University of Montana - September 15, 2006
Repurchases, Employee Stock Option Grants, and
Hedging
Daniel A. Rogers
Portland State University
Elevator pitch
What’s the rationale for observed relation between employee options and stock repurchases? Partial explanation: repurchases serve as hedge against uncertainty
surrounding option obligations. Provides more “economic” justification than EPS management
hypothesis. Findings:
Employee stock option grants exhibit positive relation with repurchases.
For firms in which this relation is strongest, I find evidence consistent with an optimal hedging motive (although there might be more to the story.
Background on options and repurchases
Microsoft: “We repurchase our common shares primarily to manage the dilutive effects of our stock option and stock purchase plans, and other issuances of common shares.” - 2003 10-K Footnote 14 to financial statements.
Existing empirical evidence Substitution of repurchases vs. dividends (exec story): Fenn
and Liang (2001 JFE) and Kahle (2002 JFE). Option funding: Kahle (2002 JFE). Earnings management (anti-dilution): Bens et al. (2003 JAE)
and Weisbenner (2000 wp).
Another Story?
Granting options to employees creates uncertain future liability for shareholders. Current shareholders incur an opportunity cost when employee
stock options are exercised. Amount of the opportunity cost = (stock price at exercise date –
exercise price). Do shareholders hedge this uncertainty? Hedging strategy: repurchase shares when options are
granted. Similar to a forward purchase of currency or commodity. Strategy implies positive relation between repurchases and option
grants over time.
Example
Assume: Stock price = $20; Option granted at the money;
Cost of equity = 8%; and Dividend yield = 3% What is the ultimate cost borne by existing
shareholders? Exercise price at time of exercise – grant date price Suppose exercise occurs in 5 years: FV of $20
invested in 5 years = $25.68 ($20 * 1.055) Ex post economic cost = $5.68 if repurchase (known
at time of grant and repurchase) Ex post economic cost = ??? if no repurchase
How does situation differ from other hedging
problems? Typical hedging situation: “bad” outcomes = low
cash flows or earnings. In this case: “bad” outcome is high stock price at
time of exercise. “Normal” opportunity cost: stock price increases by
dividend-adjusted cost of equity. If stock price change between grant and exercise dates
exceeds dividend-adjusted cost of equity, repurchasing stock at grant date provides a positive payoff against excess opportunity cost.
Why might a high stock price be bad?
Jensen (2005) arguments Management “games” market expectations stock
price > intrinsic value Wealth-destroying acquisitions (Moeller et al., 2005 JF)
Employees choose when to exercise: If employees exercise options when price is above intrinsic value rent extraction.
If company repurchases stock at high prices, its alternatives to fund growth opportunities are 1) less investment, 2) tap external capital markets
What are the “traditional” incentives to hedge?
Reduction of underinvestment/distress costs Froot et al. (1993), Tufano (1998), Smith and Stulz (1985), among
others.
Tax function convexity Smith and Stulz (1985)
Increase borrowing capacity and interest tax shields Leland (1998)
Managerial motives Smith and Stulz (1985), Stulz (1984), Tufano (1996), among
others.
Does the option hedging story fit into any of these
categories? Reduction of distress and tax convexity?
Clearly, NO!
Increasing debt tax shields? Maybe Mozes and Raymar (2001 wp): issue options, issue debt
and repurchase stock.
Managerial motives? Hard to disentangle hedging motive from “underpriced
stock” story.
What about the underinvestment theory?
If assume firm monetizes opportunity cost by repurchasing shares around option exercise, then: Higher stock price less cash available for
investment at exercise date. If deadweight costs associated with new (debt)
financing, then firm might underinvest. Repurchasing stock at grant date is effective if
investment opportunities are correlated with stock price (this idea seems reasonable).
Plan of attack
First, establish if a link exists between option grants and stock repurchases. Regress stock repurchases on option grants and
other explanatory variables. If a link exists, then can optimal hedging
story explain hedging behavior? Construct a measure of “hedging” and regress
optimal hedging proxy variables against it.
Sample
151 randomly selected S&P 500 firms. Manual data collection of employee
option data. Time frame: 1993 – 2003 (or
maximum 10-K filings available from EDGAR).
Research design - Option grants & repurchases
Dependent variable = number of shares repurchased Independent variables:
log of market capitalization free cash flow market-to-book of assets capital expenditures long-term debt dividend yield stock price change stock price volatility option grants this is the variable of interest! exercised options
Data summary
Mean Median Std DevRepurchases 1.83% 0.69% 2.83%
Option grants 2.23% 1.56% 2.91%
Rep / grants 2.42 0.36 18.87
Exercises 1.08% 0.70% 1.37%
Total options 8.10% 6.93% 6.84%
Vested options 4.06% 3.39% 3.34%
Exec options 1.65% 1.24% 1.79%
Vested exec options 0.90% 0.60% 1.08%
Methodological issues
Dependent variable is censored at zero (approximately 1/3 of observations) Suggests Tobit methodology
Panel data Random/fixed effects Fixed effects model in Tobit would yield biased
estimates. Tobit random effects estimation
Key results – Base repurchases model (Table 2)
Repurchases and options: Positive relation with option grants and exercises. Robust to inclusion of other option variables (Panel B). Other option variables do not exhibit similar statistical
relations with repurchases (Panel B). Repurchases and other variables:
Positive relations with: Firm size, and free cash flow.
Negative relations with: Market-to-book, leverage, current year stock price change and
volatility.
Robustness results – Table 3
“Excess” grants are positively related to repurchases (two-stage model).
Controlling for lagged option grants back 4 years. Concurrent grants remain positively related to repurchases. Evidence of stronger relations for 1 and 2-year lagged grants.
Controlling for Bens et al. (2003) earnings management variables. Option grants remain positively related to repurchases.
How Is Hedging Defined?
Cannot directly observe “hedging” from disclosures.
Derived measure: Coefficient of variation for repurchases-to-
grants ratio = inverse hedging measure. Effectively captures a population driving
positive relation between grants and repurchases in multivariate.
Coefficient of variation and mean: Repurchases-to-grants
ratioFull sample = 140 firms (at least 1 yr of
repurchases)
Coefficient of variation Mean
Average 1.58 2.64
Median 1.40 1.13
Standard deviation 0.73 6.17
25th percentile 1.02 0.50
75th percentile 2.05 2.38
Minimum 0.24 0.01
Maximum 3.32 57.67
What variables explain option grant hedging?
Table 5 – differences between models Model 1: no industry controls Model 2: industry controls, but no control for variation of repurchases-to-
exercises Model 3: includes variation of repurchases-to-exercises, but no industry
controls Model 4: Controls for variation of repurchases-to-exercises and industry
effects Model 4 results - Effect on “hedging”:
Leverage: Negative relation R&D expenditures: Positive relation Vested exec options: Positive relation Firm size: Negative relation Market-to-book: Negative relation Exec shares held: Positive relation
What about the “non-hedgers?”
Table 6 results Difference between non-repurchasing
firms vs. those that repurchase (logit) Lower free cash flow Greater leverage More R&D
“Continuous” measure of option grant hedging Very similar results relative to Table 5
Summary of Most Notable Results
Option grants are positively related to repurchase activity. Even if not intentional, this pattern provides a hedge for
shareholders against uncertain opportunity cost of options.
Firms that exhibit less variation in repurchase activity relative to option grants also exhibit greater R&D spending and less leverage. This set of results fits underinvestment costs rationale
for hedging.