Everything You Ever Needed To Know About Stock Compensation & More Conference of Consulting Actuaries Meeting

1. Everything You Ever Needed To Know About Stock Compensation & MoreConference of Consulting Actuaries Meeting Session 46 - October 24, 2007

2. Bill Murphy North Central Sub-Area Location Leader, Performance & Reward Senior Manager, Ernst & Young 216-583-2869 william.murphy05@ey.com Your Presenter

3. Agenda • Current trends in Stock-Based Compensation • Taxation of Equity Awards

4. Primary Stock Based Incentive Vehicles • Non-Qualified Stock Options (NQSO) • Right to purchase stock at a specified price for a specified period of time. • Value is based on the appreciation in the underlying stock from the date of grant to the date of exercise. • Does not meet specific tax requirements of Internal Revenue Code (IRC) Sections 421-4222. • Restricted Stock/Restricted Stock Units (RS/RSU) • Shares (or share units) granted to employee, typically subject to vesting restrictions. • Shares retain some value even if stock price declines. • Difference: Restricted stock is transferred to the participant on the date of grant. Restricted Stock Units are promises to deliver the stock on the vesting date. • Stock Appreciation Rights (SAR) • Right to receive stock (or cash payment) equal to the appreciation of company shares between the date of grant and the date of “exercise”. • Similar to stock options, but SARs do not require the payment of an exercise price, and may be structured to prevent actual share ownership (which may give rise to financial statement implications discussed in more detail later).

5. Plan Design Issues and Considerations • Appreciation v. Full Share Awards – Should the company limit the value to appreciation from the date of grant? • Time-based or Performance-Based Vesting – Should vesting be based on continued service or achievement of some company performance conditions? • Fixed Grant or Variable Grant – Should the number of awards be fixed up front (e.g., time-base vested awards), or should performance dictate the ultimate size of the award? • Executive Stock Ownership – Does the company have share ownership guidelines for its executives and what awards count towards those goals?

6. Plan Design Issues and Considerations • Shares Available for Grant (Number and Full Share v. Appreciation Awards) – Companies need to consider shares remaining available under the stock plans and stock plan limitations when planning grants, and institutional shareholder considerations on shares reserved, share value transfer, and burn rates. • Tax Impact – Companies need to consider tax laws not just in the U.S. but in each international jurisdiction to determine most efficient stock vehicle from a tax perspective.

7. Plan Design Issues and Considerations • Accounting Implications – Companies need to consider the potential financial statement implications of the awards. • Institutional Shareholder/Disclosure Implications – What are key shareholders focusing on in equity plan design? • Administrative Considerations – Sometimes overlooked. How difficult will it be to administer the plan? Not just from an HR perspective, but in getting appropriate information for accounting, tax, etc.

8. Current Trends in Stock Compensation PORTFOLIO APPROACH • Companies continue to shift from “stock options only” culture towards a “portfolio of stock vehicles” for annual stock based grants. • Companies are increasingly relying on restricted stock/restricted stock units to deliver their stock-based compensation opportunity. PERFORMANCE CRITERIA • With shareholders increasingly focused on performance, more companies are beginning to utilize performance-based vesting/performance-based awards. • Accounting rules (along with shareholder emphasis on performance) continue to fuel this trend in that performance-contingent awards are no longer subject to adverse accounting treatment as compared with service-based awards.

9. Performance-Based Compensation • There has been an increased shareholder focus on compensation that is more closely linked to company performance. • Performance goals can relate to achievement of absolute or relative performance metrics. • Absolute • Examples: Revenue growth, Operating Income growth, Earnings Per Share, Return on Investment. • Goals are focused on the performance of the company solely. • Relative • Examples: Total Shareholder Return (TSR) and stock price appreciation relative to a group of companies or industries, or revenue growth against revenue growth of a peer group. • Goals are focused on the performance of the company compared to other companies. • Whether a condition is classified as “performance condition” or “market condition” has significant consequences to the award valuation at the date of grant as well as the manner in which the compensation cost is recognized.

10. Performance Conditions • A performance conditionmay be a condition that is based on the operations or activities of the employer that are not driven by stock price/TSR, etc. The condition may relate to the performance of the entire company, a division, or an individual employee. • A performance condition may also be in reference to the same performance measure of another entity or group of entities (e.g., growth of EPS relative to other entities in the employer’s industry) • Example: The Board of XYZ Corporation (NYSE: XYZ) recently approved the award of performance vesting restricted stock units. The award provides participants with the opportunity to receive shares if the Company's compounded annual growth rate on revenue during a three-year performance period is equal to or exceeds the Company's peer group median. If the performance goal is met, the restricted stock units cliff vest three years from the grant date. Failure to meet the performance goal results in forfeiture of shares.

11. Market Conditions • The exercisability or other terms of a stock award may be dependent on achieving a specified stock price or a specified return on the stock price (e.g., price appreciation plus dividends). FAS 123(R) refers to such conditions as market conditions. • Examples of market conditions would include those in which exercisability is dependent on or other terms are affected by: • The employer’s stock price achieving a specified level. • Achieving a specified return on the employer’s stock, the calculation of which is based on both stock price appreciation and dividends on the stock. • The employer’s stock price increasing by a greater percentage than the average increase of the stock price of a group of peer companies. • A specified stock return, which is based on both stock-price appreciation and dividends on the stock that exceeds the average return on the S&P 500.

12. Key Differences Between Performance & Market Conditions. • For awards with a performance condition, compensation expense is reversed for awards that do not vest (similar to awards with service conditions). In contrast, compensation expense is not reversed for awards with a market condition, provided the requisite service has been rendered. • For awards with a performance condition, the valuation is similar to awards with service conditions (i.e., the performance condition is not incorporated into grant date fair value calculation). However, market conditions need to be factored in the valuation of awards through the use of a lattice model or Monte Carlo simulations.

13. James LecherAssistant Vice PresidentFAS 123 Valuations AABS Global Learning

14. Development of Valuation Assumptions • Regardless of the valuation model chosen correct assumption development is extremely important to develop accurate fair values. “Garbage in Garbage out” is very applicable in option valuation. • FAS 123(R) states the valuation models must at a minimum take into account the following assumptions. * The expected life is an input to the Black-Scholes Model. In a lattice model probabilities of exercise are input into the valuation model and the expected life is a biased estimate that is an output of the model. AABS Global Learning

15. Expected Life Considerations • Factors to consider include: • Simplified Approach from SAB 107 • Vesting period and contractual term • Historical patterns of exercise and post-vesting termination • Changes in Option Characteristics (i.e. vesting or term) • Expected Volatility • Optionee Demographics • Age/Gender/Country/Job Level • Stratification • Peer Companies • Censored Data and assumptions about outstanding options • Minimum at Vesting • Maximum at Term • Most likely may be the midpoint of future remaining term AABS Global Learning

16. Expected Life – Emerging Best Practice • Courtesy of FAS123R Data in SEC Filings as Provided by Salary.com AABS Global Learning

17. Expected Life – Emerging Best Practice • Courtesy of FAS123R Data in SEC Filings as Provided by Salary.com AABS Global Learning

18. Expected Life – Emerging Best Practice • The most common approach (58%) is to assume currently outstanding options will be exercised at the midpoint of the future remaining term. • The “Simplified Approach” as outlined in SAB #107 is also popular. It can be calculated from historical data or, in the absence of historical data, can be calculated using the vesting schedule and the full contractual term. AABS Global Learning

19. Expected Volatility - Considerations • Factors to consider include: • Historical volatility • Period of Measurement (short term vs. mid-term vs. theory of mean reversion) • Removal of data? • Implied volatility • Liquidity – Open Contract vs. Trading Volume? • Historical Implied Volatilities? • Peer companies - especially for companies with little history • Blended Volatilities – Developing supportable weights AABS Global Learning

20. Expected Volatility – Emerging Best Practice Courtesy of FAS123R Data in SEC Filings as Provided by Salary.com AABS Global Learning

21. Employee Stock Options Valuation Models • Common Stock Option Pricing Models - Black-Scholes Model -Binomial Models -Simulation Models Each Model has pros and cons with respect to accuracy, ease of application, and modeling flexibility. Regardless of the model used FAS 123(R) specifies that all of the following assumptions must be used to calculate the grant date fair value of an equity award. AABS Global Learning

22. Black-Scholes Formula - Pros and Cons The Black-Scholes model offers the following advantages / disadvantages for FAS 123(R) purposes. AABS Global Learning

23. Black-Scholes Formula - Pros and Cons Historical data from periods that had unusual stock price movements can cause anomalies in the Black-Scholes model. This is due to treating time after grant as the only driver of exercise. AABS Global Learning

24. Black-Scholes Formula – An Alternative Approach Option pricing theory states that the fair value of an option increases at a decreasing rate as the term of the option increases. This causes the Black-Scholes Model to overstate the fair value by approximately 3% to 7% since all options are assumed to be exercised at one point in time. AABS Global Learning

25. Alternative: Multiple-Point Black-Scholes Improves Accuracy • Fundamental inaccuracy of traditional Black-Scholes is use of single-point expected life • Collapses all expected exercise activity to one date • Paragraph A30 – “estimating the fair value of an option based on a single expected term that effectively averages the differing exercise and post-vesting employment termination behaviors of identifiable groups of employees will potentially misstate the value of the entire award” • A multiple-point approach overcomes this problem • Utilizes actual historical exercise option data by individual • Can incorporate historical unexercised option data • A distribution of exercise activity is created • Each option is valued as a European option at its actual or assumed exercise date • Creates weighting of results based on number of shares • Results are summed and averaged • Fair value will always be lower than traditional (i.e. single-point) Black-Scholes, as marginal value of option gets smaller with time AABS Global Learning

26. Aon Multiple- Fixed Point Error Expected Contractual Dividend Standard Point Black- Company Black-Scholes Term Life Term Volatility Yield Deviation Kurtosis Scholes Company A 13.28% 12.26% -7.65% 4.29 10.00 16.11% 3.36% 2.3590 (0.3778) Company B 36.33% 35.14% -3.28% 5.84 10.00 35.04% 1.00% 2.3883 (0.9050) Company C 44.89% 43.29% -3.57% 4.89 10.00 46.46% 0.00% 2.2066 (1.0080) Company D 20.30% 19.19% -5.50% 4.77 10.00 25.11% 3.23% 2.4140 (1.0919) Company E 55.82% 54.28% -2.76% 5.05 10.00 61.06% 0.00% 1.8318 0.0940 Company F 31.02% 30.30% -2.33% 6.19 10.00 22.98% 0.59% 1.8428 (0.5975) Company G 42.39% 40.53% -4.38% 4.07 10.00 47.19% 0.00% 2.1376 (0.6557) Company H 59.50% 56.28% -5.40% 5.26 10.00 64.71% 0.00% 2.4881 2.0704 Company I 55.99% 53.77% -3.95% 4.74 10.00 63.45% 0.00% 2.0822 2.4096 Company J 37.19% 36.80% -1.04% 3.61 5.00 43.00% 0.00% 1.0037 7.2713 Company K 49.44% 48.02% -2.87% 3.91 7.00 60.78% 0.00% 1.5042 3.0813 Company L 62.34% 59.27% -4.92% 5.11 10.00 70.77% 0.00% 2.1348 2.7479 Company M 56.73% 54.31% -4.26% 5.81 10.00 56.44% 0.00% 1.9833 2.3470 Company N 48.78% 44.90% -7.96% 3.81 10.00 59.95% 0.00% 2.4418 1.5674 Company O 29.83% 29.38% -1.50% 4.93 6.00 29.38% 1.20% 0.8175 5.7420 Company P 54.23% 52.06% -4.00% 5.30 10.00 57.30% 0.00% 2.2993 2.2516 Company Q 43.17% 42.31% -1.99% 5.69 10.00 38.55% 0.00% 1.9894 2.6006 Company R 67.55% 64.43% -4.61% 5.17 10.00 79.74% 0.00% 2.1908 2.6212 Company S Examples: Multiple-Point Black-Scholes 17.02% 16.37% -3.87% 4.23 10.00 21.34% 3.20% 1.8860 2.5333 Company T 28.65% 27.88% -2.67% 5.31 10.00 28.56% 1.65% 1.9172 2.3716 Average 42.72% 41.04% -3.93% 4.90 9.40 46.40% 0.71% 1.9959 1.7537 AABS Global Learning

27. Results: Multiple-Point Black-Scholes • Of the 20 companies, the fixed point Black-Scholes had an approximate 4% error term • Aon has observed that a fixed point Black-Scholes model creates the greatest error for companies that have the following characteristics: • High Dividend Yields • Short Historical Average Lives • Large Standard Deviations of Exercise Behavior • Small or Negative Kurtoses AABS Global Learning

28. Binomial Models Two Types of Binomial Models Barrier Model – - Standard Binomial Model - Cox, Ross, Rubinstein Binomial Model (CRR Model) - Hull/White Binomial Model Exercise Probability Model -Regression Based Models - Hazard Rate Based Models AABS Global Learning

29. Binomial Models • Barrier Models - The barrier models differ in how they assume that the option will be exercised. When the “Barrier is Broken” the option is exercised. • Time Based Barrier - Standard Binomial and CRR Binomial Model -These models are identical when there is no dividend. The fair values from these models will converge to the Black-Scholes model if small enough time increments are used. -The CRR model is slightly more refined on a dividend paying stock. It allows for early exercise in instances where it would beneficial financially to exercise the option early. • Stock Price Based Barrier - Hull/White Model - The Hull-White Model uses stock price as a barrier. - Several studies have shown that time after grant is a more significant than stock price as a driver of exercise. - This model is not considered to be as accurate as a time based barrier model by many FAS 123(R) practitioners. AABS Global Learning

30. Binomial Models Exercise Probability Model -Still in early stages of development -Can use many drivers of exercise - Time after grant, proximity to vesting, stock price appreciation, observed historical volatility are seen at this time to be the most critical drivers of exercise behavior. - Regression models can be used but it may be difficult to capture the stock price path dependent nature of employee exercises. - Many believe that hazard rate modeling will be necessary for accurate modeling. - Currently the use of hazard rate modeling is being examined by a specialized S.O.A. task force. The construction of standardized tables of exercise rates is being studied. -This would be similar in nature as the use of standardized rates of retirement, termination, disability, and mortality are used for pension plan accounting under FAS 87. AABS Global Learning

31. 3.125% 15.625% 31.25% 31.25% 15.625% 3.125% Black-Scholes and the Traditional Binomial Model * Simplification such that there is an equal probability of downward and upward movements. This is not the case generally as the probability of upward and downward movements are governed by the volatility, the dividend yield, and the discount rate. Illustration comparing closed-form Black-Scholes model with a traditional binomial model (present value of future cash flows) Traditional Binomial Model Black-Scholes Probability * @ AABS Global Learning

32. Binomial Models Binomial option pricing models offers the following advantages/disadvantages for FAS 123(R) purposes: AABS Global Learning

33. Simulation Models • Use to value exotic options that can not be valued with Black-Scholes or Binomial models. • Extremely flexible - Can be used to model extremely complex instruments. • Can capture correlation of stocks which is important when modeling relative performance plans. AABS Global Learning

34. Simulating Stock Price • The first step to any simulation valuation model is to simulate the stock price over the term of the option. • The most common technique is to use Geometric Brownian Motion. The underlying distribution of stock prices is log-normal with a drift coefficient. • Alternatively one can model the returns via a normal distribution which yields the identical price distribution as above. AABS Global Learning

35. Simulating the FAS 123(R) fair value - A Real-World Example – An absolute Performance Plan • Company ABC grants restricted shares that are subject to the following vesting schedule. TSR is defined as the total shareholder return assuming that all dividends are reinvested in the company stock. • The following economic assumptions will be used to calculate the grant date fair value: AABS Global Learning

36. Simulating the FAS 123(R) fair value - A Real-World Example - Continued • The following spreadsheet has been constructed to model this instrument: AABS Global Learning

37. Simulating the FAS 123(R) fair value - A Real-World Example - Continued • The prior sheets shows one iteration of this simulation. After re-simulating 250,000 times the model yields the following results: • Please note that the simulated fair value is NOT calculated by simply multiplying the grant price by the vesting percentage. Doing so would substantially understate the grant date fair value by ignoring the correlation between vesting events and stock prices. AABS Global Learning

38. Simulating Models – Pros and Cons Simulation models offer the following advantages / disadvantages for FAS 123(R) purposes: AABS Global Learning

39. Current Modeling Trends AABS Global Learning

40. FAS123R Valuation: Model Use Comparison FAS 123 and FAS 123(R) • FAS123 - Valuation guidance requires use of the Black-Scholes or a binomial model - As of 2002, 5 companies of the S&P 500 used a binomial model - Black-Scholes was generally perceived to overstate the value - Companies ignored the error term because it didn't affect earnings • FAS123(R) - Allows both Black-Scholes or binomial models, however states that the design of a lattice model more fully reflects the characteristics of an ESO- - As of January 10, 2007, 402 companies have publicly disclosed the use of a binomial model (see Appendix A) - 80 of those 402 companies that now use a binomial model are components of the S&P 500. AABS Global Learning

41. FAS123R Valuation: Binomial Adoption Year AABS Global Learning As of 9/10/2007

42. Steve Zwicker,Senior Consultant FAS 123R Accounting Rules Overview

43. Introduction / Agenda This presentation provides an overview of the accounting rules under Statement of Financial Accounting Standards No. 123R – Share-Based Payment. Topics are presented in the following main categories: • Measurement Issues • Attribution Issues • Tax Issues • Other Issues

44. Measurement Issues Scope of Standard

45. Measurement Issues Scope of Standard

46. Measurement Issues Equity versus Liability Awards

47. Measurement Issues Equity versus Liability Awards

48. Measurement Issues Equity versus Liability Awards

49. Measurement Issues Grant Date

50. Measurement Issues Reload Features