Description

I am long the shares of Synopsys (SNPS) and think it is a good holding for the long-term.

Fellow VIC member ‘madmax989’ recently wrote-up its peer Cadence Design Systems (CDNS US), and SNPS itself was also written up by other members three times previously.  I recommend the reader to re-visit that prior work for an explanation of the business model.  I believe EDA/IP vendors are some of the best businesses in the market.

In this write-up I will focus on 3 discrete elements of my thesis on SNPS, below. 

1. Software Integrity business likely to be divested, and business could be made more efficient

For the past 10 years, both organically and through acquisitions, SNPS has expended a lot of effort to build up its Software Integrity business.   This was in noticeable contrast to CDNS, whose focus was more on building-up a greater presence in areas like CFD (computational fluid dynamics) for auto/aerospace/medicine and molecular design for biomedicine.  In prior years, SNPS's Software Integrity business grew >15% but in most recent quarters this decelerated to <8%.   The Software Integrity business brings down the profitability of the company.  SNPS’s EDA and IP divisions both generate adjusted operating margins of >30% but Software Integrity is c.15%.  In recent years, as Software Integrity’s margin ramp and revenue growth both began to disappoint and investors’ appreciation of CDNS’s contrasting expansionary focus on physics/medicine grew, this increased the investors’ perception of Software Integrity as ‘de-worsifying’ SNPS and being a distraction.  I believe synergies with EDA/IP were tenuous at best.  So I think there was some relief when 3 weeks ago SNPS announced it is exploring strategic alternatives for the Software Integrity business.  Ideally this should mean a sale.  

What price could the Software Integrity business fetch today?  I estimate that since 2014 SNPS spent $1 billion on business and asset acquisitions in this segment.  The largest deals were Coverity, Black Duck and Codenomicon.  For instance, Black Duck was bought for $547M in 2017, when Black Duck was on a $50-60m revenue run-rate (i.e. 10x EV/sales).  SNPS’s Software Integrity business currently has LTM revenue of $525M revenue and LTM adjusted op profit of $76M.   Within the equity markets, perhaps cyber-security companies are the best comp.  Over the past few years, cyber companies have gone from being glamorous and well-loved, to being considered the worst segment of software, to now being in vogue again.  So from that POV, the timing is good.  Recurring high-profile news about cyber-ransoms and data-breaches in the market are also helpful.  Cyber companies currently trade between 6x-12x EV/sales and generally >25x EV/adj EBIT.  If Synopsys can sell its Software Integrity business for $3.5B (6.4x EV/sales, 46x EV/adj EBIT) I think that would still be accretive for strategic acquirers who have adequate existing sales infrastructure and can take significant cost out.  If the sale is to a financial buyer, the price would likely be significantly less.   In any case, the net result is that SNPS will be a more simplified business with a more focused management team.   In recent years SNPS’s management philosophy was to take the prodigious FCF from the core EDA & IP businesses and spend them on Software Integrity sales & marketing and buy more software targets.   Putting this division up for sale is a belated admission that the effort hasn't really worked out.    With Software Integrity gone, this will free up management to focus more tightly on the existing business’s growth and margin profile.  Madmax989 in his recent CDNS write-up produces several charts showing SNPS’s inferiority on revenue/employee, margins and % ROE.  Getting rid of Software Integrity is conducive to focusing SNPS’s management on narrowing this gap.   This in turn should support SNPS’s multiple (see next point, below). 

2. Valuation differential is favorable

SNPS trades at a persistent discount to CDNS.

The general perception in the market in recent years has been that this is mostly deserved but I think this will fade, and some of the reasons for this are explained below.  

3. The multiple is warranted

Technology increasingly embeds itself into our daily lives.  For private individuals and for businesses, the effective costs of not being proficient at technology continue to increase.  Semis are a key driver of technological progress, and equity in a critical EDA oligopolist like SNPS is essentially a leveraged bet on said technological progress.  All of the signposts for the continued strong growth of EDA are there.  Designers like NVDA and AMD continue thriving and are growing the numbers of electrical engineers on payroll. Hyperscalers ie MSFT, AMZN and GOOG, also are thriving and also continue growing their in-house silicon design teams.  Semi design startups continue getting successfully funded by a mix of VCs and strategic investors.  The AI software startup fundraising environment is a feeding-frenzy.   Governments are increasingly recognizing the strategic value of semis and are finally beginning to disburse some of the subsidies from the packages that they have been signing; manufacturing is first to benefit, but eventually funding will arrive to establish new R&D institutes and boost EE/STEM grad intakes.   To drive performance/power/area/cost improvements, the semi design and fabrication industry is moving beyond pure 2D planar shrink and moving into 3D architectures like GAA as well as chiplets / advanced-packaging, increasing the complexity of design and manufacturing significantly.  With regards to AI ( - and for various reasons I do not think we are there yet, and I do not expect us to get there in the short term -) eventually it will enable SNPS to offer the client new tools helping to design and verify significantly faster; more value means that SNPS will be able to charge for it, boosting average ASPs.   Altogether - the fundamental value & importance of what SNPS provides to the world will only grow.  So the sales CAGR of >10% can be extrapolated beyond the traditional 5 year investment forecasting period.      

To sum up - yes, SNPS costs 61x forward GAAP EPS in the market, but I believe it has at least 5 more years of >10% top-line growth ahead of it and critically, beyond that period it will still continue growing at a healthy clip, too.  Operating leverage and share repurchases will translate to EPS growth higher than that.  The company has very little debt and significant net cash; this financial structure is inefficient given the solid business model, and I think the company is a lot more leverageable.  In 5 years time (Dec 2028) I have the company doing $22 GAAP LTM EPS.  Getting a 10% IRR would require an exit multiple of at least 40x LTM GAAP P/E which I think for this business will remain defensible.   

Catalyst

Industry design starts

Earnings releases

Conclusion of the strategic options evaluation process for Software Integrity