Salesforce: Dilution vs. Growth

Salesforce’s high valuation has been upheld for years by the company’s first-mover advantage and ability to increase its marketshare in the corporate migration to cloud-based information management software.  This has enabled Salesforce to avoid cash expenses, by using their stock as a substitute.

On February 24, 2011, Salesforce.com released 4Q FY 2009 earnings and its 8-K [pdf] updating guidance for next year to GAAP earnings of $0.08 to $0.11 and non-GAAP earnings of $1.35 to $1.38.  This is seemingly a large enough difference to warrant questioning during the earnings conference call.  Below is a word cloud from the questions and answers portion of the earnings call transcript.

Figure 1 - February  24, 2011 Earnings Call Q&A Word Cloud    

There was not one question regarding GAAP or non-GAAP earnings over the next year.  The words GAAP, earnings, dilution, or margins were not even uttered in the Q&A portion of the call.

Most investors seem willing to ignore dilutive stock policies during this “early land-grab phase” for business-cloud market share.  They believe once the market matures, Salesforce can fire all its staff, maintain market share, and cash in on 80% gross margins.  I do not see that golden scenario playing out, but even if it does I am confident Salesforce will dilute away the benefits to current shareholders.

Issuing stock to raise cash for expenses has become engrained into Salesforce’s business model.  Salesforce expects to accrue $1.57 per share in stock-based compensation expenses during this fiscal year.  Built into Salesforce’s share price is the expectation of strong revenue growth for a number of years.  Salesforce needs its sales team, consulting team, R&D team and their stock-incentives to acheive this growth.  Below, is a chart of shares outstanding since 2005.

Figure 2 - Historical Shares Outstanding

The window through which current shareholders benefit from the golden scenario unfolding is quickly closing shut.  Based on the stock’s performance post-earnings, some investors have read the 8-K and found the writing on the wall.  After closing at $134.32 on February 24th, CRM jumped after hours and opened the next day near its highs for the day at $146 before selling off to close at $138.83.  Since failing at this resistance level, CRM has struggled to find significant support.

CRM projects the weighted average diluted shares outstanding to equal 145 Million over the next year.  That means about 150 Million shares outstanding by Jan 31, 2012.    I think CRM will underperform the market until CRM shows a decreasing rate of shareholder dilution and an increasing trend in GAAP earnings and net margins.  I am short out-of-the money CRM LEAP calls.

Chart: VIX Futures Front Month Daily Roll

As a follow up to VXX in the 'New Normal' I have graphed the VIX futures front month daily roll yield based upon settling prices going back to 2004.  Although charts are available of the VIX short-term futures index historical performance, I have been unable to find a chart of the historical daily roll.  This chart helps visualize how the current consistently steep term structure did not exist in the 'old normal.'  However, late 2006 into early 2007 came pretty close with the cash VIX trading near ten.

The daily roll yield since July 2010 has averaged negative 12%.  This is the lowest 6 month average roll yield since VIX futures began trading in 2004.  For perspective, the average roll yield over the entire graph is negative 5%.

With a single exception on November 6, 2010 of negative 2.5%, the daily roll yield has not breached negative 6.5% since July 6, 2010.

An average of ~5% of the contracts in VXX must be rolled daily (100% of contracts / 22 trading days per month).  VXX provides current contract concentrations updated nightly.  A long-term investor's return in VXX is largely determined by the average daily front month roll yield during the investor's holding period.

Disclaimer: Dataset created via Excel manual process

Source: CBOE settle prices

Note that December 2004, and April, July and September 2005 did not have contracts.  In these cases, I graphed the difference between the two front-most contracts.  Some of the noise in 2004 and 2005 can be attributed to this.

Roll Yield = (Front Month - Second Month)/Front Month * 100

Disclosure(s): Short VXX at time of writing

VXX in the "New Normal"

The VIX futures (quotes) term structure is upward sloping, or in contango, the majority of the time and only enters backwardation during VIX spikes.  Someone used to a 1-2% front month negative roll yield in the oil futures may be shocked by the 10-20% front month negative roll yield that can exist in the VIX futures market.


To summarize from the VXX prospectus, the VXX rolls its VIX futures holdings daily to maintain holdings with a weighted average of 30 days until expiration. Someone with an infinite time horizon who never suffers a margin call from a VXX short is almost guaranteed to make a large return.  Unfortunately, no investor has infinite margin along with an infinite time horizon.  The fact that VXX could easily double in the course of a week makes people afraid to short VXX, as it's still easy to borrow.  


The key to shorting VXX is to make it a pure mathematical trade and give up trying to time an entry point.  One such strategy is to dollar cost average 10% of your portfolio into a short VXX position over the course of a year.  Then add to the position to maintain the short as 10% of your portfolio.  I have begun doing this by selling out of the money Jan '12 LEAPS, and using the delta of my positions as a measure of my short exposure.


Although 2008 changed everything, a look at historical VIX futures prices (available since the contract began trading in 2004) makes it clear the steep term structure that exists today was not around during the "old normal."  After the trauma inflicted upon money managers in 2008 they are willing to pay a hefty price for VIX futures as portfolio insurance.  It has become popular to be prepared for the next black swan by holding VIX futures or calls, and the expected lose in these positions is considered a new cost of doing business.


Until the VIX futures term structure flattens when the cash VIX is under 20, I will remain short VXX.  I think the steep term structure is here to stay though.  Otherwise, the insurance would be too cheap.  The possibility that something "Too Big to Bail" is only a few months away is too great.

Disclosure(s): short VXX at time of writing

2010 Annotated Performance Chart

Wish more people were as candid.

Graph from an Excel sheet where I keep track of account and S&P 500 daily close.  I'm annoyed that brokers don't provide this functionality.  Any brokerage could implement historical performance charting and index comparisons in a day.  My guess is they don't want their customers to be put off by poor absolute or relative performance.  Sites such as mint.com do a good job of filling this historical performance reporting gap, but get very confused about the value of a portfolio with short options or equities positions.

Futures Check

Looks like we're in for a futures ramp job to start the new year... as someone is trying to continue the 2010 pattern of a big up day on the first trading day of the month.  It has been even more pronounced when the first trading day of the month falls on a Monday.