Software and Services – Bottom Fishing

Adobe and ServiceNow are at bargain prices.

• The software and services sector has become inversely correlated to the AI trade, making it a great hedge against a correction, and at the same time, two companies in particular have caught my attention, which are Adobe and ServiceNow.
• Many thanks to William DeGale and BlueBox Asset Management, which provided the catalyst for this idea.

AI Scare Trade

• Most of the sector has seen no fall in earnings or growth, but their share prices have halved due to what is known as multiple compression.
• The trading pattern of any publicly listed stock is a function of the market’s confidence in the company’s profits, and particularly, are a factor of how fast it can grow as well as how reliably it can grow.
• It is this second factor that the market has lost faith in, as at a high level, it is easy to see how AI would allow companies to create their own software and no longer buy it in from outside.
• One of the reasons why there is no real software industry in China is that software engineers are cheap and plentiful, and so many companies hire software engineers and have completely customised code built internally.
• One argument for the “AI Scare Trade” is that AI makes software much cheaper and easier to write, and so many companies will no longer need to buy it, and the US market becomes more like China.
• Furthermore, if employees become more productive, one might need fewer of them, and as software companies generally charge on a per-seat basis, one can see how earnings might stop growing or go into reverse.
• This is why the multiple of earnings (PER) that the market is willing to pay for the sector has halved, resulting in a 50% decline in the share price of most of the sector.
• This fear has gripped the entire sector, and so if there are companies that will benefit from AI rather than be pummelled by it, then, as they have been hit just as hard by indiscriminate selling, they are opportunities for those with a strong stomach.

Slicing and dicing the sector

• The sector has a large diversity of companies, and so to find the AI beneficiaries trading at a discount, one needs to start sifting through the subcategories.
  • First, Consumer: which I do not want to touch with a barge pole.
  • This is because the consumer side does not depend on quality, reliability, litigation risk and so on and is very price sensitive.
  • Hence, if a consumer is offered reasonable image generation for free, that consumer will cancel any service that he or she was previously paying for.
  • Second, Data provision: which I also do not want to go near.
  • AI makes data gathering far easier than it has been in the past, and so I suspect that companies that make a living digging up data and selling it are going to get into difficulty.
  • We have seen this already with the valuations of Gartner and Forrester (who get a portion of their revenues from selling data) being crushed.
  • Third, IT & Customer support: which are functions that are relatively low on the skill set required to provide them.
  • These are far easier to replace with AI, and so I suspect that these companies are also going to have a difficult time.
  • Fourth, AI deniers: which is a segment of the market that will simply deny that AI will impact them.
  • These companies will fail to invest to get ahead of the wave and then get crushed as it breaks over them.
• This leaves me with companies that sell software to companies and not consumers, who have also embraced the AI trend to make the most of it.
• This is the place to look for those bargains as these are the ones that will thrive as opposed to merely surviving or being crushed.

Stock Selection.

• Within this bucket, I have identified two companies that I think are investing in the right way and will thrive as a result of the increasing rollout of AI across the enterprise.
  • First, Adobe: which is best known for its image and video manipulation software and its pdf document format.
  • The company has seen no impact on its services and, in fact, beat expectations at its last set of results and continues to expect the 10-12% growth that it has enjoyed for the last 4 years.
  • What sets Adobe apart is that 90% of its revenues come from companies or professionals who need pixel-level control over their creations and can not afford to be sued for copyright infringement.
  • This is what Adobe provides, which combined with the fact that Adobe has invested hard and partnered well, means that users can access all the models they want from the usual names.
  • This allows them to have complete control and legal indemnification while still using the image generation models that they want.
  • This is why I think that companies will actually buy more Adobe software, not less, as AI becomes more widely used in the creative process.
  • Adobe trades at 11.0x 2026 and 9.8x 2027, which is a real bargain for something that I think will thrive as a result of AI and not get killed by it.
  • I have assumed that the consumer business goes to zero over the next 5 years, which is why my estimates are a little bit below consensus, but at this price, it doesn’t matter.
  • Using a 9% discount rate, I calculate Adobe’s present value to be $440, offering around 70% upside.
  • I have started building a position in Adobe.
  • Second, ServiceNow: which offers software that allows all of the software silos within an enterprise to talk to each other, making all the data of the business available and actionable from a single platform.
  • This is precisely what any company that wants to make use of AI effectively needs to do, and the company has been quick to partner up with all of the model providers.
  • No company in its right mind is going to let an LLM roam all over its data and operations without an adequate level of oversight and control, which is precisely what ServiceNow offers with its AI Control Tower product.
  • Hence, I think this idea of subscribing to Claude and getting the bot to write all of one’s corporate software is a recipe for disaster, and while a consumer might try it, no serious enterprise will.
  • Hence, ServiceNow gives a company the tools to link everything together as well as the framework from which to run the AI agents in a way that doesn’t risk sending all of one’s corporate secrets to competitors by accident.
  • This is why I think that ServiceNow will continue to grow and may even accelerate its growth as AI rolls out across enterprises.
  • ServiceNow is also headed by the legendary Bill McDermott, who started his career with Xerox, selling photocopiers to offices in Manhattan and is arguably the greatest enterprise salesman in the industry.
  • This is exactly what one needs when working to convince the C-suite that this is how AI should be deployed, and this is his forte.
  • The fact that he has extended his contract to 2030 and seems to be in excellent health adds to my confidence, although there is an element of one-man risk to this company.
  • The financial performance is showing no sign of slowing down and may actually accelerate, but I have not included this.
  • The company is trading on 26x 12-month forward PER, down from an all-time high of mid-60s, and is showing every sign of continuing to grow at around 20%.
  • Using DCF with a discount rate of 9%, I value ServiceNow at $233 per share, some 113% above where the shares are now trading.
  • I have started building a position in ServiceNow.
• Overall, I have attributed a 5% weight to this idea in my portfolio, with the position split 60% ServiceNow and 40% Adobe when the purchases are complete.

AI Hedge.

• The broad sell-off of the entire sector, with almost all of it now in oversold territory, has created an inverse correlation with the AI trade.  
• For example, even though Salesforce’s results were a little bit disappointing on Thursday, the shares rallied 4% with the rest of the sector for no other reason than Nvidia was selling off, taking the rest of the AI trade with it.
• Hence, the sector has become a good hedge for anyone wanting to take a position that is related to an AI bubble and the consequences of it bursting.
• I think the best way to do this is by taking a position on the sector through an index or an ETF rather than stock selection, as the selections I have made here could easily fare worse than my highest expectations if the AI bubble bursts.