The Partnership Flip: Why 99/5, What the DRO Is For, and When the Safe Harbour Doesn't Apply

A developer that builds a project earns tax credits and depreciation it frequently cannot use. A financial institution that can use them has no interest in building anything. The partnership flip exists to put those two parties in the same entity for long enough to move the benefit, and then to get the investor out again.

The familiar numbers — 99/5, a five-year minimum, a 4.95% residual — come from a wind safe harbour that the IRS has confirmed does not apply to solar. Most of the market structures solar flips against guidance written for a different credit, and then relies on general partnership principles instead.

This is the first post in a new series on tax structuring, and the first of three on tax equity structures. You will get how the allocation actually works, the difference between yield-based and time-based flips, what a deficit restoration obligation is and why it has grown, the five safe harbour requirements, and why solar sits outside them.

ℹ️ Note: This describes how these structures work in practice. It is not tax or legal advice — partnership allocations are highly fact-specific and depend on the full agreement.

What Problem Does a Flip Solve?

Ownership. Norton Rose Fulbright puts the constraint plainly: "Tax benefits can only be claimed by the owner of a project. Partnerships offer flexibility in how economic returns can be shared by the partners."

Those two sentences contain the whole design. The credit follows ownership, so the investor must genuinely own part of the project — a loan will not do. But ownership brings economics the investor does not want and the developer does not wish to give away permanently. A partnership resolves it by separating tax items from cash, and by making the arrangement temporary.

The developer finds an investor capable of using credits and depreciation. Both become partners. The partnership allocates tax items disproportionately to the investor until a target is reached, at which point the allocation flips and the developer can buy the investor out.

How Does the Allocation Actually Work?

Disproportionately, and differently for tax and for cash. Norton Rose describes the standard arrangement: "the partnership allocates 99% of income, loss and tax credits to the tax equity investor until it reaches a target yield. Cash is shared in a different ratio. After the yield is reached, the investor's share of everything drops to 5%."

The phrase to hold onto is "cash is shared in a different ratio." This is the single most misunderstood feature of the structure. A 99% allocation of tax items does not mean the investor takes 99% of the distributions. Tax items and cash run on separate splits, and the developer typically retains far more of the cash than its 1% tax allocation would suggest.

That separation is what makes the structure work commercially. The investor is buying tax capacity, not an operating business. The developer is giving up tax benefits it could not use, not the economics of its project.

Yield-Based or Time-Based?

Two mechanisms for deciding when the flip happens, and the market is heavily weighted to one.

	Yield-based flip	Time-based (fixed) flip
Flip trigger	Investor reaches a target after-tax return	A predetermined date
Market share	"All wind deals and about 80% of solar deals"	A minority — "at least one major investor"
Typical timing	Six to eight years in solar	"Usually after five to five-and-a-half years"
Who bears underperformance	The developer — flip is delayed	The investor — flip happens anyway
Sponsor certainty	Low: the date moves	High: the date is fixed

Norton Rose confirms the split: "All wind deals and about 80% of solar deals are yield-based flip transactions," while fixed-flip structures are used by a smaller number of investors who "flip to a 5% interest on a fixed date, usually after five to five-and-a-half years."

The risk allocation in the fourth row is the commercial heart of the choice. In a yield-based flip, if the project underperforms, the investor simply takes longer to reach its target and the flip is pushed out — the developer waits longer for its economics. In a time-based flip the date arrives regardless, and the investor absorbs the shortfall.

That is why a fixed flip is more expensive and why developers who can obtain one often prefer it. It converts an open-ended exposure into a known date.

What Is the Flip Measured Against?

A target after-tax yield on the investor's investment, computed across everything it receives — credits, depreciation deductions, taxable income allocations and cash distributions.

Two consequences follow, and both are frequently missed in a first model.

The yield is after-tax. Credits are worth their face value against tax; deductions are worth the marginal rate. A model computing the flip on pre-tax cash flows will reach the wrong date by a wide margin.

Underperformance is asymmetric. Because the flip is a yield hurdle rather than a date, a poor generation year does not reduce the investor's return — it delays the flip. The investor is largely insulated; the sponsor's back-end is not. The sensitivity worth running is therefore not the effect of underperformance on investor yield but its effect on flip timing and hence on sponsor IRR.

What Is Tax Equity Actually Pricing?

Three separate streams, of which only one is the credit. The investor's target yield is computed across the investment tax credit or production tax credits, the depreciation deductions, the allocations of taxable income or loss, and its share of cash distributions.

That composition explains two things that puzzle people coming to these structures from ordinary project finance.

The headline yield looks low. A target after-tax yield in the mid single digits would be unattractive for an equity investment bearing construction and merchant risk. It is not unattractive for a largely tax-driven return with a defined exit, much of which arrives as credits in the first year or two.

Depreciation is not a rounding item. Accelerated depreciation on a capital-intensive asset is a substantial part of what the investor is buying — which is precisely why the deficit restoration obligation matters, and why an investor that will not accept a large DRO pays less for the same project.

This is also the reason the flip survives alongside credit transfers. A transfer sells the credit alone. A flip monetises the credit and the depreciation, which is a materially larger pool of benefit. A sponsor with no tax capacity that sells only its credits has left the depreciation stranded.

What Is a Deficit Restoration Obligation?

A promise to put money in later so that losses can be taken now. Norton Rose defines it as "a promise to contribute more money to the partnership when the partnership liquidates to cover any negative capital account."

The mechanism exists because of a constraint in partnership tax: an investor can generally only be allocated losses to the extent it has capital account to absorb them. Accelerated depreciation on a capital-intensive asset generates deductions far larger than the investor's contribution, so without something more the allocations would stop precisely when they are most valuable.

A DRO solves that by making the investor liable to restore a negative capital account on liquidation. The investor can then be allocated losses beyond its contributed capital, because there is a real obligation standing behind the deficit.

Two features matter commercially.

They have grown. Norton Rose notes that "DROs today can reach 50% to 70% of the tax equity investment," and attributes the pressure directly: "falling wholesale electricity prices are forcing them to these levels."

They are enforced as real obligations. The IRS requires substance, including net worth documentation and "commercially reasonable provisions for enforcement and collection." A DRO that the investor could not actually honour is not a DRO.

ℹ️ Note: A DRO is a contingent liability of the investor, not of the project. But it shapes the project's structure — the size of the DRO an investor will accept determines how much depreciation can be allocated, which determines the price it will pay, which determines how much tax equity the deal can raise.

What Does the Safe Harbour Require?

Five conditions, set out in Revenue Procedure 2007-65. Norton Rose lists them:

• Residual interest. "The tax equity investor must retain at least a 4.95% residual interest after the flip."
• Minimum term. "The flip cannot occur more quickly than five years after the project goes into service."
• Purchase option pricing. "Any option to buy the investor's interest must be for fair market value or a fixed price that is a good-faith estimate at inception of what the fair market value will be at time of exercise."
• Investment timing. "The investor must make at least 20% of its total investment before the project is put in service."
• No put. "The investor cannot have a 'put' to require the sponsor to purchase its interest."

The underlying question all five serve is whether the investor is genuinely a partner bearing entrepreneurial risk, or a lender wearing a partnership costume. A guaranteed exit at a fixed low price, no money at risk before operations, and a right to force a buyout would describe debt. Each condition removes one of those features.

The purchase option rule connects directly to the termination values post earlier in this series: options struck at anything other than fair market value are disfavoured because they undermine the claim that the investor was ever the owner. The same logic, in a different document.

Why Doesn't It Apply to Solar?

Because the safe harbour was written for production tax credits, and the IRS has said so. Norton Rose records the position: "The IRS said in an internal memo released in June 2015 that the flip guidelines do not apply to solar projects or other projects on which investment tax credits are claimed. The memo said to apply general partnership principles to test whether the investor is really a partner."

That leaves solar in an odd place, and it is worth being clear about what it does and does not mean.

It does not mean solar flips are improper. It means there is no bright-line safe harbour, and the analysis reverts to whether the investor is genuinely a partner under general principles — substance, risk, and the reality of the economic arrangement.

It also does not mean the 2007-65 conditions are irrelevant to solar. In practice much of the market structures solar deals to look like the safe harbour anyway, because a structure resembling one the IRS has blessed for wind is easier to opine on than one that does not. The conditions function as market convention rather than as protection.

The practical consequence for diligence is that a solar flip cannot rest on "we satisfied the safe harbour." It has to rest on a partnership analysis, which is a longer conversation and a different kind of opinion.

What Does the Sponsor Actually Get Back?

A project it already owned, after waiting. That framing is worth keeping, because the flip is frequently described as though the sponsor gains something at the flip date — it does not. It recovers what the structure temporarily set aside.

Three things change at the flip. The sponsor's share of tax items rises from 1% to 95%. The cash split typically moves in its favour. And the purchase option becomes exercisable, allowing the sponsor to buy out the investor's remaining interest and own the project outright again.

That option is where the sponsor's terminal value sits, and its pricing is constrained rather than negotiated. It must be struck at fair market value, or at a fixed price representing a good-faith estimate made at inception of what fair market value will be at exercise. A sponsor cannot contract for a cheap exit, because a cheap exit is evidence the investor was never really an owner.

The practical consequence is that sponsor economics in a flip are heavily back-ended and somewhat open-ended. The near-term cash is reduced, the tax benefits are gone, and the value returns in a lump at a date determined by the investor's yield and at a price determined by a valuation nobody controls.

That profile is why sponsors care so much about flip timing, and why back-leverage exists.

How Does the Flip Interact with Debt?

Awkwardly, and the resolution is usually to move the debt somewhere else.

Tax equity sits inside the project partnership and is exposed to that partnership's economics. Debt raised at the project level ranks ahead of the partnership's cash flows, which reduces and delays the distributions the investor was relying on to reach its yield — and, in a downside, can divert cash entirely. Investors are correspondingly resistant to leverage sitting above them.

The market answer is back-leverage: debt raised not by the project company but by the sponsor's holding company, secured on the sponsor's membership interest and serviced out of the sponsor's share of distributions. The tax equity investor's position is unaffected because the borrowing sits outside the partnership entirely.

That structure is efficient and it relocates the risk rather than removing it. Back-leverage is serviced from the residual cash the sponsor receives after the partnership's own waterfall, which is the most volatile slice of the project's economics and the one most exposed to a delayed flip. Lenders to a back-levered position are therefore lending against exactly the cash flow that a yield-based flip puts at risk.

Two consequences for modelling. The debt sizing exercise covered later in this series has to be run on sponsor-level distributions, not project-level cash available for debt service. And the flip-date sensitivity is not merely an equity return question — it is a coverage ratio question for the back-leverage lender.

There is a related instrument worth naming: a tax equity bridge loan, which funds the portion of the tax equity commitment paid only at or after placed-in-service, and is repaid when that contribution arrives. It is short-dated and sized against a commitment rather than against cash flow, so it behaves differently from either project debt or back-leverage.

How Do You Model a Flip in Excel?

The modelling question is almost never "what does the investor get?" — that is fixed by the target yield. It is "when does the flip happen, and what does the delay cost the sponsor?"

The inputs

Assumptions, labelled as such:

Tax equity investment                 $75,000,000
Target after-tax yield                6.50%
Pre-flip tax allocation               99% investor / 1% sponsor
Post-flip tax allocation              5% investor / 95% sponsor
Pre-flip cash split                   shared on a separate ratio
Expected flip                         year 6
DRO cap accepted by investor          50% of investment

The DRO constraint

Max_DRO           = 75,000,000 × 50%                     = $37,500,000

This is a binding constraint on the structure, not an output. If the depreciation schedule drives the investor's capital account more negative than $37.5m, the excess losses cannot be allocated — which reduces the value of the deal to the investor and therefore the price it pays.

Flip timing sensitivity

Flip in year 6   → sponsor reaches 95% of tax items and full cash from year 6
Flip in year 8   → two additional years at 1% of tax items

Delay_Cost = PV(sponsor's foregone share over the delay period)

Run flip year across the columns against generation outcome down the rows. The shape to look for is that investor yield is nearly flat across the grid — that is the point of a yield-based flip — while sponsor IRR falls steeply. All the variance lands on one party.

The comparison worth putting to a board

Yield-based flip   → expected flip year 6, range 6 to 9 depending on performance
Fixed flip         → flip year 5.5, certain, at a higher cost of tax equity

Value_Of_Certainty = sponsor IRR difference between the expected case and
                     the downside case under the yield-based structure

That number is what a fixed flip is actually worth. It is usually larger than the price premium quoted for one, and it is rarely computed before the structure is chosen.

ℹ️ Note: Model tax and cash on separate allocation rows from the outset. Building a single "99%" line and applying it to both is the most common structural error in a first-pass flip model, and it overstates the sponsor's early cash by a wide margin.

To run the full version — flip date solved against a real after-tax yield, DRO tested against the depreciation schedule, and sponsor IRR across a generation distribution — prompt Dezzmond with your allocation terms and tax assumptions.

What Do Investors and Lenders Actually Check?

• Is the investor a genuine partner? Especially for solar, where the safe harbour does not apply and general principles govern.
• Is the purchase option struck at fair market value? A bargain option threatens the entire characterisation.
• How large is the DRO, and is it supported? Net worth evidence and enforceable collection provisions.
• What is the flip date under a downside generation case? Not the base case.
• Are tax and cash allocations modelled separately? They run on different ratios.
• Did the investor fund at least 20% before placed-in-service? A safe harbour condition and, for solar, a fact supporting partner status.
• Is there any put? Its presence undermines the structure.

Frequently Asked Questions

What does the 99/5 split mean?

The partnership allocates 99% of income, loss and tax credits to the tax equity investor until it reaches a target yield, after which its share of everything drops to 5%. Cash is shared on a different ratio throughout.

What is the difference between a yield-based and a time-based flip?

A yield-based flip occurs when the investor reaches its target return; a time-based flip occurs on a fixed date, usually after five to five-and-a-half years. All wind deals and roughly 80% of solar deals are yield-based.

Who bears the risk if the project underperforms?

In a yield-based flip, the sponsor. The investor still reaches its target yield, just later, so underperformance delays the flip rather than reducing the investor's return.

What is a deficit restoration obligation for?

It allows the investor to be allocated losses beyond its contributed capital, by promising to restore a negative capital account on liquidation. DROs today can reach 50% to 70% of the tax equity investment.

Does the flip safe harbour apply to solar?

No. The IRS confirmed in a 2015 internal memo that the guidelines do not apply to solar or other investment tax credit projects, and that general partnership principles apply instead to test whether the investor is really a partner.

Closing: A Structure Built Around One Sentence

Everything in a partnership flip follows from the fact that tax benefits can only be claimed by an owner. If credits could be sold outright, none of this would exist — no 99/5 allocation, no deficit restoration obligation, no five-year minimum, no fair market value option.

That is worth holding in mind, because credits increasingly can be sold outright. Transferability under section 6418, which the next sub-series takes up, does exactly what the flip was invented to work around, and does it without a partnership at all.

The flip has not disappeared and will not — it monetises depreciation as well as credits, which a transfer does not. But the two structures now compete, and the comparison is the subject this series builds toward.

Next: sale-leasebacks and the inverted lease, and when each is used instead of a flip.

Sources: Norton Rose Fulbright — Partnership Flips: Structures and Issues · Norton Rose Fulbright — Solar Tax Equity Structures · IRS — Revenue Procedure 2007-65