174 M. Baker et al.
3.4. Financial policy
Direct empirical tests of how optimism and overconfidence affects financing patterns is
not extensive. Existing work addresses capital structure and financial contracting.
3.4.1. Capital structure
The basic optimism model predicts a pecking order financing policy, as pointed out by
Heaton (2002). Thus, much of the existing evidence of pecking-order policies, from
Donaldson (1961)toFama and French (2002), is at face value equally consistent with
pervasive managerial optimism. And the notion of pervasive managerial optimism does
not seem farfetched. InGraham’s (1999)survey, almost two-thirds of CFOs state their
stock is undervalued while only three percent state it is overvalued. Such responses are
all the more striking given the fact that the survey was taken shortlybeforethe Internet
crash.
To distinguish optimism from other explanations of pecking order behavior (for
example, adverse selection as inMyers and Majluf, 1984), a natural test would use
cross-sectional variation in measured optimism to see whether such behavior is more
prevalent in firms run by optimists. To our knowledge, exactly this test has yet to be con-
ducted, but certain results inMalmendier and Tate (2005, 2006)have a closely related
flavor. First, and as noted above, firms run by optimists (as identified by the Malmendier
and Tate options-based proxies for optimism) display a higher sensitivity of investment
to internal cash flow. Second, managers classified as optimistic show a differentially
higher propensity to make acquisitions when they are not dependent on external equity.
3.4.2. Financial contracting
Landier and Thesmar (2005)examine financial contracting between rational investors
and optimistic entrepreneurs.^21 They highlight two aspects of contracting with op-
timists. First, because optimists tend to inefficiently persist in their initial business
plan, the optimal contract transfers control when changes are necessary. (Kaplan and
Stromberg, 2003, find that contingent transfers of control are common features of ven-
ture capital contracts.) Second, because optimists believe good states to be more likely,
they are willing to trade some control and ownership rights in bad states for greater
claims in good ones; in this sense, the optimal contract “pays the entrepreneur with
dreams”. Ultimately, optimists may self-select into short-term debt, as it transfers pay-
ments and control to the investor in states that seem unlikely to occur, while realistic
entrepreneurs prefer less risky long-term debt.
Landier and Thesmar find some empirical evidence of this separation in a data set
of French entrepreneurs. Among other results, they find that the use of short-term debt
(^21) Manove and Padilla (1999)also consider how banks separate optimists and realists. They focus on the
overall efficiency of the credit market.