Anatomy of a Partnership Flip: Tax, Cash, Control and the Flip
Before credits could be transferred for cash under §6418, a sponsor with no tax appetite had a structural problem.
The project could generate a large tax credit and years of depreciation, but neither item was useful to a sponsor that could not absorb it. A bank or other taxable investor might value those tax attributes, but the tax law did not simply let the sponsor hand over the credit and keep the project untouched.
The investor needed to be inside the ownership structure.
That is the reason the partnership flip exists.
The project is owned through a partnership. The sponsor and tax investor are both real partners. The agreement then changes how tax items, cash and residual economics are shared over time.
There is no single “99/5 split.” There are several different layers:
ownership
tax items
investment-credit basis / qualified investment
cash distributions
§704(b) capital accounts
outside tax basis
liability allocations
DRO exposure
control / consent rights
flip trigger
purchase option / exit
Some of those layers can move together. Others deliberately do not. That separation is the structure.
ℹ️ Note: Revenue Procedure 2007-65 is a safe harbor for certain §45 wind PTC partnerships. It is not a universal tax-equity safe harbor and does not apply to §48 ITC partnerships. Solar and other ITC flip structures therefore require a separate partnership analysis under the Code and regulations.
Meet Project Redwood: One Deal Through the Whole Article
Use one illustrative solar partnership so every rule has a number attached to it.
Qualified investment / assumed credit basis $200m
Illustrative investment-credit rate 30%
Gross investment credit $60m
Tax-equity investor contribution $75m
Pre-flip investor share of relevant tax economics 99%
Pre-flip cash split 40% investor / 60% sponsor
Post-flip cash split 5% investor / 95% sponsor
Illustrative investor target after-tax yield 7.0%
Base-case modeled flip end of Year 6
Annual distributable project cash, base case $10m
These are teaching assumptions, not market terms and not a statement that every ITC allocation can simply be set to 99%. The legal allocation mechanics described below still have to support the result.
Redwood immediately shows why one percentage cannot model a flip.
If the investor's share of qualified investment supports a 99% share of the $60 million credit, its Year-1 credit value is:
$60m × 99% = $59.4m
But if Redwood distributes $10 million of project cash that year under the assumed 40/60 cash split:
Investor cash $4.0m
Sponsor cash $6.0m
The investor can therefore receive $59.4 million of credit and only $4 million of cash in the same period. That is the structure in one example: ownership, tax value and cash are related, but they are not the same waterfall.
The Flip Was Built to Move Tax Attributes Without Giving Away the Project Forever
A developer usually wants three things from tax equity:
- monetize tax credits it cannot efficiently use;
- monetize depreciation that would otherwise sit unused; and
- recover most of the long-term project economics after the investor earns its agreed return.
Those goals point in different directions.
To receive tax benefits through a partnership, the investor must actually be a partner. But the sponsor does not want to permanently sell most of the project.
The flip solves that tension by changing economics over time.
Construction / funding
↓
Tax investor admitted as partner
↓
Pre-flip period
Investor receives a large share of relevant tax economics
Cash follows a separate waterfall
↓
Flip trigger
Target yield, date or other negotiated condition
↓
Post-flip period
Sponsor receives most residual economics
Investor retains a smaller continuing interest
↓
Possible sponsor call / negotiated exit
The legal and tax analysis is not “did the percentages flip?” It is whether the investor held a bona fide partnership interest, whether the allocation rules were respected and whether the investor bore real economic risk while receiving the tax benefits.
The 99/5 Shorthand Comes From a Narrow Wind Safe Harbor
Revenue Procedure 2007-65 is the source of much of the vocabulary people use when describing partnership flips.
It applies only to partnerships claiming §45 production tax credits from wind.
Within that safe harbor:
- the developer must retain at least a 1% interest in each material item of partnership income, gain, loss, deduction and credit throughout the partnership;
- the investor must retain, while it remains a partner, a minimum interest in each material item of partnership income and gain equal to at least 5% of the largest percentage interest it had in those items;
- the investor must make a minimum unconditional investment equal to at least 20% of its fixed plus reasonably anticipated contingent capital contributions;
- at least 75% of those expected capital contributions must be fixed and determinable rather than contingent;
- the developer cannot have a qualifying purchase right exercisable earlier than five years after the facility is placed in service;
- purchase rights must satisfy the safe-harbor pricing requirements;
- the investor cannot have a contractual right to force someone to buy its partnership interest; and
- the investor's credit allocation cannot simply be guaranteed.
If the investor's largest share of relevant income and gain is 99%, then 5% of that maximum is 4.95%.
That is where the famous 4.95% residual comes from.
It is not a free-floating rule that every renewable tax-equity investor must always own exactly 4.95%. It is the mathematical result of applying the safe harbor's minimum-interest test to a 99% maximum allocation.
Even the Safe-Harbor Example Does Not Have One Cash Split
The IRS example in Revenue Procedure 2007-65 is useful because it destroys the idea that “99% tax = 99% cash.”
In the example, the investor receives 99% of gross income/loss and §45 credits during the early periods.
But cash changes separately.
During one period, the developer receives 100% of project cash while the investor receives the 99% tax allocation.
In another period, the investor receives 100% of cash until it reaches its agreed after-tax return.
After the flip, the investor moves to 5% of both cash and tax items and the developer receives 95%.
The exact example is not a market template for every transaction.
The lesson is more important than the percentages:
Tax allocations and cash distributions are separate economic schedules.
Redwood makes that concrete. Its illustrative investor receives 99% of relevant early tax economics but only 40% of the $10 million cash distribution: $4 million, not $9.9 million. The sponsor receives the other $6 million even while most tax value is moving to the investor.
A model that has one “tax equity %” row applied to everything is structurally wrong before any assumptions are entered.
PTC and ITC Allocation Mechanics Are Not the Same
The old version of this post treated tax credits as though every credit were simply allocated 99/1 before the flip.
That is too broad.
Production credits
Section 1.704-1(b)(4)(ii) says that tax credits that arise from partnership receipts must be allocated in accordance with the partners' interests in the partnership when the credit arises.
For §45 PTCs, Revenue Procedure 2007-65 provides the safe harbor for certain wind partnerships.
The §6418 transfer regulations later used a partnership-flip example in which one partner receives 99% of income, gain, loss, deduction and credit before the flip and 5% afterward. The regulations treat the credit allocation as respected where the Rev. Proc. 2007-65 safe-harbor requirements are satisfied.
Investment credits
The ITC mechanics are different.
For investment-credit property, the regulations focus on each partner's share of the basis / qualified investment in the partnership's credit property.
Section 1.46-3(f) generally ties that share to the ratio in which the partners share general profits, unless all related items of income, gain, loss and deduction with respect to the property are specially allocated in the same manner and that allocation is respected under §704.
So an ITC model should not contain a loose input called ITC allocation = 99% without showing the related basis / qualified-investment allocation mechanics that produce that result.
Return to Redwood. Its assumed $200 million of qualified investment produces a $60 million credit at 30%. Only if the partnership's allocation mechanics support the investor's 99% share would the illustrative investor receive $59.4 million of credit value. The number is easy; establishing the allocation is the tax work.
The percentage may end up looking familiar. The legal path is different.
Solar Does Not Get the Wind Safe Harbor
IRS Chief Counsel Memorandum 201524024 considered whether a solar partnership claiming §48 energy credits could rely on Revenue Procedure 2007-65.
The answer was no.
The memorandum states that Rev. Proc. 2007-65 applies only to §45 wind PTC partnerships and does not apply to §48 energy credits.
The memorandum is also explicit that it may not be used or cited as precedent.
That means two things at once.
First, a solar partnership cannot say: “We satisfy Rev. Proc. 2007-65, therefore our structure is safe.” There is no such solar safe harbor.
Second, the memo still shows how IRS Chief Counsel analyzed a real solar structure: whether contributions were sufficiently fixed, whether the investor had a put, and whether sponsor protections effectively removed the investor's risk that the anticipated ITC would not materialize.
The broader issue is partner status and partnership allocation principles, not mechanical compliance with a wind checklist.
A Partnership Flip Is an Ownership Structure, Not a Credit Purchase
That distinction is why guarantees matter.
In the wind safe harbor, the investor cannot be protected against loss of its minimum investment through arrangements with project participants.
No person may guarantee the investor the right to a §45 credit allocation.
The project company must bear resource risk rather than having the developer simply guarantee the wind resource.
The 2015 solar Chief Counsel memo focused on similar economics. The structure under review protected the investor against a tax-credit shortfall and included insurance for the expected credits; Chief Counsel identified those protections as facts that prevented the arrangement from satisfying the wind safe harbor even if the safe harbor had applied.
The practical lesson is not that every indemnity is forbidden.
Tax-equity documents contain representations, covenants and indemnities.
The question is what risk the investor is actually retaining.
There is a difference between:
Sponsor indemnifies investor because sponsor breached a tax representation
and:
Sponsor guarantees investor receives the expected credit regardless of project facts
The second starts to look less like equity ownership and more like a financed tax purchase.
Capital Accounts, Outside Basis and DRO Are Three Different Concepts
This is where first-pass flip models usually become confused.
They are related. They are not interchangeable.
§704(b) capital account
The capital account is part of the partnership allocation framework.
It tracks economic capital for purposes of testing whether allocations have economic effect under the partnership regulations.
Large depreciation allocations can push an investor's §704(b) capital account toward or below zero.
Outside basis
Outside basis is the investor's tax basis in its partnership interest.
It is governed by separate rules.
It generally starts with contributions and can increase for income and certain shares of partnership liabilities; distributions and losses reduce it.
Section 704(d) generally prevents a partner from deducting partnership losses beyond its outside basis. Excess losses are suspended until basis becomes available.
A negative §704(b) capital account does not automatically mean negative outside basis. And a positive capital account does not prove the investor has sufficient outside basis to deduct all losses.
Deficit Restoration Obligation
A DRO is an obligation to restore some or all of a deficit in the partner's capital account if the relevant liquidation occurs.
Under §1.704-1, a valid DRO can support economic effect of allocations to the extent of the restoration obligation.
But a DRO must be real.
The regulations disregard obligations that are not legally enforceable, are impermissible bottom-dollar payment obligations, or are structured in a way that indicates an attempt to avoid the obligation. Relevant factors include commercially reasonable enforcement and collection provisions and financial-condition documentation.
A DRO therefore should not be described as “extra outside basis that lets the investor take more depreciation.”
A better model keeps separate schedules for:
§704(b) capital account
outside basis
§752 liability share
DRO amount
deductible vs suspended loss
The Flip Trigger Is an Economic Engine, Not a Date Cell
Partnership agreements can use different flip mechanisms.
A common structure uses an agreed investor after-tax return as the trigger.
Others can use a fixed date or other negotiated conditions.
There is no IRS rule saying every renewable flip must use a particular target yield.
If the deal uses a yield trigger, however, the model has to solve it correctly.
The investor-return schedule needs to include the economic items the agreement actually counts, potentially including:
- credit value;
- depreciation and other tax deductions;
- taxable income allocations;
- cash distributions;
- capital contributions;
- indemnity or true-up amounts; and
- timing.
The after-tax nature of the calculation matters.
A dollar of credit and a dollar of deduction do not have the same value.
A deduction is worth the tax rate applied to the deductible amount. A credit offsets tax dollar for dollar, subject to the investor's own limitations.
That is why a flip model cannot simply run XIRR on cash distributions.
The Sponsor's Real Risk Is Often Flip Delay
Redwood's base case reaches the illustrative investor target at the end of Year 6.
Now run a downside case. Assume distributable project cash falls from $10 million to $7.5 million a year, and the combined reduction in cash and tax value moves the contractual yield flip from the end of Year 6 to the end of Year 8. The exact yield calculation would use the partnership agreement's full after-tax cash-flow series; here we isolate the sponsor consequence.
If Redwood had already flipped, the sponsor would receive 95% of the $7.5 million annual cash:
Post-flip sponsor cash
= $7.5m × 95%
= $7.125m per year
Because the flip is delayed, Years 7 and 8 remain in the assumed pre-flip 60% sponsor cash split:
Pre-flip sponsor cash
= $7.5m × 60%
= $4.500m per year
The delayed flip therefore keeps:
$7.125m - $4.500m = $2.625m
away from the sponsor in each of Years 7 and 8.
Across those two years:
Sponsor cash deferred by flip delay
= $2.625m × 2
= $5.250m
That $5.25 million is before discounting, back-leverage effects or any later catch-up in sponsor economics.
Now the phrase “flip moves from Year 6 to Year 8” has a dollar consequence. For a back-leverage lender, the same $2.625 million annual reduction is a debt-service-coverage issue, not merely an equity IRR sensitivity.
A useful sensitivity therefore runs:
Generation / credit downside
→ investor after-tax return
→ flip year
→ sponsor distributions by year
→ sponsor NPV / IRR
→ back-leverage DSCR
The investor target is the trigger. The sponsor absorbs the timing.
Transferability Changed the Question, Not the Partnership Rules
Section 6418 changed the market because eligible credits can now be transferred for cash to unrelated taxpayers.
That means a sponsor can monetize a credit without admitting a tax-equity investor solely for the purpose of using the credit.
But transferability does not make partnership flips obsolete.
A credit transfer monetizes the credit. It does not automatically monetize depreciation.
And the final §6418 regulations expressly contemplate partnerships making transfer elections while partnership allocations still operate under §704(b).
One final-regulation example uses a wind partnership with a 99/1 pre-flip allocation and 5/95 post-flip allocation. The partnership transfers one partner's otherwise eligible credit share and allocates the resulting tax-exempt income under the partnership rules.
That is useful because it shows the current world is not tax equity OR transferability.
It can be:
partnership economics
+
credit transfer
depending on the transaction.
The Purchase Option Is About Substance as Much as Exit Value
Sponsors often focus on the call option because that is the path back to concentrated ownership.
Revenue Procedure 2007-65, as modified by Announcement 2009-69, allows qualifying purchase rights within the wind safe harbor only under specific conditions.
The developer cannot have a contractual right to purchase the property or partnership interest earlier than five years after the facility is placed in service.
The purchase price must either:
- be no less than fair market value determined at exercise; or
- if determined in advance, be a price the parties reasonably believe at the time it is set will not be less than fair market value when exercise occurs.
The investor cannot have a contractual right to force someone to buy its partnership interest under the safe harbor.
That last point matters.
A sponsor call preserves investor upside and downside in residual value. An investor put can start to create a protected exit.
For solar, again, the wind safe harbor is not controlling. But the underlying substance question — is this a real equity interest with genuine residual risk? — remains central.
The Five-Year Number Means Different Things in Wind and Solar
It is easy to say “tax equity has a five-year minimum.” That is too imprecise.
In Rev. Proc. 2007-65, the wind safe harbor restricts the developer's contractual purchase right during the first five years after placed-in-service.
For ITC property, there is also a separate five-year recapture regime under §50.
Those are different rules serving different purposes.
A model should keep separate fields for:
ITC recapture period
safe-harbor call restriction, if applicable
contractual flip trigger
contractual purchase-option date
If all four happen to cluster around a similar date, that is a transaction outcome — not one legal rule.
Control Usually Does Not Flip the Same Way Tax Economics Do
A tax investor needs a real partnership interest.
That does not mean it manages day-to-day operations.
In a typical structure, the sponsor remains the managing member while the investor receives consent rights over major decisions.
From a financing perspective, the useful way to model this is a reserved-matters matrix:
sale of project
new debt
amendment of material contracts
change in tax elections
related-party transactions
material litigation settlements
bankruptcy actions
major capex outside approved budget
changes that could impair credits
A closing checklist needs to know which action requires lender consent, tax equity consent, both or neither.
The capital structure is not only a waterfall. It is a veto map.
Project-Level Debt and Tax Equity Pull on the Same Cash
Debt inside the project partnership sits ahead of equity cash distributions.
That can reduce or delay the cash used in the investor's yield calculation.
If debt service consumes more operating cash:
less distributable cash
→ slower investor yield build
→ later flip
→ later sponsor residual economics
Back leverage moves the borrowing outside the project partnership and instead lends against the sponsor's distributions.
That avoids putting the debt directly ahead of the tax investor inside the partnership.
But it does not remove risk.
It concentrates the back-leverage lender on the sponsor's residual cash — the same cash most exposed to a delayed flip.
A back-leverage model should therefore size on sponsor distributions, not project CFADS, and then stress the flip date.
How I Would Build the Model
I would not start with a “99% / 1%” assumption.
I would start with separate engines.
1. Ownership / legal periods
Funding date
Placed-in-service date
Pre-flip period
Flip trigger
Post-flip period
Earliest call date
ITC recapture dates, if relevant
2. Tax allocation schedules
Income
Gain
Loss
Deduction / depreciation
PTC allocation, if applicable
ITC basis / qualified-investment share, if applicable
3. Cash waterfall
Operating expenses
Debt service, if inside project
Reserves
Tax distributions / preferred distributions, if any
Sponsor priority distributions, if any
Investor distributions
Residual split
Do not infer cash from the tax split.
4. Investor return engine
For an illustrative yield-flip:
Investor contributions
+ cash distributions
+ credit value
+ tax value of deductions
- tax on income allocations
± true-ups / indemnity items included by agreement
=
after-tax investor cash-flow series
Then solve the flip condition.
5. Capital / basis schedules
At minimum:
§704(b) capital account
outside basis
share of liabilities
DRO amount
deductible loss
suspended loss
Do not let one schedule stand in for the others.
6. Sponsor economics
Sponsor cash pre-flip
Sponsor cash post-flip
Flip delay cost
Call price / residual value
Sponsor IRR
Sponsor NPV
Back-leverage DSCR, if applicable
The investor target return is only half the transaction.
Follow Redwood Through the Flip
Redwood now reconciles the structure in one view:
| Item | Pre-flip | Post-flip |
|---|---|---|
| Illustrative investor share of relevant tax economics | 99% | 5% |
| Investor cash share | 40% | 5% |
| Sponsor cash share | 60% | 95% |
| Annual base-case distributable cash | $10m | $10m |
| Investor annual cash at $10m project cash | $4.0m | $0.5m |
| Sponsor annual cash at $10m project cash | $6.0m | $9.5m |
| Base modeled flip | End of Year 6 | — |
The Year-1 tax side is different again:
Illustrative project credit $60.0m
Investor share, if allocation respected 99%
Illustrative investor credit value $59.4m
Year-1 project cash $10.0m
Investor cash share 40%
Investor cash $4.0m
So the investor-return engine cannot be investor cash alone. It needs the contractual after-tax series: contribution, credit value, usable deductions, tax on income allocations, cash distributions and any true-ups or other included items.
Redwood's downside case is the practical reason to do this correctly. Moving the flip from Year 6 to Year 8 reduced sponsor cash by $5.25 million over Years 7 and 8 under the assumptions above. A model that simply toggles 99% to 5% in Year 6 cannot show that risk.
The Diligence Questions Are More Specific Than “Is This a Flip?”
A lender, sponsor or tax investor should be able to answer:
- Which credit is being monetized — PTC or ITC?
- Is Revenue Procedure 2007-65 actually applicable, or merely being used as a reference point?
- If PTC, how is the credit allocation tied to §1.704-1(b)(4)(ii)?
- If ITC, how is each partner's share of basis / qualified investment determined?
- Are cash distributions modeled separately from tax allocations?
- What investor minimum investment is actually at risk?
- How much of the investor commitment is fixed versus contingent?
- Which guarantees or indemnities could insulate the investor from project or credit risk?
- Is there an investor put?
- When can the sponsor call the interest, and how is price determined?
- What is the investor's §704(b) capital account trajectory?
- What is its outside basis?
- Is a DRO being relied on, and is it legally enforceable and economically credible?
- What losses are suspended under §704(d)?
- What is the flip year in the downside case?
- What does that delay do to sponsor IRR and back-leverage coverage?
- Could the partnership transfer some credits under §6418 while retaining the ownership structure?
Those questions turn “tax equity” from a label into a transaction model.
The Flip Is Really a Contract for Time
A partnership flip looks like a tax structure because the percentages are tax percentages.
Economically, it is a contract about when ownership economics move back toward the sponsor.
The investor puts real capital at risk and receives a large share of tax economics while its return is being built.
The sponsor gives up tax value and some cash today in exchange for recovering most residual economics later.
Everything difficult sits in the word later.
Later can move because generation changes.
It can move because credits differ from forecast.
It can move because depreciation or loss limitations change the investor's realized tax value.
It can move because the cash waterfall changes.
It can move because debt sits ahead of distributions.
And when later moves, the sponsor's value moves with it.
That is why the flip should never be modeled as one percentage toggle.
It is a time-dependent ownership, tax, cash and control system.
Transferability has made the alternatives better. It has not made that system simpler.
Primary sources: IRS Revenue Procedure 2007-65 · IRS Announcement 2009-69 · IRS Chief Counsel Memorandum 201524024 · IRS final §6418 transfer regulations · 26 CFR §1.704-1 · IRS Publication 541