and the numbers behind them change every month; an annual plan is out of date by the second quarter.