November 3, 2009

Competitive analysis of IT companies

Warning: A long post on the competitive analysis of IT companies (low in entertainment value :) ). So please get a cup of coffee or tea before you continue further

I recently received a comment from madhav

The question I have on outsourcing kind of IT companies like NIIT, Infosys, TCS etc is, "where is the moat?".

Every company seems to be into everything that happened yesterday, today or will happen in the future. All companies are generally present in all geographies, across all industry sectors etc. To top up the challenge, the "asset" of such IT companies are their people, but the employees keep hopping between the competitors and there is hardly anything preventing them from doing so. So where is the moat or where is the long term advantage? This also leads to the question - how do you value such a company?

This is an interesting question and there are several ways to answer it. I will try to answer it, by first doing a porter's five factor model analysis on IT companies (for more on this model you will have read this book). I will then use the conclusions from this analysis to answer madhav's question and see if we can value these companies.

The porter's five factor model has the following five factors, on which the moat of a company can be analyzed (by the way, I do this analysis for every investment I do)

  • Entry barrier : Level of entry barriers in the industry to a new entrant
  • Level of rivalry : Level of competition within the existing companies
  • Supplier power : bargaining power of suppliers
  • Buyer power : bargaining power of buyers
  • Substitute product : presence of substitute products

I have a spreadsheet uploaded in Google groups, wherein I had done a similar analysis some time back for multiple industries. It is dry reading, but I think a useful document (for me). I am reproducing some parts below for this post, for the IT industry with appropriate updates.

Entry barriers: This factor can be analyzed in detail based on multiple sub-factors. I have listed the analysis in the table below. The summary of the analysis is in the first row

ENTRY BARRIER - No. 1 Factor deciding industry profitability

  • Moderate to high switching costs
  • Barriers due to economies of scale especially in the volume business
  • Some barriers due to vertical based competency (BCM / Insurance )

Asset specificity

Low. Mainly buildings and facilities.

Economies of Scale

Economies of scale important in recruitment, training and staffing, especially for outsourcing

Proprietary Product difference

None - IPR / knowledge base for vertical is the only differentiator

Brand Identity

To a small extent for specific verticals. However not too critical

Switching cost

High

Capital Requirement

High now, especially for the mid-size and large deals

Distribution strength

NA

Cost Advantage

High - but available to all. Scale adds to this advantage

Government Policy

NA

Expected Retaliation

High

Production scale

NA

Anticipated payoff for new entrant

Moderate at the low end

Precommitted contracts

High

Learning curve barriers

Moderate

Network effect advantages of incumbents

None

No. of competitors - Monopoly / oligopoly or intense competition (concentration ratio )

Intense competition



The above analysis clearly shows 2-3 main sources of competitive advantage. Scale is critical in this business as the larger companies tend of have cost advantages due to economies of scale and can also provide the requisite resources for large engagements. In addition, these companies can afford to spend higher amounts on marketing and sales. The second source of advantage is customer relationships (long term contracts). This advantage is not set in stone, but it a very critical asset. For ex: After the scandal, the key value in satyam, was existing client relationships and Mahindra paid for that. Ofcourse this asset does not have as much life as fixed assets and can be lost much more easily.

Level of rivalry:

RIVALRY DETERMINANT

Medium rivalry. However firms in the industry due to low exit barriers do not engage in destructive competition. Moderate to high growth has kept price based competition low in the past

Industry growth

moderate

Fixed cost / value added

Low

Intermittent overcapacity

Low

Product difference

Low

Informational complexity

Medium to Low

Exit Barrier

Low

Demand variability

Low


The above analysis shows that the level of rivalry has been high, but not destructive till date. Most companies in the sector earn high return on capital and are fairly profitable. This has been mainly due to high growth in the industry and low fixed costs (they can cut our salary and bonus when the demand drops :)). Due to multiple companies in the industry, the long term returns in the industry are bound to trend lower (read that as profit margins).

Supplier power

SUPPLIER POWER

None - Input is manpower

Differentiation of input

None

Switching cost of supplier

None

Presence of substitute

None

Supplier Concentration

None

Imp of volume to supplier

None

Cost relative to total purchase

None

Threat of forward v/s Backward integration

None


If you work in the IT industry, you are the supplier. Supplier power – zip, nothing..doesn't exist. Yes, companies say employees are their asset etc etc. We all know the reality. Employees are the raw material for the industry like steel and copper (sorry if I hurt your feeling by comparing you to a commodity :)). Most companies pay for this commodity based on what the market prices it.

Buyer power

BUYER POWER

% Sales contributed by Top 5 account. High for smaller companies

Buyer conc. v/s firm concentration

Varies for companies. Tier II companies have higher Buyer conc.

Buyer volume

High for Tier II companies

Buyer switching cost

High for buyers

Buyer information

High

Ability to integrate backward

Low. The reverse is happening


Buyer power is clearly a bigger issue for smaller companies. The large IT companies have consciously tried to diversify their revenue to reduce dependence on any specific client. This is a key variable for a company. If the buyer concentration is high, the vendor can get squeezed and will not be able to make high returns.

Substitute product

Substitute product

Substitution is feasible with another vendor. However switching costs are high. Hence repeat business is key variable

Price sensitivity

High for low end work

Price / Total Purchase

High

Product difference

Low

Switching cost

Medium

Buyer propensity to Substitute

Medium to high


Substitution of one vendor with another is a key competitive threat for each company. Clients typically have multiple vendors to ensure that they can maintain competition and keep the prices low. Till date, the competition has not been destructive and most companies have made decent returns in the past.

Conclusion

The broad conclusion one can draw from the above analysis is that IT companies do enjoy a certain degree of competitive advantage. The source of this advantage is no longer the global delivery model (everyone does it) or the employees (all the companies source from the same pool). The key sources of competitive advantage can be summarized as follows

  • Switching cost due to customer relationships
  • Economies of scale
  • Small barriers due to specialized skills in specific verticals such as insurance, transportation etc
  • Management. This is a key source of competitive advantage in this industry and explains the wide variation of performance between various companies operating in the same sector with the same inputs and under similar conditions.

Inverting the question

Let's assume for argument sake that the industry does not have a competitive advantage and is similar to the steel or cement industry (which by the way has some competitive advantage). In such as case, the industry would be characterized by intense competition and low returns on capital (low ROE). This has not been the case for the last 15 odd years and most companies especially the larger ones have maintained fairly high returns on capital. This variable alone shows that the industry has some level of competitive advantage – especially the larger ones.

Valuation

The above analysis is clearly a backward looking exercise. Valuation on the contrary requires a forward looking estimate. Can we arrive at any conclusion from the above analysis?

It is difficult to arrive at how each company will evolve over the next 5-10 yrs (the typical duration required for a valuation). However we can arrive at some general conclusions

  1. As in other industries, the return on capital for the industry should come down over the course of next 5-10 yrs
  2. The industry could split in two levels – the large SI (system integrators) such as Infosys, Accenture, Wipro, IBM etc and the niche players. Both these type of players should enjoy a decent level of profitability.
  3. The industry is likely to diversify and expand into new geographies, but the future growth is unlikely to be as high for the big players.

The above conclusions are my educated guess and are as valid as anyone else's. However based on these conclusions I would propose the following

  • The large SI like Infosys, WIPRO etc should continue to do well. However, these companies would see only moderate growth in profit. As a result I would be hesitant in giving a PE of more than 25 to these companies.
  • The attractive returns in this sector are to be made with the small niche players. These companies, if they can be indentified early enough, are likely to have high growth and profit. However this is a specialized form of investing, requiring deep skills in the specific sub-segments.

Are you still reading? Wow!! ..If I have not put you to sleep, leave me a comment :)

27 comments:

mkd said...

Hi Rohit,

Nice analysis. Insightful and crisp.

I agree with the point that profit margins will trend lower for the bigger established players.

Is it possible to draw a parallel with Accenture/IBM on how their margins trended over a period of time relative to their services business size and maturity? Compared to them it seems that most of top homegrown IT companies are enjoying margins from la-la land.

However as with any market situation, there will be upward pressure on wages. Moreover 2008 has forced lot of companies to look at their costs more closely and to try and consolidate. While this would mean bigger volumes for bigger players, it would probably be at cost of margins.

If companies purely stick to services business, then they can be treated like utilities over the next few years. Low growth rates and predictable free cash flows. Hopefully resulting in predictable dividends.

I would personally be not very comfortable at buying such business at 25 PE (especially in inflationary environment). However it completely depends upon individual's perception.

But one should look out for unknowns like - IP based selling, Product selling, Transaction based selling etc. All of which, would break the linear relation between top line and cost.

Rohit Chauhan said...

Hi mkd
this is the fastest comment ever !! almost 5 min after the post.

points you raise are great. i will cover them in detail in the next post

personally, i think the indian SI should eventually converge to the same margins as an accenture ..maybe a few % higher.

on IP based stuff..i doubt the large indian companies have it in their DNA to get into that meaningfully

A PE of 25 is my upper estimate of intrinsic value ..not really a buy poitn

rgds
rohit

Anonymous said...

Hi Rohit - Great analysis (was awake at the end of post)!!

What do you think of pure BPO players? With ongoing economic crisis, do you think India BPO players will get any advantage like more business / better margins?

Thanks
Prashant

KD said...

Hi Rohit,
A good analysis, industry for which people started writing a separate valuation perspective!!
Personally I believe that due to cost structure and management quality, Indian SI would be having higher growth and earning higher bps margins than Accenture / IBM.
Niche players (or lets call them "product" based cos.) may have more advantage, but in long term may be acquired by SI's looking for higher growth. In any case your judgement of <=25 PE remains valid.

Anonymous said...

You sell yourself short. You are not as boring/dry as you might think.

Unknown said...

Dear Rohit,

You have missed the most important variable in the analysis. IT services are basically labor arbritrage games. You must take currency variations also into account. Now, that most of our IT services businesses are from USA and dollar being a doomed currency, what will happen to the profitability or even sustenance?

Amit Mehta said...

Hi Rohit

I agree with your conclusion that returns will be made in small niche players and there are several such available such as Infotech, KPIT Cummins, Rolta etc..any thoughts on which would do well?

regards

Anonymous said...

quite an interesting analysis. found it useful. thanks!

Anonymous said...

Why make things complicated when it comes to investing? KISS.....kep it simple and stupid....thats the key!!!!

Unknown said...

Hi Rohit,

Talking of IT companies here are a couple of my picks.
1) Tata Elxsi
2) KPIT Cummins

I have bought tata elxsi and have never regretted it.Almost a 100% return from a price of 82 it reached during the march lows.Also I came across KPIT from a great writer in Outlook money magazine.Even in the US I am hearing great things from its client and also the work environment is great.
It hit a low of Rs.20 I donno where it is now. These are the 2 IT companies I have been attracted to.

Thanks,
Hari

Rohit Chauhan said...

Hi prashant

i think pure BPO players are in a position IT companies were 7-8 yrs back. they should do well..i own hinduja TMT in that space and they continue to do well

rgds
rohit

Rohit Chauhan said...

Hi KD
i think the indian SI's will continue to have a margin/ cost advantage for some more time ..how long ..dont know. their overheads are still lower than other vendors such as accenture etc
rgds
rohit

Rohit Chauhan said...

hi nihal
just joking ..ofcourse industry analysis would hardly be an idea of fun for most people :)

Rohit Chauhan said...

Hi anwar
outsourcing is a definitely a cost arbitrage. thats true for all outsourcing ..but that does not mean that indian companies have a competitive advantage on that compared to other companies. that was the source of advantage a few years back ..now almost all major SI such as IBM have access to the same advantage ..so cost differentail is not a sustainable source of competitive advantage which is the focus of my post

on the point of dollar as a doomed currency ..its anyone's guess ..if not dollar then what ? currencies are always relative ..its dollar v/s ???

yen or euro ..each have their own problems.

the contention is that the developed countries could be in for a long term slowdown ..well then it impacts all companies and the lower the cost structure the better. so indian vendors will do better on that aspect than the US based ones

rgds
rohit

Rohit Chauhan said...

Hi amit
sorry ..dont have much insight into niche players ..not really within my circle of competence

rgds
rohit

ms420 said...

Hi Rohit,

Thanks for taking the time to write a post on based on my question. There is tons of useful information in your write up.

I really like the analysis you have provided using the porter's five factor model; I didn’t know about such a thing and I will probably buy his book.

Entry Barrier:
Until now, I thought that any body with a computer and programming skills could enter IT development business and so entry barrier was low. However I have to agree with you that barrier entry is Medium due to factors like domain expertise and scale across regions to be able to take up a large scale project or engage with companies across geographies. Based on this analysis I agree that we cannot compare a company like Infosys to a small IT company of 10 people and compare them purely on cost.

Switching Cost:
You have mentioned that switching cost is high. I am not sure how you arrive at such a conclusion. Even if this is true for an on-going project where a new company needs time to come to speed, this would probably not hold true for a new project which hasn’t started. In such a case an RFP can go to multiple vendors and any vendor could be chosen on the merits rather than the reason that there is an existing relationship with a vendor.

Supplier Power:
Though man power is the only raw material, isn’t the same being bid up by different buyers (companies) in a bull market? As a result, in a bull market is there is a impact on the margins due to salary pressure?

Just wondering that if you were to analyze Infosys or Wipro as a possible investment, they would probably rank similar on the factors you have mentioned. Would one have to look at financial ratios, leverage and the like to determine the right investment?

Thanks a lot once again.

Unknown said...

In the IT discussion, I guess we got side-tracked from the original topic - Rejected Ideas!

Rohit - what do you think about Man industries? Is it a Graham stock - discount to BV, decent dividend play? Or is it a value trap?

Thx,
Arun

Rohit Chauhan said...

Hi matt
very valid point. the objective of the post was to do a five factor analysis alone. non- market factor are crucial too ..though i have yet to find a model to do it and hence i end up doing it in a general manner. will have a look at the article in the link

rgds
rohit

Unknown said...

Excellent article on doing industry analysis. Would you please provide further explanation of each sub-category on Porter's 5-force analysis or suggest a link?

Great work and keep it up.

Unknown said...

how can I access the ggl spreadsheet you referenced in the article?

Thanks

Watch Investments said...

Excellent Analysis. I have added this article as a link on my website. I am maintaing this website as a collection of all value investing blogs/articles/news.

http://watchinvestments.blogspot.com/

Sandesh said...

wouldnt considering all the five forces make it too complex to analyyze?
why not just focus on EOS and customer captivity?
i guess entry barriers alone would suffice to tell whether the industry has moat
http://www.chillibreeze.com/articles_various/IT-Industry.asp#a

Rohit Chauhan said...

hi dipesh
you will have to read the book for further understanding of the sub points

watch investments - thanks for the link

hi sandy
yes in some cases you are right. prof greenwald in his books has said the same point. i still find doing the 5 factor useful in a lot of cases

rgds
rohit

Indrajeet Singh said...

Great Work Rohit

Just one thing Do you work full time on investing?

Rohit Chauhan said...

hi indrajeet
no i dont ..it's a side thing ..mainly to compound my own capital fast

rgds
rohit

$udeep said...

ohh man this all things takes time to me to understand .As i am in search of investors and entrepreneurs,i need to learn more from this blog most informative...
thanks...
http://www.venturegiant.co.in

Unknown said...

Hi Rohit,

It is excellent to see such an insightful analysis. But the currency risk still rankles me. I am not able to get any analysis on IT industry if the exchange rate goes below 35Rs to USD.

The possibilities are real. If the real estate collapse had not happened in US, then it would have already happened in my analysis. Remember the oil surge in 2007.

I have a deep suspicion that JP Morgan and Goldmann Sachchs exited the real estate portfolio and betting heavily on dollar depreciation as the fundamentals of US economy is very very weak. The only way to bring down the USD is through oil trade.

It never really matters to India if the oil trades at 200 USD per barrel as long as the exchabge rate is 20 RS to USD. Only the Americans would suffer. But the entire economic scenario would go through a sea change and nobody is ready for that yet. But it cannot be postponed indefinitely.

My interest is to see what will happen to IT industry in such a scenario?

And what would happen to industries that are import intensive like refineries and petrochemicals, coal power plants and export intensive like IT, BPO, iron ore and pharmaceuticals?

If we have some sort of a primer on that it would be really useful as we cannot rule out this scenario in near future...