The Ira Sohn conference is a big deal in the world of hedge funds. It has become known as the place where industry luminaries announce their picks to the world, such as when David Einhorn explained his short position in Lehman during the earliest innings of the global financial crisis. 

Sohn also hosts an annual Idea Contest that anyone can enter. Among the biggest names on Wall Street participate as judges. It's a rich source of ideas that all investors should be aware of. But its recommendations should not be followed blindly.

This year's winner, Twilio (TWLO -0.78%), may be a case in point. 

Two analysts walk though a data center.

Image source: Getty Images.

Twilio's cloud communications platform makes it easier for developers to write software that can incorporate phone calls and text messages, enable two-factor authentication, and communicate with a product's customers via other channels. Notably, it's how Uber lets users call a driver within the app.

The winning pitch's author, Albert Wang, is a recent Princeton graduate who now works for Matrix Capital in San Francisco. Among other things, he argues:

  • One-fifth of the world's developers are now using Twilio, an increase of 10 times in three years;
  • Twilio is investing heavily to build exceptionally reliable infrastructure that will be difficult for competitors to replicate;
  • Management has made savvy acquisitions to build Twilio's profitability and defensibility;
  • Twilio's developer-first customer acquisition strategy is cheaper than the traditional approach of focusing on executives -- and it will create much higher margins as it scales.

Yet Wang fails to comment on the company's most basic issue: Twilio has never earned a profit. Not once.

In fact, Twilio has repeatedly diluted shareholders, in part to fund its losses. Since 2011, it has lost over $238 million in free cash flow (cash from operations minus capital expenditures) while issuing over $2.6 billion in new stock. 

Bulls argue Twilio is simply doing what all Silicon Valley giants did in their early days to quickly dominate their markets. Famously, Amazon lost money every year from inception until the fourth quarter of 2001. Twilio's CEO Jeff Lawson, a former Amazon employee himself, uses the phrase "Day One" thinking, a term borrowed from Jeff Bezos. 

But the 1990s also reminds us that for every Amazon there were thousands of "dot gones" -- stocks that never posted a profit until investors lost everything. 

So which is Twilio? Should you jump aboard? Or wait for Day Ten? 

Valuation, valuation, valuation

Warren Buffett is fond of saying that there are no bad risks, only bad rates. Ultimately, much of good investing comes down to making smart decisions about what price to pay.

One counterargument to Twilio's lack of profitability would be that it is profitable -- in the narrow sense of cash from operations. If we assume all of Twillio's capital expenditure (what a company spends to buy, maintain, or improve its assets after normal expenses are accounted for), is optional -- that is, all growth, no maintenance -- then the company is posting a $45 million annual profit.

Even if we assumed this, Twilio would still be trading for an astronomical 1350 times earnings at recent prices.

But the assumption cannot ultimately hold water. Twilio's capital expenditures, as Wang's pitch itself points out, are going to critical new infrastructure and acquisitions that are likely required to remain competitive.

In fact, very unusually for a public company, Twilio's depreciation and amortization costs have consistently exceeded its capital expenditures, sometimes by more than three times. As Buffett has explained in his concept of "owner earnings," D&A costs are often the best estimates of a company's real maintenance capital expenditure needs that an outside investor has access to. We ignore them at our peril.

So what? We got sales!

Ah, you say, the more common way to value tech stocks is as a multiple of revenue. That is true, and, here, Twilio is selling for a more reasonable-sounding 27 times trailing-12-month sales. This is roughly the multiple Amazon sold for when it went public.

But even in those early days, Amazon had already grown its revenue from $511,000 to $15.75 million -- a 31-fold increase -- in just one year. By 1997, Amazon was already generating more revenue in one quarter than it had the entire previous year.

Twilio has not demonstrated revenue growth at anywhere near these levels. It has compounded yearly sales at an impressive but by no means parabolic 54% since 2017. That includes during its recent successes in the pandemic, which managers at hedge funds like Lone Pine (who judged Wang's presentation) see as an accelerant. 

Unlike 1990s Amazon, Twilio has also pursued acquisitions to grow sales to even these levels, indicating it may already be near the limit of its organic growth potential. 

Wang's sole slide on valuation suggests Twilio will command a higher multiple once margins expand. But even this is a circular argument.

Bulls like Wang are really betting on a new company.

An unproven moat

In 2018, Reid Hoffman, founder of LinkedIn, published Blitzscaling. In it, he recounts strategies that helped him, Elon Musk, and Peter Thiel succeed at PayPal in the 1990s -- ideas that have influenced generations of Silicon Valley companies. Chief among them? Scale at all costs. Forget about profitability; forget, if you can, about regulation. Just grow!

The real moat for any new technology business, the thinking goes, usually comes from size, such as through an economy of scale.

Superficially, this appears to be exactly what Twilio is pursuing, spending capital to horizontally integrate, build infrastructure, and acquire new customers however it can.

But an economy of scale typically requires a company to grow bigger than both its suppliers and its customers. Today, Twilio already sits between giants: Oligopoly telcos like AT&T on the one hand and Big Tech behemoths like Alphabet on the other. 

Bulls like Wang are betting on future margin expansion, what Buffett terms pricing power. But the reality is that Twilio's pricing power remains very unproven. It's not clear why companies like AT&T or Google could not enter Twilio's business in the future -- unless its margins remain unattractive enough that they do not want to.

On this point, TMF's Leo Sun makes two great observations:

  • Twilio's non-GAAP gross margin (that is, the margin number management gets to calculate any way it wants) shrank in 2020, even during this pandemic period of apparently exceptional growth.
  • Major telcos are beginning to charge A2P ("application-to-person") fees for each text message sent to a customer using a product like Twilio. This could significantly change the company's unit economics going forward.

Both suggest Twilio's pricing power may already be waning.

Exceptional shareholder dilution

It would be naive to say that you should never buy a stock that is losing money. As TMF's Brian Withers points out, even if Twilio is burning cash, it has enough capital to do so for several years.

Yet even this timeframe may not be as long as it seems: Twilio's TTM FCF losses plus stock-based compensation expense total over $511 million, while tangible book value is just under $2.9 billion. That's an implied runway of less than six years!

The bigger problem is that a company with large financing needs can do well in its business and still devastate shareholders. This is especially true when it has a history of dilutive financing. Such businesses can be far more lucrative businesses to work for than to own.

To date, Twilio has diluted investors to the tune of $5.6 billion -- over 10% of its market cap and 3 times its 2020 revenue. Twilio issues new shares so aggressively that it wants investors to simply ignore stock-based compensation when it reports "non-GAAP earnings" each quarter.

If Twilio stopped awarding developers and other top employees these shares, would they stick around? What would happen to the company's prospects if they left?

Comparably, if Twilio could no longer dilute its stock to buy new companies, would it be able to continue acquiring them? What would happen to its sales growth if it could not? 

Twilio is far from the only new tech darling guilty of gluttonous stock-based compensation. It has sadly become rampant behavior, affecting companies from AirBnB to Palantir

But even compared to these peers -- and certainly compared to proven wide-moat players like Google -- Twilio's ratio of stock-based compensation to free cash flow seems extreme. 

As the table below indicates, Twilio is unique for having a negative ratio of stock-based compensation to free cash flow. Meanwhile, its ratio of cash from operations to stock-based compensation clocks in at 11X, more than 64 times Alphabet's.

USD rounded to nearest million





TTM stock-based compensation

$494 million

$3,385 million

$1,515 million

$13,966 million

TTM cash flow from operations (CFO)

$45 million

$1,482 million

$69 million

$80,859 million

TTM free cash flow (FCF)

-$18 million

$1,451 million

$62 million

$58,536 million

Ratio of stock-based compensation to CFO





Ratio of stock-based compensation to FCF





Source: Company SEC filings and author's calculations

The verdict

Bulls like Wang are right: Twilio is a potentially world-changing business. But, right now, it may not be the best investment. 

Lack of profitability alone should not dissuade anyone. But Twilio's steep valuation, unproven moat, recently changing unit economics, and abundant history of shareholder dilution are all cause for concern. 

This is likely not the best time to jump in.