Calculating How Much Tax Equity Can Be Raised

Calculating How Much Tax Equity Can Be Raised

June 01, 2008 | By Keith Martin in Washington, DC

Many developers of renewable energy projects in the United States are struggling to model their projects correctly so that they can calculate how much tax equity can be raised to help pay the project cost.

This requires including four blocks of figures in the computer model.

A developer can raise an amount of tax equity equal to the present value of four items discounted at the target internal rate of return required by the tax equity investor. The four items are the tax credits the tax equity investor will receive, the cash it will receive, its anticipated tax savings from depreciation and any interest deductions, less the taxes it will have to pay on its share of taxable income from the project. Target returns in the US tax equity market had dipped below 6% unleveraged and after taxes. They are headed back up and are currently in the mid-6% to low 7% range for one-off wind and solar photovoltaic projects. They are lowest for portfolios of projects where there is risk diversification across geography and by equipment type. They are 200 to 250 basis points higher in deals where there is project-level debt as the equity will require a premium against the risk that it will squeezed out of the deal before its target return is reached.


The chief financial officer of a renewable energy company must cover the capital cost of his or her project through a combination of true equity, tax equity and debt.

The US government pays as much as 63% of the capital cost of a typical wind farm and 56% of the cost of a solar project through tax subsidies.

Few developers are in a position to use the subsidies because of inadequate tax base.

The most common way to get value for them is through a “partnership flip” transaction. The developer brings in an institutional equity investor to own the project as a partner with the developer. The investor puts up a share of the capital for the project and is allocated 99% of the economic returns until it reaches a target internal rate of return, after which its interest drops usually to 5%, and the developer has an option to buy out the investor’s remaining interest for fair market value determined at the time. Cash may be distributed 100% to the developer until it gets back the capital it has in the deal, after which cash is distributed 99% to the investor until the flip.

In solar deals, the tax benefits might be transferred instead by selling the project to the institutional investor and leasing it back.

Lease structures do not work for wind farms and geothermal projects. They work in theory for biomass projects, but any lease would be an “inverted” lease where the investor is the lessee and the developer is the lessor.

The Internal Revenue Service issued guidelines for partnership flip structures in October 2007. The agency said it is okay with the structure, but that anyone straying outside the guidelines should expect to be subjected to “close scrutiny” on audit. The guidelines were addressed to partnership flip deals involving wind farms. However, the market has followed them in other types of projects.

At most, 99% of the tax subsidies can be transferred to an investor in a partnership flip deal.

In practice, the percentage may be smaller.

Any tax subsidies that cannot be transferred to the investor can be carried forward by the developer for up to 20 years.

Each partner in a partnership flip transaction must track its “capital account” and “outside basis.” These are different ways of measuring what each partner invested and took out of the deal. If either measure goes negative, then it is a sign that the partner took out more than his fair share. They are also a limit on the capacity of the investor to absorb tax benefits. Consequently, it is important to model both accurately.

Capital Accounts

A partner’s capital account starts with the cash he or she paid to buy into the deal or contributed to the partnership. It also includes the fair market value of any property contributed.

There are two forms of partnership flip deals.

In deals with larger developers who build projects on their own balance sheets, the investor usually pays a purchase price to the developer directly to buy an interest in a limited liability company that owns the project. This is called the “purchase model.” The investor usually waits to buy into the deal after it is already in service, except in solar deals where the investor will not be able to claim a 30% investment tax credit on the project unless he is a part owner before the project is in service.

In deals with smaller developers who must borrow from a construction lender to build a project, the investor commits at the start of construction to make a capital contribution at the end of construction in exchange for an interest in the project company. This is called the “capital contribution model.” The project company uses the capital contribution from the investor to pay down the construction debt.

In either case, the investor takes an opening “capital account” equal to what he pays the developer or contributes to the project company to buy into the deal.

The project company does not exist for tax purposes until the investor funds. Until then, it is usually a limited liability company with only one owner: the developer. Such companies are “disregarded” for tax purposes until they have at least two owners.

When the investor funds, the project company turns into a partnership for tax purposes. In the purchase model, the investor is treated as having purchased an undivided interest — or percentage of — the completed project from the developer and contributing it to a new partnership with the developer. The developer contributes the share of the project he retained. In the capital contribution model, the investor is treated as having made a capital contribution to a new partnership in exchange for an interest. The developer is treated as if he contributed the entire project.

In both models, the investor has an opening capital account equal to his cash payment.

In the purchase model, the developer has an opening capital account equal to the fair market value of the share of the project it retained. In order to calculate that, set up a fraction. The numerator is the amount paid by the investor. The denominator is the fair market value of the entire project. In most deals, the fair market value of the entire project is determined by a desktop appraisal at closing using discounted cash flows. The fraction is the share of the project the investor purchased; the developer retained one minus that fraction.

In the capital contribution model, the developer has an opening capital account equal to the fair market value of the entire project. However, he must subtract the opening capital account of the investor plus the amount of any term debt that will remain outstanding at the end of construction. By subtracting these amounts, the developer ends up with an opening capital account equal to the equity value he has in the project. His opening capital account is the claim he would have on the project assets if the partnership were to liquidate the next day. The lender would have a claim for the outstanding debt. The investor would have a claim for the capital the investor contributed. The developer has a claim for what is left.

Capital accounts are a fluid concept. They go up and down each year to reflect partnership results.

Add to each partner’s capital account at year end his share of income earned by the partnership. Subtract the losses he is allocated and cash he is distributed. In other words, increase the capital account each year as the partner suffers detriment; having to report income is a detriment (because taxes will have to be paid on that income). Reduce the capital account by the benefits the partner receives; being distributed cash or allocated losses is a benefit.

The income and loss that are reflected in capital accounts are “book” income and loss.

The “book” amounts are not the same as what is reported on financial statements. They are not the taxable income and loss that get reported on tax returns, either.

Rather, they are the income or loss computed at the partnership level the same way as the taxable income the partnership reports to the IRS, except that the project is depreciated by starting with its fair market value when the investor funds (and the partnership is formed) rather than the actual cost of the project. Otherwise, the depreciation is calculated the same way as tax depreciation. Thus, the only difference between “book” income and taxable income is the depreciation used to calculate book income may be a higher amount; it starts with the fair market value of the project rather than its cost.

A partner’s capital account serves two purposes.

First, it is the claim the partner will have on the assets in the partnership if the partnership liquidates.

Second, it is a limit on the amount of losses the partner can be allocated. A partner’s capital account cannot go into deficit unless the partner is willing to contribute additional capital to the partnership when the partnership liquidates.

Most investors in the tax equity market are willing to step up to such a “deficit restoration obligation”; however, they will agree only to contribute up to a fixed dollar amount. The dollar amount is the amount of deficit that the computer model suggests will reverse itself on its own under reasonably conservative assumptions about how the project will perform. For example, in wind farms and geothermal projects, the tax benefits are largely exhausted after 10 years. After that, the partners receive both cash and taxable income. If the income to be reported exceeds the cash — for example, because cash must be used to repay long-term debt — then the partners will have “phantom” income to report from the partnership. For example, a partnership might earn $100 from electricity sales, but have to use $80 to repay debt principal; the partners must still report the full $100 in income even though they are distributed only $20 in cash. The amount of this phantom income will increase their capital accounts. An investor will usually step up to a deficit restoration obligation in the amount of the aggregate phantom income expected over the remaining life of the project.

A rough rule of thumb used to be that tax equity covered 65% of the capital cost of a wind farm. The percentage has dropped recently to closer to 50%. This is due to increasing turbine costs; most turbines are priced in euros, and the euro has gained ground against the dollar. At the same time, the output of the project does not change, so the cash and tax credits on electricity output do not change, meaning the cost of projects has increased but without a commensurate increase in the amount the tax equity is willing to pay.

Another way to deal with deficit capital accounts is to pay part of the cost of the project with debt at the project level. IRS rules allow capital accounts to go into deficit to the extent they are driven into deficit by depreciation claimed against the share of project cost funded with nonrecourse debt. Thus, for example, suppose a project costs $100 and the cost is paid with $40 in equity from the partners and $60 in nonrecourse debt. The first $40 in “equity” depreciation will usually have to be shared in the same ratio the partners contributed equity, but the last $60 in “nonrecourse” depreciation can be shared in any ratio the partners wish, as long as the ratio is consistent with some “other significant item,” like the 99-1 ratio used to allocate other partnership items. (This is an oversimplification; it is discussed in more detail later.) The reason the nonrecourse depreciation can be shared 99-1 in favor of the investor is that US tax rules require the partners to report the later phantom income tied to repayment of the nonrecourse debt principal in the same 99-1 ratio. In other words, any deficit created by the nonrecourse deductions will reverse itself because it will be matched by future phantom income.

After calculating the capital accounts, the computer model should show the balance at year end in each partner’s capital account.

It should then have another line adding back the “nonrecourse” depreciation the partner was allocated. The next line should show the balance in the “adjusted capital account.” It is the adjusted capital account that cannot go into deficit unless the partner has agreed to a deficit restoration obligation.

If a partner has a deficit in his adjusted capital account that exceeds the deficit he has agreed to restore, then the standard partnership agreement shifts any losses he was allocated that year to the other partners to prevent a deficit.

There is a common misconception in the market that investors are able to absorb the tax subsidies fully from a project simply by stepping up to a large enough deficit restoration obligation. Stepping up to such an obligation may prevent losses from being shifted to another partner, but it does not ensure the investor will be able to use the losses fully. Even if he can keep the losses, his use of them may be suspended if he does not have enough “outside basis” to absorb them fully. Outside basis is the next block of figures that it is important to calculate.

If losses shift in a year because of inadequate capital account, then most tax counsel take the position that the shift will drag production tax credits with it. The US government allows production tax credits of 2.1¢ a kilowatt hour to be claimed on electricity from wind farms and geothermal projects for 10 years after the project is placed in service. Credits of 1¢ a kilowatt hour can be claimed on the electricity from a biomass project for 10 years. The credits must be shared by partners in the same ratio they share in “receipts” from electricity sales. The IRS has not said how to determine in what ratio “receipts” are shared; receipts are not the same thing as cash. Most tax counsel assume receipts are shared in the same ratio as net income and loss for the year. Thus, if net losses are supposed to be shared in a 99-1 ratio in favor of the investor, but the investor has too little capital account in a year to absorb the full net loss in that ratio with the result that part of the loss shifts back to the developer, the production tax credits will end up being allocated that year in the actual ratio that losses were shared.

Outside Basis

A partner’s outside basis is another potential limit on the ability of an investor to absorb tax subsidies. It is the same thing as his capital account, with three exceptions.

The investor’s opening outside basis is the same as his opening capital account — what the investor paid or contributed to the partnership — but the developer’s outside basis is his “basis” or cost of the share of the project he is treated as having contributed. Thus, in the purchase model, take the fraction of the project that the developer is viewed as having retained. Multiply the cost of the entire project by that fraction. The developer’s outside basis is that fraction times the original project cost. In the capital contribution model, the developer’s outside basis is the cost of the entire project, less the capital contributed by the investor and less the amount of term debt.

Outside basis goes up and down each year in the same way as capital accounts. Add income. Subtract cash distributions and losses allocated to the partner. However, instead of using the “book” income and loss, use taxable income and loss.

Finally, a partner’s outside basis includes not only what he contributed to the partnership, but also his share of any debt at the partnership level. Put differently, his capital account is just his equity in the deal. His outside basis is his equity plus his share of debt at the partnership level.

Each partner includes a share of partnership- or project-level debt in outside basis by working down a three-level waterfall. The model should recalculate the amount of debt in each partner’s outside basis at the end of each year. It should have a line showing the outstanding principal amount of the term debt there is to put in partners’ outside bases. Then give each partner first an amount of debt equal to the nonrecourse deductions he has been allocated to date and that have not been charged back. (How to calculate this is discussed below.) Next, give the developer an amount of debt equal to the “built-in gain” or appreciation there was in the share of the project he was treated as contributing when the partnership was formed. The built-in gain gets worked off over time, so the amount to put in the developer’s outside basis on account of built-in gain reduces gradually over time. (The concept of built-in gain is discussed later in the article.) Finally, the remaining debt is shared by partners in the same ratio that income is allocated (i.e., 99-1 initially in favor of the investor and then usually 5-95 after the flip).

If one of the partners or one of its affiliates makes a loan, then that debt must go entirely into its outside basis, and any depreciation tied to such a loan must also be allocated entirely to that partner.

The model should show each partner’s outside basis at year end.

Then it should have two more lines.

If the outside basis is negative, then the model should treat the cash the partner was distributed that year — to the extent needed to get the outside basis back to zero — as an “excess cash distribution,” meaning the partner must report it as capital gain. It is inefficient to be in such a position, since the partners will have already had to have reported the income in full that the partnership earned from electricity sales. When cash is later distributed to partners, it is not normally taxed again. An excess cash distribution is a form of double taxation.

If the partner still has a negative outside basis after converting all of the cash he was distributed into an excess cash distribution, then the model should suspend the use of any losses the partner was allocated that year to close the remaining gap. The partner keeps the losses, but he cannot use them until a later year when his outside basis goes back up.

If there is an excess cash distribution, then the model should increase the “inside basis” — or basis that the partnership has in the project — by the amount of the excess cash distribution. The partners’ capital accounts should also be increased by the same amount. However, the investor’s capital account will usually increase by 99% of the excess cash distribution if it occurs before the flip, and the developer’s capital account will increase by only 1% of it. The increase must bump up partner capital accounts in the same ratio that a gain in the same amount would have been reported by the partners.

It is important in deals where the term debt will remain outstanding after the flip in the investor’s interest to check whether the flip will cause an “excess cash distribution” to the investor. When the flip occurs, the investor’s share of partnership income will drop from 99% to around 5%. This will lead potentially to a large amount of debt being shifted from the investor’s outside basis to the outside basis of the developer. The mechanism by which this shift occurs is the investor is treated as if he was distributed an amount in cash equal to the debt that shifts. If this “deemed” cash distribution exceeds his remaining outside basis at the time, then he will have an excess cash distribution that must be reported as capital gain.

In some deals, the investor’s interests flips down to 5% in two stages to try to manage this problem. The first flip is to an intermediate sharing ratio that avoids such a deemed distribution.

Minimum Gain Chargebacks

“Minimum gain” is a fancy term for a simple concept. The model will need another block of figures to track minimum gain. The concept is as follows.

Suppose two partners form a partnership. The partnership builds a project at a cost of $100. It pays the cost with $40 in equity contributed by the partners and $60 in nonrecourse debt borrowed from a bank. The partnership will have $100 in depreciation. However, the partners are really only exposed to $40 in loss in value in the project, because they can always walk away and hand the keys to the nonrecourse lender. It is the bank that is exposed to the last $60 in depreciation; depreciation represents erosion in value of the project.

As a general rules, partners are only supposed to claim losses that they really suffer.

However, the IRS will let the partners claim the full $100 in depreciation in this case on one condition: they must agree to report the “phantom” income when the debt is repaid in the same ratio they claim the “nonrecourse” depreciation tied to the loan.

Therefore, in any deal where there will be term debt, the model must track the amount of “nonrecourse” depreciation and how it was allocated to the partners.

This is simple enough to do. There should be a line showing the outstanding debt principal. Next, a line should show the inside basis, or unrecovered “book” basis that the partnership has in the project. There is no “minimum gain” until the inside basis in the project drops below the remaining debt principal. At that point, the lender is exposed in theory to a loss if the project company walks away from the project and hands the lender the keys. The shortfall, or potential loss, is the “minimum gain.”

At each year end in which the minimum gain increased, the amount of the increase is the amount of book depreciation the partners were allocated that year that is considered nonrecourse depreciation, or depreciation that reflected an erosion in value to which the lender is exposed. In the first year in which the gap starts to narrow, the partnership must “charge back” income to the partners in the amount of the decrease. This income must be reported by the partners in the same ratio they were allocated the nonrecourse deductions earlier. These chargebacks are not additional income. They are simply a direction that the first amount of income that year must be shared by partners in the same 99-1 or other ratio that they were allocated the nonrecourse depreciation earlier. The remaining income for the year is allocated according to the business deal.

The point is that in any deal where term debt will remain outstanding after the flip, the investor will be allocated phantom income as the debt principal is repaid in a 99-1 ratio. It will be allocated income in that ratio at a time — after the flip — when it is being distributed only 5% of the cash.

This will take the investor’s return backwards.

Investors in such situations insist on one of at least two fixes. One is the investor might insist on being distributed enough cash to cover his taxes. He will also need additional income to restore his capital account. (Cash distributions reduce his capital account; income pushes it back up.) It is an iterative calculation to figure how much additional cash and income the investor needs to remain whole.

Alternatively, the investor might only credit the time value of the nonrecourse depreciation in determining when he reaches his target return (rather than treat each dollar of depreciation as saving him 35¢ in taxes). It should be possible to calculate when the depreciation will reverse due to chargebacks.

Built-In Gain

There is one more concept that must be reflected in the model in order for it to work properly. It is called “section 704(c) adjustments.” Once again, the concept is simple.

Suppose two partners form a 50-50 partnership. The deal is that each must contribute $100. Partner A contributes $100 in cash. B contributes an asset worth $100 but that has been fully depreciated. This is not a fair deal for A, since the partnership will have a gain of $100 one day when it sells the asset that B contributed, and A will have to pay taxes on 50% of that gain.

Section 704(c) of the US tax code addresses this by requiring B to pay taxes on the full $100 in gain when the asset is sold. However, it also requires B to try to make it up to A in the meantime without waiting for the asset to be sold. B makes it up to A by shifting depreciation to which B would otherwise be entitled to A until A has received $50 in deductions.

In many partnership flip deals, the developer is viewed as contributing appreciated assets, and there is not enough depreciation to shift to make A, the investor, whole.

There are three ways to make section 704(c) adjustments. Under the “traditional” method, the partnership allocates A the full amount of tax depreciation each year up to the “book” depreciation that A was allocated. For example, suppose the partnership has $20 in book depreciation, but only $15 in tax depreciation in a year, and the business deal is the investor gets 99% of everything until the flip. The investor gets 99% of the book depreciation, or 99% x $20 = $19.80. The investor also gets all $15 in tax depreciation. Since there is not enough tax depreciation to shift, when the partnership sells the project or liquidates, any remaining built-in gain that was not worked off by shifting tax depreciation will have to be allocated to the developer.

The traditional method is common in the purchase model where the investor buys an interest in the deal by making a payment directly to the developer. In the contribution model, the “remedial method” is more common. Under the remedial method, the investor gets an amount in tax losses equal to the “book” depreciation he is allocated. If there is not enough tax depreciation, the investor still claims a tax loss equal to its book loss, and the developer must report an offsetting amount of taxable income. For example, in the example earlier where the investor is allocated $19.80 in book depreciation in a year, but there is only $15 in total tax depreciation, he would be able to claim a tax loss of $19.80 under the remedial method. The developer would have to report income equal to the gap of $19.80 minus $15 = $4.80.

Use of the remedial method has the effect of requiring the developer to pay taxes on any appreciation in the project when the tax equity funds over the same period the project is depreciated.

Pre-Tax Return

The model should also calculate the pre-tax return for the investor. Most investors require a pre-tax return of at least 2%. This means that the cash and production tax credits the investor is projected to receive, when discounted at 2%, must equal or exceed the amount of his investment. Investors want at least this level of return to ensure that they are not viewed as having invested solely for tax benefits. The IRS confirmed in guidelines in October 2007 that production tax credits can be treated as equivalent to cash. They are a substitute for higher electricity prices.

Most equity investors appear also to treat the investment tax credit in solar deals as a cash equivalent for purposes of the pre-tax return test. The IRS has not taken a position yet.

Solar Issues

In solar projects, the parties are entitled to an investment tax credit in place of the production tax credits that are claim in wind, geothermal and biomass projects. The investment credit is claimed in year one when the project is placed in service. It is 30% for projects placed in service by December 2008. It drops to 10% after 2008. Congress is expected to extend it at a 30% level, but may not get around to doing so until 2009.

The credit must be shared by partners in the same ratio they share in income in the year the project is placed in service. However, because solar deals generate losses for at least four years due to depreciation allowances, it is important to hold the 99-1 sharing ratio used in the first year in place for at least a year after the deal starts generating income, lest the IRS argue that the 99-1 ratio for sharing income was illusory.

Investment credits vest over five years at the rate of 20% a year. If the solar project is sold or the investor disposes of his interest in the partnership during the first five years, then any investment credit he was allocated will be recaptured to the extent it has not yet vested. A reduction by more than a third in a partner’s sharing ratio for income will also lead to recapture of any unvested investment credits.

A partnership must reduce its basis in any solar project by half the investment credit on the project. For example, if a project cost $100 and it qualifies for a $30 solar credit, then only $85 can be recovered through depreciation. The depreciable basis is reduced by $15, or half the solar credit. The partners must also reduce their outside bases and capital accounts by the same $15. They do so in the ratio they are allocated the tax credit. If the tax credit is later partly recaptured, then the basis adjustment is commensurately reversed.

Keith Martin