Changes in Tax Reserves in Anticipation of FIN 48

1
Changes in Tax Reserves in Anticipation of FIN 48

• Jennifer Blouin, Cristi Gleason,
• Lillian Mills and Stephanie Sikes

2
Setting the stage what are tax reserves?

• Firms pay lower tax TODAY due to tax positions
  claimed.
• The corporation might have to pay more tax
  SOMEDAY as a result of tax enforcement.
• Increasing reserves reduces financial earnings
  TODAY.
• Pre-FIN48, practice varied widely. Management
  discretion. Possible earnings management.
• Gleason/Mills02, Dhaliwal/Gleason/Mills04,
  Blouin/Tuna07

3
Setting the stage FIN48 Background

• FIN 48 issued July 13, 2006. Effective for
  calendar companies on January 1, 2007.
• More conformity in measurement, new disclosure
• Probable effect of
• New recognition rules, absent prior earnings
  management
• ? ? reserves
• New disclosure rules ? reserves
• Change in unrecognized tax benefit (UTB) due to
  FIN 48 adoption should be recorded in
  stockholders equity at adoption.

4
Research question

• Do firms adopt FIN 48 opportunistically?
• Do firms with excess reserves at FIN 48 enactment
  release reserves through earnings in quarters
  prior to FIN 48 adoption rather than as a
  stockholders equity adjustment at adoption?

5
Motivation

• Provides evidence of the degree to which tax
  reserves provided a cookie jar.
• Use or lose setting permits powerful test of raw
  earnings incentives.
• FASBs use of stockholders equity method for
  changes creates timing incentives that may
  generalize to other settings.

6
Data

• 100 largest 100 smallest non-financial/non-utili
  ty firms with a following of five or more
  analysts.
• We used large/small dummy to proxy for likely
  over/under reserve in preliminary tests.
• Collected tax reserve disclosures from 10-Q/10-K
  for 2005-2007Q1.

7
Data Disclosed Reserve Changes

• Large Firms Small Firms
• ? ? ? ?
• 2005 Annual 5 36 5 8
• 2006 Q1 2 8 0 2
• 2006 Q2 2 18 1 2
• 2006 Q3 4 20 0 4
• 2006 Q4 3 36 4 5
• 2006 Annual 6 54 3 11
• FIN48 Adoption 41 39 39 5
• 2007Q1 10 13 6 0
• 1/1/07 aggregates 78 billion balance,
• (2 billion) net decrease.

8
Research Design

• Logit regression
• DecreaseReserve (preFIN48) a0 a1
  Over-reserved at Enactment a2
  DiscloseSettlement a3LogSales
• a4 Jurisdictions a5 LastYearClosed
• a6 UTB/Sales a7 ETR
• a8(UTB-Expected Loss)/Sales e
• H1 predicts positive a1.

9
Results

• Main result Over-reserved at enactment is
  positively associated with Decrease in Reserve in
  2006Q34 (p lt 0.01).
• Control variables
• DiscloseSettlement is positive and significant (p
  lt 0.01).
• LastYearClosed and ETR are each negative and
  significant (p lt 0.10).

10
Robustness Tests

• Robust to
• replacing dependent variable w/expected rather
  than actual release.
• using excess reserve at adoption rather than
  imputed excess at enactment.
• using continuous dependent and explanatory
  variables.
• including industry indicator variables.
• eliminating loss firms.
• eliminating firms with zero reserve at enactment.
• Weak evidence that releases are more likely if
  over-reserved firms would have missed Q4
  consensus forecast without decrease in tax
  expense.

11
Conclusions

• Over-reserved firms were more likely than other
  firms to release tax reserves through earnings in
  the quarters preceding FIN 48 adoption rather
  than through stockholders equity at adoption.
• Amounts are economically significant. Firms that
  appear to have excess reserves at enactment
  released about half of their aggregate 8.6
  billion of excess reserves in Q3 Q4 of 2006.
• This reporting behavior is contrary to the FIN 48
  transition rules, but consistent with preference
  to record good news in earnings but bad news
  through stockholders equity.
• Results may generalize to other changes where
  FASB provides for equity effects at adoption.

12
Concepts and Constructs
Historical Earnings Incentives and IRS
probabilities
Opportunistic Reporting Surrounding FIN48
Adoption
Conceptual
Over- Reserved at FIN 48 Enactment
Decrease Reserves through Earnings in
Quarters Prior to Adoption
Operational
Settlement, Firm size, of Jurisdictions, Last
Year Audit Closed, ETR, (UTB-Expected Loss)/Sales
Control
13
Over-reserved at Enactment
FIN 48 Announcement
2005Q4
2006Q2
2006Q1
2006Q3
2007Q1
2006Q4 1/1/2007
accumulation period to determine over-reserved at
enactment
14
Recording Tax Reserves

• Suppose that a company estimated that it
  currently owed 100 million in taxes to the IRS.
• The total tax liability was reduced by 15
  million attributable to an uncertain (aka
  aggressive) tax position that the firm thinks
  it has a 40 chance of sustaining if discovered
  and audited by the tax authorities.
• Basic Entry
• DR Income Tax Expense 100
• CR Income Taxes Payable 100
• Tax Reserve Entries

Pre-FIN 48 Firm may have reserved anywhere from
zero to 15 assume 4 DR Income Tax Expense
4 million CR Tax Reserve 4
million
Post-FIN 48 Since less than 50 chance of
prevailing DR Income Tax Expense 15 million
CR Tax Reserve 15 million
15
Recording Tax Reserves

• Suppose that a company estimated that it
  currently owed 100 million in taxes to the IRS.
• The total tax liability was reduced by 15
  million attributable to an uncertain (aka
  aggressive) tax position that the firm thinks
  it has a 40 chance of sustaining if discovered
  and audited by the tax authorities.
• Basic Entry
• DR Income Tax Expense 100
• CR Income Taxes Payable 100
• Tax Reserve Entries

Pre-FIN 48 (FAS 109, FAS 5) Firm may have
reserved anywhere from zero to 15 assume
4 DR Income Tax Expense 4 million CR Tax
Reserve 4 million