Description

Totvs, in our view, is a value tech investor’s dream – a dominant, growing ERP software franchise with 90% recurring revenue, 85% software gross margins, 50%+ market share, and 95% customer retention trading at 3x recurring software revs.  The company generates strong FCF, minimal stock based comp, is capital light, has excellent governance, and pays out 60% of its net income in dividends. The company is in the midst of a license to subscription transition which we believe is artificially deflating growth and profitability.  We believe software revenue growth, which has recently grown mid-single digits, should inflect to double-digit growth next year. We think this is the most attractive software value in the market today. Insiders, who have been buying the stock recently, seem to agree.

If this sounds too good to be true, it is, to a degree – Totvs is listed in Brazil, and 95%+ of its revenues originate there.  We are not emerging markets investors, but for reasons we will outline in detail, think you can largely hedge the “Brazil” risk, leaving you with an exceptional, growing software franchise at a very attractive price.  Further, we think there are reasons to believe growth and profitability are on the verge of inflecting, based largely on contract provisions tied to inflation indices, not on any recovery in Brazil. For those concerned with Brazil risk and how we hedge, please skip to the bottom where we address this.  For others who are undeterred, read on:

Why the opportunity exists:
Software businesses generally are great businesses – highly predictable, high switching costs, cash flow generative with in long-term secular tailwinds.  Most software businesses can be run very profitably either as part of a larger organization or at a more mature stage of growth, which is why software businesses tend to trade at a multiple of software revenue (especially if they are growing fast or subscale).  We think the best way to value software businesses tends to either be on free cash flow or as a multiple of recurring software revenues (ex-license / service revs).

Latin America has two publicly traded software companies, both of which are listed in Brazil (Totvs and Linx).  The analysts that cover these names are Latam or Brazil focused telco analysts and generally do not have much experience or familiarity with software businesses globally.  When Totvs does roadshows, they meet entirely with emerging markets investors, who are not as versed in software businesses or license to SaaS transitions, which Totvs is in the middle of.  Furthermore, Totvs has a small breakeven hardware business and service business that have been shrinking modestly and is masking the underlying growth of its recurring software streams. The SaaS to license transition has also depressed revenue growth and operating margins and further make the stock “screen” poorly.

Software valuations have increased dramatically over the last 7 years.  Totvs’ valuation has moved the opposite direction:

7 years ago Totvs traded at 4x revs vs. global software comps at 3x.  Today global software comps trade over 6x forward revs whereas Totvs trades at less than 2x.  Totvs has a slightly disproportionate amount of zero margin service revenue relative to peers (all in GM of 62% vs. 71% for peer).  If we take 300M of service revenue out of our valuation to reflect this discrepancy, we’d still have Totvs at 2.3x revs vs. peers over 6x.

You can further make the argument that Totvs is much slower growth than the average software biz, which is fair.  If we isolate on companies growing revenues 5-15% a year that are profitable, the average forward sales multiple today drops to about 4.5x, still well in excess of the multiple afforded to Totvs.  We’d also note that when Totvs was trading at in excess of 4x revs, it was only growing 10-15% a year.

Overview of Totvs business:
Totvs has a dominant position in the ERP software market in Brazil and has had this position for many years:

Totvs is the original ERP in Brazil and over time has acquired smaller competitors.  They have strong positions in numerous industry verticals, including Manufacturing, Retail, Financial services, Agriculture, and Healthcare.  Its brand is strong and a source of national pride – we’ve spoken to competitors and customers who tell us they only buy Totvs solutions. Interbrand, a global brand consultancy, ranks the Totvs brand 20th most valuable in all of Brazil.    

Competition:
Despite competing against behemoths like SAP & Orcale, Totvs has done an excellent job capturing and holding onto market share.  Totvs has several advantages that have allowed it to fend off competition, other than brand and national heritage:

1) Localization – Brazil is an unusual market with complex tax and labor laws that often vary by each of Brazil’s 26 states.  Totvs built its software from the ground up catering to these nuances, whereas SAP and Oracle tried to adapt existing global software to the Brazil market.  While SAP has done a reasonable job with this (after many years of effort), Oracle has not been able to overcome this issue and for these reason has minimal local customers outside of those large enterprises based elsewhere that use oracle in their local subsidiaries

2) Pricing / Implementation / Support – Because of its legacy heritage, Totvs has a deeper consultant and installation ecosystem to support implementations vs. SAP & Oracle.  We’ve heard a R$1M SAP license may cost R$2M in professional services to implement, whereas a Totvs implementation often costs less than R$1M and can be done faster.

3) Strong legacy share and high switching costs – ERP software is incredibly sticky, which is why you will sometimes still walk into an office and see someone working in a program that looks like it’s in a DOS environments.  Almost everyone in the company uses it, and the disruption and distraction of a new ERP implementation are massive. Although this is purely anecdotal, I’d argue that more ERP migrations and implementations have been cited as creating business issues by companies on conference calls than all other types of software implementations combined.

4) Less robust start-up / cloud environment – Brazil is not Silicon Valley.  There is a much less robust incubator, private equity, and strategic acquirer presence in Brazil than in the US.  There are really only two meaningful acquirers of small Brazilian software businesses – Totvs and Linx. Everyone else is too small to matter, and Oracle / SAP generally do not buy local software businesses.  We believe there is much less risk to Totvs being disrupted by competition than exists in the US. If it does emerge, there is a good chance Totvs can buy anyone that poses a serious threat to them at a reasonable price, given the lack of competition for buyers.

A revenue model in transition

Totvs has LTM revenue of 2.25B, of which 1.5B is software (for simplicities sake we ignore the hardware and services revenue).  65% of software revenue is maintenance revenue, 11% is license, and 24% is subscription. 75% of license revenue is to existing customers, the majority of which we believe is for additional seats of existing implementations, rather than upsells.  If you consider this part of the business mostly recurring in nature (tied to business / employee growth), upwards of 95% of Totvs software revenue is recurring. Over the past 3 years, Totvs has begun to aggressively transition away from license revenue towards subscription revenue:

The math of this is pretty straightforward:  under the license model Totvs recognized all license revenue upfront and then roughly 30% of that as maintenance revenue each year afterwards.  Under the subscription model, Totvs recognizes only 57% of the original license revenue up front on an annual basis. We believe the breakeven point is at about 4 years at which the company begins to generate more from a subscription customer than they would a license customer.  In the meantime, this depresses revenue and operating profit.

As anyone who has researched SaaS transitions knows, the math is even worse than this on a quarterly basis.  In the quarter of a $1M license sale, Totvs would previously would recognize $1.075M of revenue ($1M license fee plus one quarter of maintenance revenues).  Under the new model, Totvs only records one quarter of subscription revenues ($143k) in the period a deal is closed. This is 7x less in the quarter than the ammt that is recognized under the license revenue model.  As the company has transitioned to this model, license revenue declined, while subscription revenue and bookings have been growing in excess of 30% a year since 2013.

We believe this conversion in rev rec has hurt revenue by 75M on an LTM basis.  Without it, we believe EBITDA would also be 20% higher than it is today.

Forthcoming Inflection in revenue growth
We believe Totvs will return to double digit revenue growth in its software business within the next year, assuming no material improvement in the economic environment in Brazil.  One of the most interesting elements of an investment in Totvs has to do with the way its maintenance and subscription contracts are structured. Contracts are adjusted each year, on their anniversary, for inflation.  Totvs’ maintenance revenues reset based on the IGP-M index, and subscription revenues reset based on the IPC-A index. Historically this results in annual price increases of 6-10% of maintenance contracts:

This was not the case for most of last year and early this year, where inflation turned negative.  In conjunction with this downturn in inflation, maintenance revenues turned negative:

Furthermore customer churn, which historically has been about 4-5%, has expanded to 8% in the last couple years due to the difficult economic environment in Brazil.  Assuming continued inflation in the mid-single-digit range we expect Totvs’ maintenance revenues to begin increasing again by Q4 of this year eventually accelerating to MSD growth in Q2 of next year, driven by the majority of contracts having repriced by then.  This combined with continued subscription growth should propel revenue growth to double digits in software by the middle of next year.

Insider Buying
In June 2018 insiders purchased over R$3M Real of stock.  We estimate members of the board purchased R$2M (inclusive of the CEO, who we assume bought the bulk of shares) and members of the executive committee outside the CEO purchased R$1M.  In Brazil, insider purchases are disclosed in aggregate (e.g. Board of Directors or Executive committee) but not by individual, for safety / security reasons. We believe the CFO and several other senior execs invested a meaningful amount for themselves personally at nearly R$28/share in June 2018.  
Insider buying in Brazil is not particularly common, and has less signaling power than buys in the US do since individual buyers are not disclosed. Totvs’ discloses total management cash compensation, including the CEO, at approximately R$13M for 2017.  This implies that management used 25% of pre-tax cash compensation in 2017 to purchase shares this year.

Conservative accounting understates earnings relative to peers
US software companies are notorious for paying substantial amounts of stock comp and, in many cases, capitalizing software development that probably should be expensed.  Academically, ignoring these costs artificially overstates profitability – the more a company capitalizes software and pays employees in stock comp vs. cash the more it can inflate its adjusted ebitda.  

Totvs is very conservative in this regard relative to the median software company and also relative to the 1st quartile.  It also is favorable relative to its only Brazilian peer, Linx:

Valuation & Scenarios:

Bear Case:
It’s relatively easy to see how we make money here: If we are right and Brazil’s economy doesn’t go haywire, we should make somewhere between a modest amount to a lot of money.  If Brazil faces a situation with meaningful inflation and currency devaluation, given the currency and market hedge, combined with the inflation protected contracts and the defensive nature of the business, we should be okay.  I’d argue the worst environment for Totvs is the one we experienced last year, with minimal inflation and a still shaky economy. If we assume subscription revenues slow dramatically and maintenance and license revenues continue their slow bleed, we would expect the business to look largely flattish over the next couple years from a revenue and ebitda standpoint.  Then the question becomes how much this could de-rate. We think the likely answer is not very much, given Totvs is already trading close to its historical lows on an EV / Revs basis:

We’d eat about 300bp of financing on the forward currency contract.  If you want to say this could re-rate to 2.5x recurring software revs, below anything this traded at historically and below just about any other software company in the world, we see downside to $21 a share.

Base Case:
In the most likely scenario, we see maintenance revenue growth rebounding as the effects of inflation flow through, license growth remaining stagnant, and Totvs’ subscription business continues to grow robustly.  In this scenario, we see a return to double-digit growth in software and ebitda margins rebounding to the high teens by 2020:

Note that the historical revenue multiples we have included above are inclusive of nearly 500M of low to no margin service revs.  Excluding all hardware revenue (but including services revenue) and valuing the business at 4x revs, where it traded for most of 2010-2013, and closer to global peers, we would arrive at a stock price in excess of $50 a share, up over 100% from here.  This would also have the business trading roughly in-line with public peer Linx, which is experiencing similar rates of organic growth and worse cash flow conversion.

Bull Case:
There was a time not long ago where emerging markets in general were all the rage, driven by the belief in a growing global middle class.  Brazil GDP was growing MSD (and had done so for the greater part of 2003 to 2013). While we don’t underwrite a return to some modest growth in Brazil after 3 years of a nasty economic environment, it doesn’t seem outside the realm of possibility.  What could Totvs look like in that scenario? The biggest delta we believe, in this scenario, is seat growth in underlying customers as Totvs’ customers return to growth. Seat growth historically was a meaningful driver of license revenue (and associated maintenance revenue growth).  Despite transitioning newer customers to subscription offerings, most of Totvs installed base is still on license based deals and must buy additional software licenses if it adds employees who will use its software. As the economy tanked, sales of license revenue to existing customers dropped dramatically:

Although Totvs does occasionally upsell additional modules to existing customers, our understanding is that the bulk (75%+) of license sales to existing customers are from seat growth.  We think a decent proxy for license sales from additional seats should be as a % of the maintenance revenue base. As seen above, this ratio has dropped dramatically over the past several years.  Interestingly, as the economy has stabilized, this ratio has been improving.
In our upside scenario, we assume license revenue from existing customers increases again to 18% of maintenance revenue, license sales to new customers remain flat, subscription revs continue to grow 30% a year, and maintenance revenues grow as a function of new license sales, 6% inflation, and 5% churn.  Doing so would result in the following financial picture:

Ascribing a 4.5x revenue multiple, ex hardware revs, to 2020E revenue would get you a stock price of $70 a share, nearly a triple from today’s levels.

Brazil Risk:
Emerging markets have been a landmine this year.  Brazil is no exception. We have no opinion on who will win Brazil’s elections next month or where its currency is headed.  Country and currency risk is the single biggest risk to an investment in Totvs, but is also reasonably easy (and cheap) to hedge.  For smaller funds or PAs, EWZ is a Brazil ETF with no borrow cost that can hedge both currency and market risk. For larger funds, your prime broker may be able to create a custom small/mid cap Brazil basket.  Forward contracts on the currency are also readily available and generally pretty cheap (we estimate this costs about 250-300bp a year on the notional exposure you want). Furthermore, as we will go into in greater detail later, the entirety of Totvs recurring software revenue are indexed to inflation, which helps protect from a hyperinflation scenario.  Totvs software is also critical to the functioning of the businesses that use it, and if they stop paying most cannot continue to function as a business. The general resilience of the business can be seen in Totvs’ revenue performance the last several years relative to Brazil GDP, which had its biggest decline in nearly 3 decades from 2014 to 2016:

We would also point out that Totvs has excellent disclosures, English financials and English conference calls.  8 of the 9 board members are independent. There is no controlling shareholder. As far as governance goes, I would argue Totvs would be above average in the US, especially for software businesses, which frequently have dual class shares or founder CEOs who control the board.

Catalyst

**Upcoming Roadshow in US targetting software/tech vs. emerging market investors likely in early October.  Contact Sergio (IR) at [email protected] or UBS for details.  Will post more detail as this firms up.

**Resolution of Brazil election and removal of political uncertianty 

**Continued growth in recurring revenue

**Inflation escalators kicking in heading into 2019 and return to double digit revenue growth, along with associated operating leverage

**Call option on some type of economic recovery in Brazil