Study: Notes on Stock Based Compensation (SBC)

SBC Valuation can be Grossly Wrong

The most common types of SBC are Restricted Stock Unit (RSU) or Performance Stock Unit (PSU). They granted to employees and top executives and the value of such grants are recognized in the P&L as an expense, it is also reflected in the cashflow statement. It is important to understand how the value is calculated. RSUs typically have a vesting schedule and the fair value is calculated on the initial grant date under a straight line schedule (or accelerated amortization schedule).

For example, you received RSUs of a company during which the stock traded at $100/share. The company granted you $100k worth of shares (1,000 shares). If the vesting schedule is recognized in a straight line over four years, you will receive 250 shares of the company after each year of employment for the next four years.

No matter what happens to the stock price afterwards, the company will recognize $25k of SBC expense in each of the next four years. Therefore, the SBC number you see in the P&L and cashflow statement is based on the estimates on the initial grant date. Now, after year 1 of vesting, you received 250 shares and sold them immediately. You are about to receive the next 250 shares after year 2, but the stock price fell 90% and is trading at $10/share today. So these 250 shares would be worth $2.5k currently, but under GAAP, the P&L will recognize the same $25k of SBC expense; a gross overestimate!

The converse is true. If the stock price rose to $1,000/share, then the true economic cost of SBC is $250k, but the P&L underestimates by recording the same $25k of expense.

This means GAAP does a poor job in estimating SBC reality when things are very volatile.

Extreme volatility is far too common for individual stocks, especially for tech stocks in the post-COVID period. At first, upside volatility in 2021 and then downside volatility in 2022.

In other words, the actual economic cost of SBC was underestimated in 2021 and likely to be overestimated in 2022. So it’s not surprising that SBC as % of revenue did not go down in 2022 because it was overestimated.

Dilution

Since there are limitations of SBC numbers we see on financial statements, can we just look at dilution each year?

Unfortunately, this approach has limitations too. We suspect many investors tend to infer too much information from a specific year’s dilution than there actually is. If a company diluted its shareholders by 1% this year, it may not necessarily mean that it is the intended strategy to dilute shareholders by 1% or less. It may just mean that the stock is trading at high multiples, and they can pay good compensation to their employees without diluting shareholders too much.

We have to ask which option is a relatively fixed number:

1. Number of shares offered to employees

2. Dollar amount of SBC per headcount

We think the answer is likely to prove to be much closer to #2. If you do assume management pays employees a fixed dollar amount of stock every year, GAAP accounting is closer to reality over the long term than modeling for dilution which can be more volatile depending on stock price in any particular year.

An illustrative compensation package at tech firms is usually like this:

1. $200k base salary

2. $450k stock vested over 4 years

3. 15% bonus on base of $200k.

Therefore, year 1 total compensation is $342.5k.

Of course, the actual total compensation is dependent on stock price. But companies don’t think about the stock component in terms of number of shares, but mostly from the point of view of dollar amount.

If they indeed want to dilute the shareholders less, they will have to lower the total compensation. When you look at big tech OPEX per headcount compared to other tech companies in Silicon Valley, it is hard not to see the material competitive advantage big tech enjoys today in recruiting which may be strengthened during the downturn.

There are only a handful of big tech today, and these companies will make sure the cost of entry remains exorbitantly high.

Common Valuation Mistake

When doing valuation, a very common approach is to model dilution in terms of share count in the interim years and multiply the terminal FCF by an exit multiple. Here, FCF = cashflow from operations less CAPEX.

This method does not recognize SBC as an expense but rather considers it as a claim on equity. Moving SBC from an expense to a claim on equity is what creates dilution.

This approach assumes there is an inherent economic strategy for companies to manage the rate of dilution. Theoretically, whether treating SBC as expense or dilution should give the same result. But in practice, we think that companies operate on dollar terms when it comes to compensation.

Instead, we prefer to deduct SBC from FCF in the interim and terminal years. This way we treat SBC as an expense because of the point we raised above: companies think about stock compensation in dollar terms and not share count.

Other Implications

Remember that SBC is an estimate on the date of the grant. The returns of the stock price between the date it was granted and the date that it vests can distort earnings, for example, with taxes. If the stock goes up, the employee earns more than the amount implied by the SBC expense, and the tax deduction from compensation increases. The greater the deduction, the lower the tax.

Reverse is true for decreased stock price. For example, META enjoyed a $1.1b tax benefit in 2021. The stock had risen at 37% CAGR in the prior three years. In 2022, the stock fell 64% and its provision for taxes increased $471m, raising the effective tax rate by 2%.