wilmott.com

I work for a Fortune 500 company shortly coming under the scope of FAS123(R), which requires the expensing of employee stock options to the income statement. The government accountants have won, and corporations are going to be required to expense their employee stock options (ESOs). Now there is a motivated corporate base that thinks stock options should not be given an unrealistic value on the income sheet. The question on everybody’s mind is “how should we value ESOs in a fair way to make everybody happy?”On the side of accountants and pseudo-analysts are those that think that any method of stock option valuation is better than none. They misunderstand that companies will respond to the law and employees will suffer due to their hastiness. These people believe that the income statement should reflect a cost, regardless of its accuracy, and believe that the Black Scholes valuation is a fair approximation to reality. The other camp believes that Black Scholes is not an accurate valuation engine for employee stock options due to the uniqueness of ESOs, such as the vesting schedules, long horizon, illiquidity (nontransferability), employee forfeiture etc. For a quick background, check out this article: http://accounting.smartpros.com/x49842.xml. Cisco created ESORs in attempt to create a market traded instrument, but they have fallen rather flat. In my view, there needs to be a valuation metric to make the SEC and FASB happy.As a reader of Wilmott since the beginning, I had a thought. Maybe there are some readers out there who could figure out a method to value ESOs more accurately than traditional Black Scholes. The rules for valuation are in the FAS document and have been clarified by both the FASB and the SEC.So to make a short question long, I am wondering if there would be a warm reception from the Wilmott readership for a competition to value employee stock options. I was thinking generous cash prize and publication for the best paper. Any comments?

Black Scholes is in general a very poor model for ESO's as it fails to take into account the unique aspects of an ESO (most of which reduce the option value and thus the option expense). We have developed a trinomial lattice implementation of the Hull & White 2004 model which reduces option expense versus Black-Scholes and is fully compliant with both FAS 123 and its European conterpart IFRS 2. The model can cope with both single and multi-vesting ESO's.The model is free to use online at http://www.esomanager.com/pricing_new.aspx and features full online help to guide the user through the valuation process. Detailed documentation on the model can be found on the Resources page.

There is no single method of valuing ESO's - I have about 15 different models that I use, depending on the terms and conditions, and the nature of the employees. As Judo mentions, the Hull & White method (which I have seen implemented by several people, including FinCad) and its analytical equivalent do have features not covered by Black-Scholes. I'm less sure about FAS123, but IFRS 2 takes the approach that the model used has to be materially correct. So if Black-Scholes gives a materially correct result, it is fine. But equally, even Hull & White, if inappropriate in the circumstances, cannot be used.Because of this emphasis on materiality, the answer to the question "what is the best method?" is a weigh-up between simplicity, objectiveness and precision. There are some incredibly complex methods out there, designed by actuarial firms, but I wouldn't call them better in all cases.

glipman raises some good points - IFRS2 is less specific about models than FAS123 and any model could be either invalid or vaid depending on the circumstances (FAS 123 allows for the use of BS but states a preference a lattice model).biggest problem as I see it for a lot of models (including HW2004) is estimating inputs such as future forfeiture rates and suboptimal exercise factors., which are more behavioural than almost any other derivative inputs.PS : sorry not to be in touch Guy, haven't done much on ESO's recently as I've been working on other projects

What is actually most important for transparent accounts is not the model you use but the underlying assumptions.http://www2.eycom.ch/publications/items ... n.pdflooks at this in some detail.I dont know much about accounting for employee options, but the accounting standards for pesnions miss out simple factors like - durationand convexity of liabilities and longevity assumptions which mean it is very difficult to compare firm with firm or re-mark to market if you get shifts in the bond markets or stress test.Problems with option accounting are clearly lack of transparency on vol assumptions (or underlying pdfs).Its worth noting that there is going to be a trade off between accuracy and transparency. for example I doubt anybody belives FX prices follow lognormal distributions but the industry standard for quotes is to use Garman-Kohlhagen.As somebody once said - accuracy before precision......

On this issue I have implemented the FASB 123 compliant stock options pricer (here) using the approach of Hull with Binomial and Trinomial Trees as described in his work. Anyway, this seems to be the standard approach used within software implementations since it is relativily straight forward.