Yesterday Microsoft announced some downgrades to its current employee benefits. Most noticeably, it changed the terms of its Employee Stock Purchase Plan (ESPP):
"Effective July 1, 2004, the ESPP program will be modified as follows:
- The discount for stock purchased though the ESPP will be adjusted to 10% from the current discount of 15%.
- The discounted purchase price will now be based solely on the market price on the last day of the offering period; the price on the first day of the offering period will no longer be taken into consideration.
- To increase the flexibility of the program, and better meet the financial needs of employees worldwide, you will now be able to purchase shares on a quarterly basis, rather than twice a year."
I have always been an active participant of ESPP because the plan offers very attractive financial returns. In an analysis I did back in August 2003, I calculated the annualized after-tax gain for money put into ESPP to be at least 58%, if participants will sell ESPP shares as soon as they are available for sale.
Now the new terms changed the landscape. In Microsoft's own confession, it estimated that the move will save the company and its shareholders $60 million every year. Detailed analysis shows that the new terms are good news for some, but bad news for more.
A quick calculation on the annualized after-tax returns based on new terms:
1) The average time (in months) money is tied to the ESPP program: (2.5 + 2.0 + 1.5 + 1.0 + 0.5 + 0) / 6 + 0.5 = 1.75 months.
* Microsoft pays biweekly at around 15th or 30th. Also sale proceeds from ESPP is available at around Jan 15th or Jul 15th if employee sells the share immediately.
2) Absolute Gain: 1 / (1 - 10%) - 1 = 11.11%
3) After-Tax Return: 11.11% * (1 - 25%) = 8.33%
* assuming the marginal tax rate is 25%
4) Annualized After-Tax Return = (1 + 8.33%) ^ (12 / 1.75) - 1 = 73.13%
Yes, it is not a mistake, the return under the new terms is higher than that under the old terms. That is because the average investment period is shortened from 3.25 months to 1.75 months, and my calculation of the old terms do not include the potential upside from the look-back feature.
Therefore, the term change is actually good news for those who cannot max out their ESPP contribution percentage (15%) because of liquidity issues (a lot of money will be tied for up to six months). Under the new terms, they can probably contribute to the 15% limit and reap a better gain.
But for most others (like me) who contribute to the maximum, the absolute after-tax return is reduced from 13.24% of the total contribution to 8.33% of the total contribution, or close to 0.74% of the total eligible salary. Adding back the potential upside from the look-back feature, Microsoft immediately reduced its compensation cost of these active participants by around 1%. Plus, with the smaller margin of safety (11.11%), it is more risky to execute the immeidate-sales strategy I advocate.
Anyway, I will still stay in the program for now. The annualized after-tax return of above 70% still makes ESPP a good place to mark money.
P.S. The other two benefit changes are relatively unharmful:
1) Employee copayment for brand-name drugs that have an FDA-approved generic equivalent, and
2) 12-week parental leave (4-week paid and 8-week unpaid) now should be used in the first six months since child's birth (compared to 12 months in the current plan)
In the copayment example, it is almost a Pareto improvement because employees do not loss anything (albeit some small inconvenience to ask for a new prescription in some cases). The shortened period for parental leave usage is also reasonably reflecting the spirit of the parental leave itself.)