Accenture results showed a drop in billings….for all their talk about future Ai revenues their billings number shows the reality…….if the biggest IT company is not getting orders what to expect from the rest……It might be undervalued…..but need to get some indication whether orders are coming for them otherwise underperformance of IT might continue.
Discl: Personal opinion, no IT positions as of date.
This seems to reflect in how only AI/datacenter and related themes stocks seem to do exceptionally well in an otherwise stagnated market. Shows investors are crazy in private space as well and flock at same place…the only difference is that shrewd ones have among private ones have a solid profitable exit plan on day 0 of investing….
coming to AI and IT services thing….although (i have disclosures in earlier posts that) indian IT is my core holding and have been adding to positions…was viewing an interview of vishal sikka on cnbc yesterday where he mentions that some project where 500 odd team worked 12 hours a day and for around 3 years created something and sometime back he trained AI and it completed that entire project in a single day….it took him several days to come out of that realisation….
what this means to me currently is that for indian IT opportunities will be huge no doubt but the way talent will be used and billed will change dramatically….a gradual shift would be ideal but cant say….
Also i feel they can play a leading role in such shifts but no company would want to do radical changes in way they operate so while they understand & acknowledge the change, they will take it as it comes being ready with action plans on every milestone that hits them….
I maybe wrong in all above….
although i have exited completely out of the IT Trade; my points remain majorly the same :
This doesn’t mean that IT services will benefit.
At gross level this is what is happening:
Software service demand is not reducing. It is slow growth.
But automation and AI is leading to lower headcounts. e.g. 2 engineers being replaced by 1 engineer + 1 AI license(which is about 200 USD)
Now customer knows this maths, they are not going to continue paying for 2 engineers. So topline takes a hit. At best you can maintain your bottom-line.
Market does not like de-growth. Investors would rather buy bonds. Accenture dividend yield is equal to long term bond yields.
Indian IT companies are trading at 50% premium to US IT companies due to scarcity premium and they are huge % of Indian Mcap, not easy to churn for institutions at one go.
Ace investors could see the writing on the wall. He called it more than a year back
There is a tendency to believe that, GCC in India is competition to IT services. It is partially correct in my opinion.
Some of these global companies have their Indian development centers from 1996-97 days, such as Microsoft Hyderabad. Same is case with Adobe, J P Morgan (Malad West/Goregaon East) since 2007-08, Few German companies at NESCO since 2010-11, Few MNC(s) were there in Pune since late 1990(s). Also, Cognizant, Accenture, Capgemini are there in India, since past 2-3 decades. So this competition is old now. It seems that, the scale of competition has increased.
At the same time, most of these Global Companies generally keep premium Innovation, Product R&D, Conceptualization of Next Gen Products, and the Core Software Design and Development platforms in Developed Nations, and Generally use Indian development centers for Tier-2, Tier-3 platforms, products, infrastructure management, testing work. They look at these Indian development centers as Low Cost Development Centers for non-premium Products and Platforms.
I think, the context here is different. While MNC(s) will continue their presence in India, The real revenue limitation for Indian IT services might happen due to AI tools. Will they charge same for the Product / Platform / Service developed using AI and Non AI Tool OR Do they have to give discount for that ? Many companies already use Output Based / Fixed Cost models for charging customer and the impact of person hours is not always 1:1 co-relation.
We have to observe at least next few quarters to understand the impact of AI, on revenue, fixed cost models, margins, PAT and EPS.
well he is maverick in Investing industry. I respect him a lot, hardly any interview of him i miss. But, all this is always true in hindsight, his earlier calls were also about the auto industry due to pressure faced via EVs world over, then he also had a view about Ola and physicswallah etc etc.
All in all he is still a great investor and I respect him a lot, but singling out like this does not work.
Further, people like Rajiv Thakkar have the opposite view if you look at the portfolio construction of PPFAS.
We always need to look the silent evidence that nobody talks about with respect to such hindsight success stories and related stuff.
My views are based on my personal experience.
IT stocks may show some value bounce back once in a while. Chances of it showing degrowth is higher than chances of it getting back to double digit growth.
If bounce back is what one is looking at, it is better played thru Accenture and Cognizant at ~10 PE rather than paying ~15 PE for Indian companies.
I just quoted Samir Arora to give context.
Just a personal opinion…I do not see IT competing with AI for efficiency, cost or rather anything. AI is in itself a part of Technology and this part helps IT services or any business become more efficient and do “more” work & produce “better” results.
So, for a patient company of any industry, AI will be a tool to do better & more work and same is true for IT services….
Btw these days the hype of AI has reached extreme levels that in some school’s curriculum they have introduced subject “AI” in class 5. Add to it, I just listened to lesson where they were talking about using AI in farming, when actually what they are using is “technology”.
In a way, this indicates that, when any term becomes extremely popular, even every school/college/all industries/all banks/all insurance companies will also may claim that they are also using AI!! This is one of the reasons, why an investor should ignore many of such hyped claims and focus only on Core Business of IT services industry.
Check whether they are using AI for :
(1) Customer Satisfaction.
(2) Customer feedback and participation in Sprints/Iterations.
(3) Product Backlog and Sprint Backlog generation.
(4) Iteration and Sprint Planning.
(5) Customer Requirements, Mapping it to User Stories and Prioritization.
(6) Design Framework and Consistency of all the code with this framework.
(7) End to End Product Design, Architecture, Component Design, API framework.
(8) Integration with External API / Cloud / Government API / Income Tax API, and many other Cloud Platforms.
(9) Security Audits.
(10) Security Compliance / Data Compliance / Data Protection Act Compliance, and N number of such compliance checks.
(11) Procurement Planning and Execution.
(12) Selection of right Tool (AI and Non AI Tool) - Can AI be used for this without bias ?
There are so many things / tasks / activities which are beyond Design + Coding + Testing + Go Live. Also, there is massive Legacy Code written in COBOL, C++/UNIX, JAVA which has been in place for 20+ years as well. Can it be replaced by AI Code within few weeks and months ? Is that Risk worth taking for large MNC(s) ?
Once we can look into all these cases and see how much AI can be used, it will give us an idea about how much % of efficiency can be practically gained with AI in next 1 Year or so ?
Same hype was created for SaaS, PaaS, RPA and It was estimated that, Millions of Jobs will become redundant by 2015. We need to check whether that has happened ? What % of people actually lost jobs due to these 3-4 technologies. RPA was supposed to be the game changer for human being, has it lived up to that expectation ?
I think currently some Media and Experts are making some high level statements without probably knowing the Full IT Stack, Day to Day Work and Actual Reality at ground level.
I may be wrong in my analysis. It might happen that, 80% of above things can be done by AI Tools, but it may be more costly or it may generate incorrect logical blocks.
