Form 8697

I have a client that needs to file Form 8697 for tax year (i.e. “Filing Year”) 2018. The only prior year at play is 2017 and only one long-term contract is at play. Call it Contract #1. Said contract started in 2017 and ended in 2018. Turns out, because of actual vs. estimated differences associated with this one long-term contract, he recognized $100k of income in 2018 that “should have been” recognized in 2017. So, we do the lookback computation, compute the added tax for 2017 and then compute the interest charge. Easy enough.

Client then proffers the following: “Hey Jeff, I had one long-term contract open at 12/31/18, which we’ll call Contract #2. This job started in 2018 and ended in 2019. Let’s say that when we go to File Form 8697 for “Filing Year” 2019, it turns out that on this one contract, I recognized $100k in 2019 that should have been recognized in 2018 because of actual vs. estimated differences. I won’t owe the lookback charge for Filing Year 2019 because if you recall, with respect to Contract #1 I recognized $100k too much income in 2018, right? According to my math, if I recognized $100k too much income with Contract #1 in 2018, and $100k too little income in 2018 from Contract #2, these cancel each other out, no?”

The 2019 over-recognition may cancel out the 2018 under-recognition but lookback is only done when the contract is complete. Things I always consider when doing lookback calculations are: 1) Type of entity 2) Type of work performed under the contracts 3) De minimis election 4) AMT (in the case of individuals with long-term contract adjustment on 6251) 4) average gross receipts. There are instances where there is no interest due because of one of these considerations.

By "lookback charge" I'll assume you mean the interest on line 10.

For 2018, he pays interest on the tax on the $100k that got reported in 2018 instead of 2017.

For 2019, he'll pay interest on the tax on the $100k that he'll report in 2019 instead of 2018.

Seems to me he and IRS are even at that point. No offset.

As an aside, if rates or brackets changed, and I'm guessing here, as long as the allocations to wrong years were just because of wrong estimates but made in good faith, the tax rate difference doesn't matter, i.e., doesn't get adjusted; it just affects the interest.

The 2019 over-recognition may cancel out the 2018 under-recognition but lookback is only done when the contract is complete.

Client’s actual 2018 TI (Taxable Income) was $225k. That amount includes the $100k recognized in 2018 with respect to Contract #1 (that should have been recognized in 2017 because actual differed from estimates). Client’s actual 2018 tax was, of course, computed on the $225k of TI.

Now pretend 2019 just ended, the year in which Contract #2 was completed. As it turned out, $100k of Contract #2 should have been recognized in 2018. So, when we do the lookback for “Filing Year” 2019, and we look back to 2018, we need to compute the increase to the 2018 tax that results from the $100k associated with the Contract #2 lookback.

Are you saying that when we do just that, we compute the tax differential between (1) 2018 actual TI of $225k and (2) 2018 redetermined TI of $325k…despite the fact that 2018 actual TI was inflated by the $100k associated with Contract #1 that “should have been” recognized in 2017?

For 2019, he'll pay interest on the tax on the $100k that he'll report in 2019 instead of 2018.

How do we compute the tax on that $100k of Contract #2 income?

Don’t we do this: Look at the 2018 tax return, as filed, and jot down the gross tax liability. Then, we hypothetically add $100k to 2018’s existing/actual Taxable Income and compute the tax on that higher amount. The difference between the actual tax (showing on the 2018 return as filed) and what that tax would have been had 2018 taxable income been $100k higher is the “tax increase.” Then we apply an interest rate to it.

Does that make sense?

One might argue that this piece:

Look at the 2018 tax return, as filed, and jot down the gross tax liability.

…involves inflated numbers. Remember, the actual TI for 2018 included $100k that “should have been” reported in 2017.

If we pretend the 2018 actual TI, as filed, was $225k…then which is the correct calculation to determine the 2019-to-2018 lookback:

Is it Option #1 (described above)…$225k actual TI for 2018 vs. $325k redetermined TI for 2018…or…

Is it Option #2…$125k “actual” TI [since the $225k real actual TI was inflated by $100k b/c of the 2018-to-2017 lookback] vs. $225k redetermined for 2018?

Either way, the delta is $100k. We are, as you say, computing the tax increase associated with the $100k Contract #2 income that we reported in 2019, but “should have reported” in 2018. But it matters a great deal what that $100k “is added to” (i.e. the base/benchmark) in computing the tax on the base and the tax on the “Base plus $100k.”

What do the form and instructions say? Also the Code and Regs.? You look at the latter and I'll do the former. Edit - changed my mind; form confusing. I'd like special dispensation to just apply logic that ends up being fair to both sides. OK, I grant myself that.

I should mention that I missed the math class that explained "delta". In fact, I missed whatever subset of math that's in.

I'll assume that the 2 contracts mentioned (#1 and #2, I did go to that class) were the only ones incorrectly estimated each year.

2017 reported 225, should have been 325 (225 + 100), pay interest on the 100.
2018 reported 225, should have been 225 (225 + 100 -100), no interest.
2019 reported 225, should have been 125 (225 - 100), get interest refund on 100.

Think about this: Let’s pretend 2019 doesn’t exist. What do we end up with? We end up pushing $100k of TI from ’18 to ’17 for lookback purposes (i.e. to compute the add’l lookback tax/interest charge). Then we’re done. No biggie, right? We improperly deferred $100k of income from ’17 and ’18 and now we have to pay the piper.

But what if our ’18 tax bracket was massive and our ’17 tax bracket wasn’t?

Are you telling me we’ll have to pay an interest charge even when, the fact is, we ended up paying way more tax because our estimates were off? It’s like we’re getting whacked for the deferral, but the deferral actually cost us way more in pure dollars than the dollars it saved because of the time value of money…but we get no credit for that.

Anyway, back to your numbers…

2017 reported 225, should have been 325 (225 + 100), pay interest on the 100.
2018 reported 225, should have been 225 (225 + 100 -100), no interest.
2019 reported 225, should have been 125 (225 - 100), get interest refund on 100.

Makes sense, but I don’t think it works that way.

We’ll assume we reported $225k actual TI each year, as you suggest.

I think the rules operate like this:

2017 reported 225, we redetermine it to be 325 for lookback purposes (225 + 100 pushed back to ’17 from ‘18).
2018 reported 225, we redetermine it to be 325 for lookback purposes (225 + 100 pushed back from ’19 to ’18).
2019 reported 225, should have been 225.

So, we end up paying an interest charge “for” 2017 and “for” 2018 based on an increased TI of $100k in each of these two years.

If this is right, my big beef is that when we go to compute the lookback charge “for” 2018, our starting point is 2018 TI as reported (or $225k). That doesn’t seem quite right. We compute tax @ $225k and then @ $325k for 2018, then we take the difference, then we apply an interest charge to that difference. Why don’t we get to compute tax on $125k and then on $225k? We’d end up with a lower tax differential, since we haven’t moved through the brackets as much. It’s like they make us look back and then ignore the forward looking fallout of that lookback exercise…

Is this the area of computation that concerns you or is it some other area?

§ 1.460-6 Look-back method.
(a)In general -

(1)Introduction. With respect to income from any long-term contract reported under the percentage of completion method, a taxpayer is required to pay or is entitled to receive interest under section 460(b) on the amount of tax liability that is deferred or accelerated as a result of overestimating or underestimating total contract price or contract costs. Under this look-back method, taxpayers are required to pay interest for any deferral of tax liability resulting from the underestimation of the total contract price or the overestimation of total contract costs. Conversely, if the total contract price is overestimated or the total contract costs are underestimated, taxpayers are entitled to receive interest for any resulting acceleration of tax liability. The computation of the amount of deferred or accelerated tax liability under the look-back method is hypothetical; application of the look-back method does not result in an adjustment to the taxpayer's tax liability as originally reported, as reported on an amended return, or as adjusted on examination. Thus, the look-back method does not correct for differences in tax liability that result from over- or under-estimation of contract price and costs and that are permanent because, for example, tax rates change during the term of the contract.

It is the area of computation that is troublesome. Or, should I say, imprecise.

The rules do a good job of identifying the issue: We’re using estimates, so we might end up deferring (or accelerating) tax. Fair enough. But the rules do not contemplate the corollary effect of the deferral or acceleration. The example I gave was a situation involving deferral, so we look back and compute the added tax “for the prior year” (call it 2017) and the interest charge thereon. If we were in a 10% bracket in the lookback year (2017), and $100k was deferred, the deferred tax is computed to be $10k. If that $100k deferred income was actually taxed in 2018 at a 30% rate, we end up paying $20k more in tax than we “should have.” The government actually benefited from the deferral here, in a real big way. Yet, we still owe the lookback charge and we get no relief. It makes very little sense. But, I suppose, if we flip-flop things such that we had a real high tax rate in 2017 and a real low rate in 2018, then the government gets screwed.

Thus, the look-back method does not correct for differences in tax liability that result from over- or under-estimation of contract price and costs

That’s pretty much my point. To do things my way would complicate things greatly. Doing things Congress’ way is illogical and imprecise. We have two evils. Does it make more sense to pick the lesser evil, which is probably Congress’ and not mine? Or, would it make more sense to not even have this provision in the first place?

Evil/schmevil. Congress did what it did. We are required to do what it says. What makes more sense depends on whose ox is being gored (or something like that). Is it all that different from a change in accounting method that gets spread over 4 years? Yes, it's different, but not on the point of tax effect in the various years, depending on both how the business income is going and whether tax rates have changed.

Were there no provision at all, taxpayers generally would be incentivized to weigh their estimates to more deferral, paying the tax in later years w/o interest. IRS would make lots more adjustments upon examination in order to correct this. Then we'd get into undersatement penalties too.

Actually, is there anything stopping IRS even under the existing protocol from examining and moving income around where there are really bad estimates, thereby getting penalty revenue?

Evil/schmevil. Congress did what it did. We are required to do what it says

Well, this is what Congress said in the Geneal Explanation of the TRA of 1986:

In the taxable year in which the contract is completed, a determination is made whether the taxes paid with respect to the contract in each year of the contract were more or less than the amount that would have been paid if the actual gross contract price and the actual total contract costs, rather than the anticipated contract price and costs, had been used to compute gross income

Note the words, “…in each year of the contract.” That would include the year of completion. However, the actual statute seems to depart, in a big way, from that Explanation:

Look-back method.

The interest computed under the look-back method of this paragraph shall be determined by—
(A) first, allocating income under the contract among taxable years before the year in which the contract is completed on the basis of the actual contract price and costs instead of the estimated contract price and costs,
(B) second, determining (solely for purposes of computing such interest) the overpayment or underpayment of tax for each taxable year referred to in subparagraph (A) which would result solely from the application of subparagraph (A), and
(C) then using the adjusted overpayment rate (as defined in paragraph (7) ), compounded daily, on the overpayment or underpayment determined under subparagraph (B).

I feel like I’m in the Twilight Zone…one thing saying one thing and another thing saying something different…

The stuff you excerpted from the Blue Book refers to the base on which to compute interest. It doesn't say to pay what the tax would have been with a correct estimate. If it did, it would be the same as requiring amended returns every year an estimate was off.

Which in no way affects what zone you are in.

The stuff you excerpted from the Blue Book refers to the base on which to compute interest.

…and that base (which includes all years of the contract) is far different than the base in the actual statute, which only includes those years before the contract was complete.

It doesn't say to pay what the tax would have been with a correct estimate.

Correct. But it does say that we are to determine:

whether the taxes paid with respect to the contract in each year of the contract were more or less than the amount that would have been paid if the actual gross contract price and the actual total contract costs, rather than the anticipated contract price and costs, had been used to compute gross income

The only way to figure that out is to include the year of completion in the calculation.

It pains me to admit that I have now concentrated enough to see your point. Are the examples in the regs. any help? Or the form?