The Reliability of Fair Value versus Historical Cost Information: Evidence from Closed-End Mutual Funds
Presented by:Ira GeraldinaNova Novita
Intan Oviantari
Outline
Introduction
Research Questions & Theoretical Frameworks
Methodology
Results
Conclussions
Introduction
JWA, Basis for Conclusions:
fair value more relevant than historical cost
fair value represents unbiasedmeasured (consistent within year
and between enterprise) economic effect of risk is
reflected in current periodincome
Fair value is measured at marketprice or based on estimates
(FASB)
Recognizing fair value in the Balance Sheet (JWG)
Subjective in estimate far valuefor non-traded instrument (IASC)Proponents of fair value measurement:
• Fair value valuation has more advantages than historical cost
Opponents of fair value measurement:• Reliability problems for some fair
value estimates•Suggest that these instruments are
recognized at historical cost
Research Questions
Primary Research Questions:
When compared to historical measures, is the valuerelevance of some financial instrument (Closed-End
Mutual Funds) fair values eliminated due toreliability issues relating to fair value estimation?
Research Questions-1
ONA
HC
FV
PRICE
Research Questions-2
INC
GRL
FVGL
RETURN
Research Questions-3
Reliability across fund types:
G7PUB NG7FUB
GOV
CBONDS
OTHER
UNCLASS
Theoretical FrameworkPrevious Studies
Barth (1994) • Unable to consistently document incremental explanatory power for fair value securities gains and loses• Found significant incremental explanatory power for fair value disclosures
Ahmed & Takeda (1995)
• Include a variable controlling for the interest rate sensitivity of other balance sheet items• Found that bank’s fair value securities gains and loses exhibit incremental explanatory power over historical cost measures of realized securities gains and loses
Petroni & Wahlen (1995)
• Use property casualty insurance companies• Performing test to examine whether there are cross- sectional differences for securities traded in active and thin market• do not test on fair value and historical cost gains and losses because of insufficient data
Carroll at al (2002) Vs Previous Studies
Carroll at al (2002) used closed end mutual funds which hasseveral advantages:
Provide sufficient data to evaluate the incremental informativeness of both fair value balance sheet and
income statement information
Potential problem with correlated omitted variablesIs basically nonexistent
There is great variation in the securities invested in bydifferent funds
The results support the hypotheses
Solve the problem of previous studies about omitted variables
Explain that reliability problems in measuring the fair valuesof investment are not the primary explanation for the
inconsistency in prior results
General Description
Mutual Funds
Open-End
Closed-End
General Description
Open-End:
Companies issue and redeem shares at the current market value
per share
There is no secondary trading market
Equity capitalization and netassets are directly changed bymagnitude of the traded share
price when shares sold or redeem
Closed-End:
•Companies operate more likea traditional companies
• Shares are available only fortrade in secondary market
•Subsequent purchases and salesof shares do not effect the equity
and assets of the fund
The traded price may fluctuate inresponse to supply and demand
• Trade at discount/premium
General Description
t
ttt NAV
PNAVD
Discount/premium has been widely studied over 30 years
Early explanation of discount/premium focused on problem
of measuring NAV
They derive from managerial performance that are not reflected in NAV(Boudreaux, 1973; Ingersoll, 1976; Chance, 1997)
Discount/premium explanations
Tax liability on unrealized capital appreciation does not show up in NAV (Malkiel, 1977)
Liquid asset are overvalued in NAV (Malkiel, 1977)
Indicate mispricing regarding individual investor sentiments causesystematic noise trading (Lee, Shleifer, Thaler, 1991)
Mispricing exist in the presence of rational traders who taking fulladvantages of mispricing
Accounting Requirement
Accounting Requirement
Accounting Requirement
Sample selection and Descriptive Statistics
Descriptive statistics
Table 2
Empirical tests
Empirical tests
Empirical tests
Investment Securities Fair Value Gains and Losses
Reliability Test
Reliability Across Fund Type
Investment Securities Fair Value Gains and Losses Test
Using Earning Capitalization Model to determine if investor view income statement fair value security
gains and losses as value relevance in a setting with no correlated omitted variables.
Fair Value Income
INC FVGL
Earning Capitalization Model
Rit = γ1t + γ2 INCit + γ3 RGLit + γ4 FVGLit + εit² (3)
Tabel 4. Investment Securities Fair Value Gains and Losses Test
Mean Estimate
Mean t statistic
Z statistic
γ1it 0.01 0.07 0.12
γ2 it 0.93 2.92 4.32**
γ3it 0.11 0.63 0.79
y4it 0.87 8.70 6.30**
Mean Estimate
Mean t statistic
Z statistic
γ1it Not tabulated
γ2 it 0.85 0.14 6.26**
γ3it -0.03 0.13 -0.25
γ4it 0.89 0.07 12.14**
Panel ASeparately Yearly Regression
Panel BPooled Estimation
Rit = γ1t + γ2 INCit + γ3 RGLit + γ4 FVGLit + εit²Mean R² 0,68Mean n= 65
White’s (1980) Adjusted Standar Errors as needed
Rit = γ1t + γ2 INCit + γ3 RGLit + γ4 FVGLit + εit²Mean R² = 0,67Mean n= 971
Newey-West (1987) Adjusted Standar Errors
In the close-end mutual fund setting, Carrol et al find that fair value securities gains and losses consistently provide information incremental
to historical cost income measures
Reliability Across Fund Types Test
Rit = γ1t + γ2 INCit + γ3 RGLit + γ4 FVGLit + εit² (3)Pit = β 1t + β2t ONAit + β3t HCit + β 4t FVit + it² (2)
The researcher examine the incremental informativeness of balance sheet and income statement fair value s by modifying the price and returns regressions
specified previously to allow the intercept and all of the coefficient to vary across the six fund types.
Tabel 5. Reliability Across Fund Types Test
ONA HC FV
G7PUB 1.08(11.89)**
0.09(0.68)
0.93(7.54)**
NG7PUB 0.12(0.24)
0.09(0.64)
0.78(8.35)**
GOV 1.01(8.16)**
-0.17(-1.34)
1.18(8.40)**
CBONDS 0.85(13.13)**
-0.21(-1.56)
1.18(9.26)**
OTHER 0.68(13.57)**
-0.10(-0.74)
1.00(7.89)**
UNCLASS 0.85(5.02)**
-0.10(-0.34)
1.14(4.60)**
INC RGL FVGL
0.61(1.70)
-0.30(-1.75)
1.14(10.63)**
0.23(0.46)
0.09(0.39)
0.84(10.38)**
1.18(3.12)**
1.28(3.12)**
0.41(2.00)*
2.20(4.69)**
-0.09(-0.52)
0.89(5.46)**
0.60(2.29)*
0.30(1.49)
0.50(3.72)**
1.52(6.62)**
-0.01(0.06)
1.04(8.62)**
Price Regression
Return Regression
Pit = β 1t + β2t ONAit + β3t HCit + β 4t FVit + it²Mean R² = 0,94Mean n= 1118
Rit = γ1t + γ2 INCit + γ3 RGLit + γ4 FVGLit + εit²Mean R² = 0,67Mean n= 971
Newey-West (1987) Adjusted Standar Errors
Coefficient Estimate t statistic based on Newey-West (1987) adjusted standar errors
Reliability Across Fund Types Result
Table 5 result consistent with fair value providing information incremental toHistorical cost for both balance sheet and income statement information
This finding is consistent with investor perceiving that the fair value estimatesfor equities traded in less active markets, such as private markets and markets in non G7 countries are less reliable than fair value of
publicly traded equity securities from G7 countries that are traded in active markets
Conclusion
• Investment securities fair values provide relevant information to close-end fund investors : a significant association between stock price and fair value securities gain and losses
• Provide evidence consistent with close-end investors viewing fair value estimates for G7 publicly traded equities as more reliable than those for private and non-G7 equities.
• The reliability problem in measuring fair value of investment securities are not the primary explanation for the inconsistency in prior research result, instead incomplete availability of fair value measures in other settings.
• Suggest to standard setters : their greatest challenge in requiring the recognition of all financial instruments at fair value may not be the reliable estimation of the fair value of investment security assets not traded in active market.
• This economic problem is virtually absent in their setting because all the net assets of close-end mutual funds are recognized in financial statements at their fair value or its equivalent. .