Title: Fair Value Accounting, the Financial Crisis, and U'S' Standard Setting
1Fair Value Accounting, the Financial Crisis, and
U.S. Standard Setting
- Wayne R. Landsman
- Capital Markets Institute
- Rotman School of Management
- University of Toronto
- April 23, 2009
2Overview
- Introduction
- Background of Fair Value Accounting in U.S.
Standard Setting - Are Fair Values Useful to Investors? Evidence
from research - Fair Value Implementation Issues
- Fair value accounting and the financial crisis
3Introduction
- Events surrounding the financial crisis caused
many to question fair value measurement for
financial instruments. - SFAS 157 focus on exit value caused banks to
sell assets because of writedown to distress sale
levels - Procyclicality resulted
- Before addressing this criticism, it is helpful
to review empirical literature on fair value
accounting with a focus on highlighting findings
of interest to practitioners and standard setters - I begin with brief review of fair value
accounting and standard setting -
4Fair Value Accounting in Standard Setting
- Fair Value Definition
- FASB defines fair value as the price that
would be received to sell an asset or paid to
transfer a liability in an orderly transaction
between market participants at the measurement
date - The objective of a fair value measurement is to
determine the price that would be received to
sell the asset or paid to transfer the liability
at the measurement date (an exit price).
Objective is to estimate exchange price - Implicit in this objective is that exchange price
fully captures an instruments value - In practice, fair value may not be well defined
because no active market exists for the asset or
liability. - Difficult to disentangle an asset or liabilitys
fair value from its value-in-use to the entity. - Implication of this is that the value of a swap
derivative to a bank does not depend on the
existing assets and liabilities on its balance
sheet - A pretty strong assumption!
5Fair Value Accounting in Standard Setting
- Applications to standard setting
- US (FASB)
- Disclosure
- SFAS 107 Disclosures about fair value of
financial instruments - SFAS 119 Disclosure about derivative financial
instruments and fair value of financial
instruments - Recognition
- SFAS 115 Accounting for certain investments in
debt and equity securities - No. 133 Accounting for derivative instruments
and hedging activities
6Fair Value Accounting in Standard Setting
- Valuation Techniques
- FASB issued SFAS 157, Fair Value Measurements,
which describes hierarchy of preferences for
measurement of fair value - Level 1 quoted prices for identical assets and
liabilities - Level 2 quoted prices for similar assets and
liabilities - Level 3 company-based estimates
- Firms should use market prices as model inputs
wherever possible (e.g., equity prices for inputs
to B-S model to estimate employee stock options
fair values) - Can be constructed using entity-supplied inputs
(e.g., discounted cash flow estimates) if other
models employing market inputs are not available
7Fair Value Accounting in Standard Setting
- Valuation Techniques
- Critics of SFAS 157 express both conceptual and
practical concerns - Conceptual Concern exit value may not
appropriately capture the value of an asset (or
liability) to a firms shareholders even if an
active market exists for the asset - Occurs if significant divergence between V-in-Use
and exit value. For example, V-in-Use reflects
how an asset is used in conjunction with other
assets with which it is combined to generate
income exit value ignores this interaction. - Practical Concern Because active markets may not
exist for an asset or liability, FVs will often
be based on Level 2 and Level 3 estimates - Subjective, subject to manipulation, and
difficult to verify (audit).
8Fair Values Capital Market Evidence
- Policy question typically addressed is whether
recognized or disclosed accounting amount is
relevant to investors and measured with
sufficient reliability to be incrementally
informative to investors - Policy question is often operationalized using
value relevance regressions, testing for
incremental association of accounting amount
under study in explaining cross-sectional
variation in equity share prices - Example Landsman (1986) pension study
- MVE a0 a1MVA a2MVL a3PA a4PL
9Fair Values Capital Market Evidence
- US-based Research
- Investment Securities
- Barth (1994) -- pre-SFAS 115
- Investment securities fair values are
incrementally associated with bank share prices
after controlling for investment securities book
values - But, mixed results for whether unrecognized
securities gains and losses provide incremental
explanatory power relative to other components of
income - Could be a lack of reliability of fair value
estimates or omitted variable bias (fair value
changes of other balance sheet amounts).
10Fair Values Capital Market Evidence
- US-based Research
- Investment Securities
- Barth, Landsman, and Wahlen (1995) pre-SFAS 115
- Lends support to the measurement error
explanation by showing that fair value-based
measures of net income are more volatile than
historical cost-based measures - However, incremental volatility not reflected in
bank share prices - Banks violate regulatory capital requirements
more frequently under fair value than historical
cost accounting and fair value reg capital
violations incrementally predict future
historical cost violations
11Fair Values Capital Market Evidence
- US-based Research
- SFAS 107
- disclosed investment securities, loans, deposits,
long term debt fair value estimates - Barth, Beaver and Landsman (1996), Eccher,
Ramesh, and Thiagarajan (1996), and Nelson (1996) - Focusing on BBL, basic model is
- MVE-BVE a0 a1 (FV_ISEC - BV_ISEC)
- a2 (FV_LOANS - BV_LOANS)
- a3 (FV_DEP - BV_DEP)
- a4 (FV_LTDT - BV_LTDT) e
12Fair Values Capital Market Evidence
- US-based Research
- SFAS 107
- BBL (1996) finds
- ISEC and LOANS fair values incrementally
informative to their book values - Loans fair value capture dimensions of default
and interest rate risk - Investors discount loans fair value estimates
made by less financially healthy banks (those
with relatively low regulatory capital) - Consistent with investors seeing through attempts
by managers of less healthy banks to exercise
discretion when estimating loans fair values
13Fair Values Capital Market Evidence
- International Research
- UK and Australian GAAP asset revaluations
- Easton et al. (1993) estimates annual return
regressions for Australian firms (1981-90) and
finds tangible asset revaluations have
incremental explanatory power relative to
earnings and change in earnings - Rit a0 a1 Eit a2 ?Eit a3 REVALit
eit - Barth and Clinch (1998) estimates annual stock
price regressions for Australian firms (1991-95)
to determine if financial, tangible, and
intangible asset revaluations have incremental
explanatory power relative to earnings and equity
book value (less book value of revaluated
assets) - Pit a0 a1 BVEit a2 Eit a3 REVALit
eit - Find investments and intangible revaluations are
incrementally priced, but those for tangible
assets are not - Little difference in value relevance between
appraiser and director-based estimates --
potential good news regarding level 3 estimates
14Fair Values Capital Market Evidence
- International Research
- UK and Australian GAAP asset revaluations
- Aboody, Barth, and Kasznik (1999) examines the
performance prediction and pricing implications
of fixed asset revaluations for UK sample
(1983-95). - Upward asset revaluations predict higher future
operating income and cash from operations - Current year revaluations also positively related
to annual stock returns (Rit), and current year
asset revaluation balances are significantly
positively related to annual stock prices (Pit). - Relations are weaker for higher debt-to-equity
ratio firms. - Consistent with managerial discretion
(manipulation?) affecting the usefulness of asset
revaluations made by managers of firms facing
financial distress pressures.
15Fair Values Capital Market Evidence
- International Research
- UK and Australian GAAP asset revaluations
- Black, Sellers, and Manly (1998) finds no
difference in earnings management behavior for
UK, Australian, and NZ firms (1985-95) for asset
revaluing and non-asset revaluing firms. - Evidence consistent with mitigation of incentive
to time asset sales for earnings management
purposes if gains/losses are recognized in income
when assets are revalued and gains on sale are
based on FV rather than HC. - Findings do not hold for UK firms pre-93, when
asset revaluing firms were permitted under UK
GAAP to include in income gains/losses based on
HC.
16Fair Value Implementation Issues
- Estimating FV (exit value) for assets and
liabilities is relatively easy if they are
actively traded in liquid markets. Issue is
complicated if - Active markets for financial instrument do not
exist - Exit value and v-in-use differ substantially, in
which case it becomes hard to disentangle asset
values. - For example, for even a single financial
instrument with embedded options, values of each
option depend on inter-related default and price
risk characteristics - What does exit value mean in this case?
- US Treasury paid 66 for 100 of bank assets
under TARP. Where did this level 3 price come
from? -
-
17Fair Value Implementation Issues
- Manipulation of Model Inputs
- Relying on managers model estimates of Level 3
fair values introduces informational asymmetry
problemsadverse selection and moral hazard
(information asymmetry also for Level 1 and 2
estimates) - Adverse selection
- Market cannot distinguish high and low
quality firms. E.g., how do investors
distinguish which banks have relatively healthy
portfolio of loans and which do not if credible
information regarding assets quality is
unavailable
18Fair Value Implementation Issues
- Manipulation of Model Inputs
- Relying on managers model estimates of financial
instruments fair values introduces informational
asymmetry problemsadverse selection and moral
hazard - Moral hazard
- Managers will tend to use private information to
their advantage (manage income to maximize
compensation -- by manipulating information they
disclose to securities markets and regulators). - Evidence that managers cannot resist this
temptation even in the case of model inputs
relating to unrecognized ESO expense (ABK, 2006)
19Fair Value Implementation Issues
- Manipulation of Model Inputs
- If fair value accounting is generally applied for
financial statement recognition, accounting
standard setters and securities regulators must
determine how to balance the benefit of
permitting managers to reveal private information
(mitigating adverse selection problem) and the
cost of managerial manipulation of earnings and
balance sheet amounts when selecting model inputs
(moral hazard cost)
20Fair Value Accountingand the Financial Crisis
- Passage of TARP
- Section 133 required SEC to report back to
Congress - SEC held three Roundtables in 4th quarter 2008
- Third focused on
- assessing the role SFAS 157 played in causing
bank failures, - the impact of such standards on the quality of
financial information available to investors - the process used by the FASB in developing
accounting standards and the advisability and
feasibility of modifications to such standards.
21Fair Value Accountingand the Financial Crisis
- Observations
- Need a strong independent private-sector standard
setting - SFAS 157 did not cause banks to writedown their
assets writedowns were the result of bad
lending practices - Investors need to know what bank assets are worth
- SFAS 157 procyclicality role is highly
overstated. - Procyclicality results even if no fair value
problem. - FASB/SEC September 30 release notes provision for
level 3 - Accounting standard setting is for investors
regulation of banks to ensure stability of
financial system is for bank regulators and
supervisors.
22Fair Value Accountingand the Financial Crisis
- SEC Report and the future of fair value
accounting - TARP SEC report endorsed private-sector standard
setting and fair value accounting (including SFAS
157). - However, the report also recommends
- Developing implementation guidelines
- Readdressing accounting for financial asset
impairments - Ensuring that accounting standards meet the
informational needs of investors - Congressional hearings on mark-to-market in
March - 2009
23Fair Value Accountingand the Financial Crisis
- FASB responded by issuing 3 FSPs
- FSP FAS 157-4
- FSP gives companies applying 157 more flexibility
in determining when a market for a particular
asset is inactive or market prices can be
characterized as arising from distressed sales. - Application of the FSP is likely to lead to
companies using level 3 fair value estimates
rather than level 1 market prices. - Obvious concern with FSP leading to deterioration
of information about bank asset values provided
to investors
24Fair Value Accountingand the Financial Crisis
- FASB responded by issuing 3 FSPs
- FSP FAS 115-2 and FAS 124-2
- FSP requires separate display on the income
statement of losses related to credit
deterioration and losses related to other
market factors. - However, market-related losses would be recorded
in other comprehensive income (OCI)i.e., be
excluded from income by deducting them from the
total impairment chargeif it is not likely that
the holder of the security will have to sell it
prior to price recovery. - OCI portion likely to be excluded from Tier 1
capital - Banks have great discretion in determining what
hits income and what hits Tier 1 capital.
Difficult to verify (audit). - FASB has gotten into the bank regulation business.
25Fair Value Accountingand the Financial Crisis
- FASB responded by issuing 3 FSPs
- FSP 107-1 and APB 28-1
- This FSP requires additional disclosures about
fair value, including significant assumptions and
methods - Goal is to help investors determine quality of
fair value information. - Political Pressure continues
- HR 1909 introduced by US Rep. Steve Cohen on
April 6, 2009, would require the SEC to suspend
by rule, regulation or order the application of
mark-to-market accounting