Docstoc

it it Cash

Document Sample
it it Cash Powered By Docstoc
					                 Session 2
Product Design and Research Design:
    Accrual Accounting and Cash
              Accounting

  CKGSB-PKU-Tsinghua Accounting Summer Camp

               Stephen Penman
     The Penman-Yehuda Paper


•   Issue A: Examining a basic normative
    feature of accounting design
•   Issue B: Specification in empirical
    research that emphasizes product
    design features
     Issue A: Cash Accounting vs. Accrual
       Accounting for Periodic Reporting
Normative statements (product features) implicit in
   accrual accounting:

1.    (Accrual) earnings add to shareholders’ equity
2.    Cash flow does not affect earnings or
      shareholders’ equity
3.    Cash flow decreases the value of the business.

Accounting value added is independent of the timing of
    cash
flows; cash flows are a dividend.
Issue B: (Mis)specification in Capital
          Markets Research
Reporting correlations vs. tests of specifications
                 Pit  a  bCGS it  eit ,

                            1.12
                        (t = 13.52)

    Pit  dit  Pit 1    CGS it   it

                              0.23
                           (t = 8.62)
 Specification in Capital Markets
            Research
Accounting Structure: Revenue - CGS = Gross Margin


                 Pt  a  b1Salesit  b2CGS it  eit
                                         -3.94
                                     (t = -17.74)


    P it  d it  Pit 1    1Salesit   2 CGS it   it
                                           -0.74
                                      (t = -9.48)
          The Research Question

What is the sign on b2 in the following regression?

                         Earn1       Cash Flowt
       Re turnt  a  b1         b2             et
                          Pt 1         Pt 1
                                 ?



How are earnings and cash flow priced?
Is cash priced positively?
    The Economics of Cash Flows:
          Valuation Theory
A Basic Principle of Finance (M&M):

•   The timing of cash flows does not matter

•   Cash flow streams can be repackaged with value preserved

Applied in practice:

•   Securitization repackages cash flows and keeps a book (with market-
    to-market accounting)

•   Investors trade on the book
Accrual Accounting Repackages
          Cash Flows
• Accounting reassigns cash flows to periods
• Accounting keeps a book on value
  - the balance sheet reports equity value
  - the income statement reports changes in value
• Accounting deems cash flows to be irrelevant
• Investors trade on the book
Accounting keeps the book imperfectly, and
  there’s the rub: How should an investor trade on
  the book?
        Cash Accounting for
           Shareholders
“Income Statement” (really a cash flow
  statement)

Net cash from operations – Cash to
 debtholders = Cash to shareholders

    [C - I] - F = d
        Product Feature: How Accrual
             Accounting Works
Cash                     Accrual Adjustments         Income
Flows                                               Statement
C-I            +   I + operating accruals      =        OI
  -F          +    D - financing accruals      =       NFE
  d                                                   Earnings
                             Balance Sheet
      Net Operating Assets            Net Financial Obligations and
                                                 Equity
 ΔNOA = I + operating accruals        ΔNFO = – D + financing accruals


                                        ΔBook value of equity = ?
               Accounting for Value
  ∆B = ∆NOA - ∆NFO

∆NOA = I + operating accruals

        = OI - (C - I)

∆NFO = -D + financing accruals
      = NFE - F
      = NFE - (C - I) + d
Thus,
∆B = OI - (C - I) - NFE + (C - I) - d
   = OI - NFE - d
   = Earnings - d

Free cash flow does not affect the book value of equity!
The Problem With Free Cash Flow

Home Depot, Inc. (millions of dollars)
                              1999     2000         2001        2002
Cash from Operations         1,894      2,439       2,977       5,942
Cash Investments             2,273      2,620       3,521       3,406
Free Cash Flow               (379)      (181)       (544)       2,536

General Electric Company (millions of dollars)

                               1999       2000         2001         2002
Cash from Operations          24,593     22,690       32,389       29,488
Cash Investments              42,179     37,699       40,308       61,949
Free Cash Flow               (17,586)    (15,009)     (7,919)     (32,461)
         Measurement Matters!

• The pricing of earnings depends on how
  earnings are measured
• Dividends and cash flows can have “information
  content”
• The “information content” of dividends and cash
  flows depends on how earnings are measured

  Build a specification that captures the structure of
accrual accounting but also incorporates measurement
                  Stock Returns and
                 Accounting Numbers
 Price differs from book value because of accounting
 measurement:
               Pit = Bit + (Pit - Bit),

              (Pit – Pit-1) = Bit + (Pit – Bit) – (Pit-1 – Bit-1)

 As ∆B = Earnings – d,
      Pit – Pit-1 = Earningsit – dit + (Pit – Bit) – (Pit-1 – Bit-1)

Dividing through by Pit –1
       Pit  Pit 1 Earningsit  d      B      P  Bit
                               it  it 1  it       1
          Pit 1       Pit 1   Pit 1 Pit 1   Pit 1
      P  P 1            Earningsit      d      B
       it     it
                  a  b1             b2 it  b3 it 1   it
          P 1
           it                P 1
                              it         P 1
                                          it     P 1
                                                  it
    Some Insights on Measurement
•   If earnings are measured such that
                 ∆P = Earnings – d
         then ∆premium = 0

•   Earnings measurement creates other information
       If dividends or cash flow have “information content” they must
           explain changes in premiums (ε)

•   The change in premium is “all other factors”
       other information
       expected returns
       changes in expected returns
       market inefficiency

•   b1 different from 1.0 if earnings are correlated with change in premiums (ε)
      Some Further Observations
    Bt 1
•         has an initializing role
    P 1
     t

            Initializes for forecasted returns – risk or abnormal returns
            Initializes for measurement of book value
                                 E
            Correlated with P if conservative accounting
                                t

                               t 1




• A change in premium amounts to a change in expected
  growth in residual earnings. The multiplier, b1captures
  this growth.
                      The Pricing of Earnings
                                  US Stocks, 1963 - 2001


       Pit  Pit 1          Earningsit      d      B
                     a  b1             b2 it  b3 it 1   it
          Pit 1                Pit 1      Pit 1  Pit 1



                              a            b1               b2              b3          Adj. R2
       Coefficient         0.05           1.67           -2.98            0.08            0.13
        t-statistic       (1.46)         (8.35)         (-5.64)          (5.10)




Source: Penman, S. and N. Yehuda, “The Pricing of Earnings and Cash Flows and an Affirmation of Accrual
Accounting,” Columbia University, 2004
       The Pricing of Earnings and Free
                 Cash Flows
Adding free cash flows to the regression

Pit  Pit 1          Earningsit      d      B           (C  I )it
              a  b1             b2 it  b3 it 1  b4              it
   Pit 1               Pit 1       Pit 1  Pit 1        Pit 1



                        a                b1               b2                  b3          b4      Adj. R2
Coefficient           0.04             1.69             -2.88                0.08       -0.03         0.14
 t-statistic         (1.37)           (8.38)           (-5.62)          (4.96)          (-1.12)

                                                                                    Free cash flows
         Returns to Operations and
             Operating Income
            ations PNOA  NOAt  (Pt NOA  NOAt )
Price of Oper       t



            P  Pt 1
            t
              NOA  NOA
                                        
                            ΔNOAt  Δ Pt NOA  NOAt   

                                           t   
                             OI t  (C - I)  Δ Pt NOA  NOAt   

Note : Pt NOA  Pt  NFO
        Does Free Cash Flow Add To The Price
                   of Operations?
               PitNOA  PitNOA         OI it      (C  I ) it      NOAit
                           1
                                  1 NOA   2               3 NOA1   it
                    PitNOA
                       1              Pit 1       PitNOA
                                                       1           Pit 1



                               1               2               3           Adj. R2


Coefficient    0.01            2.21            -1.10            -0.01              0.22
 t-statistic   (0.37)         (12.63)         (-24.77)         (-0.62)



                          Free cash flow reduces the price
                          of operations approximately dollar for dollar
The Pricing of “Cash From Operations” and
            “Cash Investment”

PitNOA  PitNOA         OI it      Cit       I it 1   NOAit 1
        NOA
            1
                   1 NOA   2 NOA   3 NOA   4     NOA
                                                                  it
     Pit 1             Pit 1    Pit 1    Pit 1      Pit 1


                   1.68         -0.98         1.30
                (t = 12.42)   (t = -15.53)   (t = 12.88)




                              $ for $        Added Value
Where Might You Go from Here?
Can a structured financial statement
 analysis that explains ε?
Can earnings quality metrics explain ε; does
 cash flow have a role?
What do εit look like over time?
Do εit predict future returns?
             Session 2

  Product Design and Research
Design: Product Failure of a Popular
        Asset Pricing Model
CKGSB-PKU-Tsinghua Accounting Summer
               Camp

           Stephen Penman
    A Product to be Examined
• Book-to-price (B/P) robustly predicts stock
  returns: the B/P effect
• Fama and French build an asset pricing
  model on the basis of this observation
• Caveat Emptor
 The Penman, Richardson and Tuna
              paper

What Explains the Book-to-Price Effect?
• Is it a loading on a risk factor?
• Is it mispricing?
Standard retort: Issue can’t be resolved
without commitment to an agreed-upon
asset pricing model.

Can an accountant help out?
Two Points on Which We Can Agree

1. Given market efficiency, B/P is an accounting
   phenomenon
   - The hedge fund vs. money market fund

2. Given operating (firm) risk, financial leverage
   adds to expected returns: any result to the
   contrary would be anomalous. We can’t agree
   on an asset pricing model, but a valid model
   must have this feature
An Accountant Can Help: Recognize
     a Feature of the Product
Difference between Book and Price:

    PP     NOA
                  P   ND              P  B  P NOA  NOA  ( P ND  ND)


                              P  B  P NOA  NOA
Book-to-Price Ratio:

      B NOA ND                              P NOA NOA ND
                                               NOA 
      P   P   P                               P   P      P

                            B  NOA ND  NOA      
                               NOA     NOA  1
                            P  P      P P       
How B/P “Subsumes” Leverage:
 Decomposing Book-to-Price
    B   NOA     ND  NOA     
                        1
    P   P NOA
                 P P NOA
                             

• B/P is a weighted average of the enterprise
  book-to-price ratio (NOA/PNOA) and the book-to-
  price ratio for financing activities
• Leverage introduces a non-linear relationship
  between Book-to-Price and NOA/PNOA
          Decomposing Stock Return
                          Use the Relation, FCF = d + F

                                       
       E Pt 1  d t 1  Pt   E ( Pt 1  FCFt 1  Pt NOA )  ( NDt 1  Ft 1  NDt
                                         NOA
                                                                                              
If     P NOA   NDt
        t
                  1
         Pt     Pt




       Pt 1  d t 1  Pt  Pt NOA     Pt 1  FCFt 1  Pt NOA  NDt
                                             NOA
                                                                             NDt 1  Ft 1  NDt 
     E                              E             NOA                E                      
               Pt             Pt                Pt               Pt              NDt          


              E Rt 1   E R         NOA
                                       t 1      
                                                    NDt
                                                     Pt
                                                          
                                                        E RtNOA  RtND
                                                            1      1      
    Pairing the Relations

E Rt 1   E R     NOA
                     t 1      
                                  NDt
                                   Pt
                                        
                                      E RtNOA  RtND
                                          1      1   

    B   NOA     ND  NOA    
                  NOA  1
    P   P NOA
                 P P       

                 NOAt        NDt
Rt 1    1      NOA
                         2      t
                 Pt           Pt
         Regression Results

                           NOAt        NDt
          Rt 1    1      NOA
                                   2      t
                           Pt           Pt


                    0.116        -0.022
                    (6.04)       (-2.62)

For NOAt < 1
     NOA
                    0.159        -0.045
    Pt
                    (4.27)      (-4.97)
 Returns for B/P and Leverage Portfolios
                                      ND/P Quintile                      (HIGH-


                    LOW        2          3             4      HIGH       LOW)

                                                                        -0.148
             LOW    0.017    -0.053    -0.069         -0.099   -0.131
                                                                        (-2.78)
                                                                        -0.133
              2     0.053    -0.017    -0.049         -0.061   -0.079
                                                                        (-2.74)
                                                                        -0.086
              3     0.034    -0.006    -0.048         -0.052   -0.052
                                                                        (-2.89)
                                                                        -0.077
              4     0.032    0.017     -0.034         -0.050   -0.046
                                                                        (-3.52)
                                                                        -0.079
              5     0.055    0.030      0.000         -0.051   -0.024
NOA/PNOA                                                                (-2.34)
    Decile                                                              -0.060
              6     0.026    0.016     -0.015         -0.017   -0.034
                                                                        (-3.19)
                                                                        -0.062
              7     0.039    0.016      0.008         -0.007   -0.023
                                                                        (-2.57)
                                                                        -0.056
              8     0.043    0.037      0.018         -0.005   -0.013
                                                                        (-2.05)
                                                                        -0.055
              9     0.053    0.057      0.029         0.007    -0.001
                                                                        (-2.03)
                                                                        -0.061
             HIGH   0.101    0.056      0.066         0.039    0.040
                                                                        (-1.61)

                    0.084    0.109     0.135          0.138    0.170
(HIGH-LOW)
                    (1.65)   (3.01)    (4.48)         (3.17)   (4.20)
           Did You Know?
Empirical research has never been able to
 show that stock returns are positively
 related to leverage. A fundamental,
 uncontroversial tenet of finance does not
 have empirical validation.

Can an accountant help?
  Leverage and Fundamental Rates of
               Return

                                          NDt
          ROCE t 1    a  b1 RNOAt  b2      et
                                           Bt


• For firms with favorable leverage (i.e., RNOA >
  NBC):
   – b1=0.484 (t=13.85) and b2=0.023 (t=8.36)
• For firms with un-favorable leverage (i.e.,
  RNOA < NBC):
   – b1=0.545 (t=21.09) and b2=-0.067 (t=-9.05)
  Add Leverage to Penman-Yehuda
      Unlevered Regressions
     PitNOA  PitNOA         OI it     FCFit     NOAit        NDt 1
             NOA
                 1
                        1 NOA   2 NOA   3 NOA1   4          it
          Pit 1             Pit 1    Pit 1     Pit 1       Pt 1



                         1         2        3         4         Adj. R2
1963-2005 0.02 2.16 -1.15                     0.01      -0.01         0.22
t-statistics 0.71 13.58 -22.16                0.37      -1.70
           Levered Regressions

      Pit  Pit 1         OI it     FCFit     NOAit 1   NDt 1
                      1 NOA   2 NOA   3 NOA   4          it
         Pit 1            Pit 1    Pit 1     Pit 1     Pt 1



                            1        2         3        4       Adj. R2
1963-2005 -0.01 2.60                -0.41 -0.03          0.09         0.13
t-statistics -0.41 16.54             -5.64 -1.18         9.56
          A Final Question
Can we think of asset pricing in terms of
 operating earnings at risk plus a premium
 for leverage?

Earnings move stock prices…but it depends
 on the product features of the accounting.

				
DOCUMENT INFO
Shared By:
Categories:
Tags:
Stats:
views:3
posted:9/27/2012
language:Unknown
pages:37