Thursday, October 25, 2018

Twilio - Growth Isn't Reasonably Priced

Twilio Analysis
This must mean something to somebody.


Twilio, Inc. engages in the provision of communications software, cloud-based platform and services. Its developer-first platform approach consists of programmable communications cloud, super network, and business model for innovators. The company was founded by John Wolthuis, Jeffery G. Lawson, and Evan Cooke.

Founded: 2008
Number of Employees: 996
Headquarters: San Francisco US
CEO: Jeff Lawson

Analysis Methodology:

We'll keep TWLO's business model in mind as we analyze it.  They're an enabling company - they provide tools to help other companies deliver on their objectives.  They help their customers focus on their core business by providing a robust platform where they can quickly, easily and for less money (presumably)  build the ancillary features they need to be successful, namely customer support.  They're a B2B business using a SaaS model.  They are "developer-centric".  I interpret this to mean they are focusing on the developer in their sales cycle - creating something that solves problems for them and in turn, the developers become evangelists, paving the way for a "yes" with the actual decision makers. 

TWLO and the other companies in their peer group can be interesting.  They have the ability to dominate a niche, dig a decent moat and create value for shareholders.  I read one analysis on them which claimed they are a commodity business.  I don't agree.  Part of this analysis will look at competitive advantages now and into the future, as well as risks. 

This will be a peer based analysis.  None of the peers are competitors to TWLO, but they're all (except one) in the same sector and industry - technology services and packaged software.  Most of them are like TWLO-offering services to other companies.  It will be peer-based analysis because it's the best way to give the analysis numbers context.  It's also a chance to find other opportunities.

The analysis will search for an investment story in the three main genres: value, income or growth.  It will also review the quality of earnings.

Peer Group:

Stock Name (Symbol)Closing Price-Oct 17Market Cap
Trade Desk, Inc. Class A(TTD:XNAS)$128.285.2736B
Aspen Technology, Inc.(AZPN:XNAS)$104.057.2347B
Pivotal Software, Inc. Class A(PVTL:XNYS)$18.944.7719B
DocuSign, Inc.(DOCU:XNAS)$47.526.9989B
Okta, Inc. Class A(OKTA:XNAS)$62.266.2770B
Zscaler, Inc.(ZS:XNAS)$37.914.3474B
Paycom Software, Inc.(PAYC:XNYS)$136.517.5519B
Sabre Corp.(SABR:XNAS)$24.336.4709B
2U, Inc.(TWOU:XNAS)$66.933.6860B
Fair Isaac Corporation(FICO:XNYS)$212.766.0613B
LINE Corp. Sponsored ADR(LN:XNYS)$36.298.5574B
Nuance Communications, Inc.(NUAN:XNAS)$17.514.8426B
Open Text Corporation(OTEX:XNAS)$34.879.2849B
Trend Micro Incorporated Sponsored ADR (TMICY:OOTC)$60.358.3786B
Nutanix, Inc. Class A(NTNX:XNAS)$42.297.1667B
Twilio, Inc. Class A(TWLO:XNYS)$74.017.3005B

Quality of Earnings

It's a fact that earnings can be manipulated and changed by accounting-driven decisions.  We want earnings that are persistent and aren't the result of one-off events or management tinkering.  I use an nine part quality of earnings framework based off the work of two academics, Lev & Thiagarajan.  You can read their original paper here.  You can read my adaptation here.

The framework looks at nine areas in the financial statements: inventories, receivables, capital expenditures, research & development, gross income, selling-general-administrative expenses, sales per employees, tax rate and audit opinion.  The first two, the fourth, fifth and six are compared to sales levels, capex and/or r&d are compared to industry averages (I use a peer group average as a proxy), the tax rate measures seeks to remove the effect of an earnings bump from a reduction in the tax rate and the a last one looks for a clean audit opinion.  When any of these measures give a favorable signal, it gets a score of one.  All the scores are summed to get a total out of nine, the higher the better.

Most of the companies in the group don't have inventories.  The score for them is out of eight.   The four that do have inventories (TMICY,  PAYC, LN & NUAN) their score is out of nine.

Here are the results:
TTD          5/8
AZPN       3/8
PVTL        3/8
DOCU      4/8
OKTA       7/8
ZS             6/8
TMICY     5/9
PAYC        8/9
SABR        4/8
TWOU      5/8
LN             3/9
NUAN      4/9
OTEX       3/8
NTNX       6/8
TWLO      4/8

I don't consider TWLO's score of 4/8 to be a good result.  We have a strong basis for skepticism if any of the possible investment reasons involve earnings.

Companies to keep our eye on include OKTA and PAYC.


TWLO's management identified their key metric as base revenue growth according to their Q2 2018 conference call.  We'll also look at sales per employee but we'll start with earnings growth.

Year over Year Growth in Earnings

TWLO's been beating on quarterly earnings estimates lately, but I prefer to look at annuals - annual statements are audited and quarterlies are not and there's a lot noise in quarterly statements.
Let's see how many times the company has been able to grow its earnings year over year.  Each time the company increases its earnings relative to the previous year it earns a score of one.   We'll constrain it to five years of data as that's all that we have for TWLO.  The maximum score is four.

You have to hit the "Next 10" button to find TWLO as the results are organized from highest to lowest.  They scored one.  That was actually in the most recent period, which is a bit of a boost.
PAYC earned the highest score possible four out of four years of earnings growth.  I consider that a strong result, given it also did so well in the qualities of earnings section.

(There are notes to this section of the analysis, please check the notes section at the end of the article.)

Revenue Growth

The chart below shows the year over year growth in revenue.

Security Y/(Y-1)-1 (Y-1)/(Y-2)-1 (Y-2)/(Y-3)-1 (Y-3)/(Y-4)-1
TTD 0.5189 0.7826 1.5554
PVTL 0.2238 0.482
DOCU 0.3593 0.5229
OKTA 0.6216 0.8663 1.0948
ZS 0.5127 0.5651 0.4956
TMICY 0.093 0.181 -0.0516 -0.0207
PAYC 0.3157 0.4651 0.4885 0.4027
SABR 0.0667 0.1393 0.1252 -0.1371
TWOU 0.3929 0.3707 0.3624 0.3262
FICO 0.0577 0.0508 0.0631 0.0613
LN 0.1511 0.2992 0.2216
OTEX 0.2288 0.2559 -0.015 0.1399
NTNX 0.5067 0.7236 0.8429 0.8991
TWLO 0.4388 0.6615 0.8787 0.7798
Strong absolute numbers but year over year revenue growth has been declining.  The most recent growth is half what it was two years ago.  TWOU's numbers have been accelerating
Sales per Employee
Let's look at sales per employee.

Nice results here - they've got the second highest sales per employee of the group.


Although I don't expect this to be value story, it will still be useful to look at a couple metrics price/sales (PS) and price to sales growth (PSG) to get an idea of relative price.  I'm not using the PE ratio because TWLO had negative earnings in it's most recent year.


This metric tells us what it costs to buy $1 of sales.  The higher the ratio, the more expensive the stock is.  The rule of thumb is between one and two and less than one indicates its pretty cheap.

That rule of thumb would be for a broad universe of stocks.  I would bump it up for technology stocks like this, but not as high as this group.  The average is almost 12 and TWLO is well over it at 18.  The other two companies with possibility OKTA and PAYC are expensive too, but I would use an earnings based valuation measure for PAYC as they are profitable.


This ratio takes into account the growth rate of the company in calculating its relative value.  According to Backroom Analyst, a writer a Seeking Alpha, a PSG between 0 and .2 is the sweet-spot.  Obviously as the ratio increases it gets more expensive.

We can see that TWLO is slightly above average, however we've got a couple of stocks that might be skewing the results.  When I remove AZPN (sales growth of 3%) and NUAN (negative sales growth) we get TWLO just below the average.

(There are notes at the end of the analysis).


TWLO not surprisingly doesn't offer a dividend.

Hopes & Fears

The biggest competitive advantage a company can have is to be a monopoly.   That's not the case for TWLO and I don't think they ever will be. There are other players in their niche, perhaps not doing exactly what they are but companies do have choice when it comes to building out their customer care program. 

However, they can build a moat with an excellent product and their strategy of becoming the "developer's choice" might pay off.  It looks like they are actively expanding their suite of offerings, which is a good model.  If someone comes over for one solution and they like it, the hurdle to buy more solutions is lower.  Basically the more they can tilt the "buy versus build" (in this case it's "rent" versus build) argument in their favor, the better off they'll be.

Another advantage a company like TWLO can have is to be the dominant player in their space.  I don't think they've achieved that level of success yet, but there's ample evidence this is what they are striving for it.  They've made a couple of purchases of late, including one that was announced on October 16th.  Strategic acquisitions are a good idea for them to achieve fast growth and market dominance.  In my opinion, I think they were lucky to strike this all-stock deal at this moment, given their current price.

Their 'developer-centric' model is great so long as they keep an eye on the actual decision makers, cost/benefit is always an important consideration.  The SaaS model can be expensive over the long term.   Tangential to this is the need to hang on to customers as they grow.  They lost Uber and the reason seems to be that TWLO wasn't flexible on the pricing, Uber wanted a volume discount.  I'm an outsider and I don't know the full story, but I question whether that was a good move.  It raises a red flag for me about the management, it seems imprudent.

Another risk that other analysts have identified is a behemoth coming into their space, Amazon Connect, specifically.  Meh, competition is always a risk.

Conclusion - Growth at an Unreasonable Price

The only story I see at this moment and at this price is speculation, the belief that they're going to pull something magic out of their hat and keep the party going.  I would like to see growth at a reasonable price but they don't have earnings yet and with the exception of last year, they've been actually growing their loss year over year.  They have growing revenue, but the rate of growth is slowing down.  The two value indicators we looked at point strongly to an expensive stock.

I wouldn't buy right now and I'd consider taking profits if I were long.

PAYC could be interesting - it's got a good quality of earnings score and it's actually achieving earnings which have been growing, year over year.  It's expensive though.  Further analysis if and when the price drops could reveal an opportunity.

Analysis Notes

Year over Year Growth in Earnings - not every company had five years of earnings.  TTD, PVTL and DOCU only had three years of data.  The maximum score they could earn is therefore two.  OKTA, ZS & LN had four years of data, their maximum score is therefore three.
PSG - PS is calculated as price dividend by sales per share.  G is calculated as the year over year growth in sales from Y-1 to Y.


Part of intelligent investing involves taking on risk levels appropriate to one's circumstances.  We don't know what your's are and this analysis should not be construed as investment advice.  INVRS, its parent company, its officers, directors and employees cannot be held responsible for any investment decisions you make.

Sales Pitch

You can do amazing things with INVRS - build investment models, do peer based analysis and create investment reports.  Wouldn't you like to see for yourself?  Sign up for a free two-week trial and put the effort into learning the software using our myriad of learning tools.  If you're a numbers geek with a curious and creative mind, we're your ticket to unique investment insight.
Thanks for this analysis request, we hope you've enjoyed reading it as much as we enjoyed creating it.

No comments:

Post a Comment