Is the Indian IT sector now becoming a higher Beta sector? We think so as the business becomes more cyclical with size:: JPMorgan

We advocate that investors use a higher cost of equity in discounting cash
flow streams for large-cap Indian IT than they might have been doing so in
the past. Over 2004-08, we saw Beta for the large-cap Indian IT space as
contained between 0.5-0.7 over varying time periods (1-year, 2-year, 3-year and
so on). But over the past 3-4 years, this has noticeably shot up to as high as 0.7-
0.9 for the likes of Infosys/TCS (a reasonable 0.2-0.3 increase in the past 3 years).
The reason we think the Beta for large-cap Indian IT companies is exhibiting
a rise is due to the cyclicality of the business model. Why is cyclicality of
growth getting more pronounced necessitating a higher Beta? In our view,
two factors increasingly contribute to this occurrence: 1) clients’ IT spending
patterns and 2) companies’ large account sizes (esp. of the large-caps).

�� -->


 Factor 1: Clients’ IT spending patterns. Post-Lehman, we see client spending
patterns occurring more unevenly, (or in fits and starts/peaks and troughs, in a
manner of speaking). The keeping-the-lights-on type of spending is steady but
change/transformation-related spending is what is particularly uneven or nonuniform,
resulting in a cyclical pattern.
 Factor 2: As companies get bigger, the cyclicality of growth increases. The
large Indian IT Services companies are getting increasingly well-penetrated
relative to potential addressable spend in certain verticals such as BFSI
(particularly in flagship accounts such as Citibank, JPM, BOFAML within this
vertical) - they are so strongly entrenched in their anchor clients’ IT Services
spending that they might find it increasingly difficult to defeat/defy the ebband-
flow and seasonal patterns of large clients’ spending patterns.
 All of this is likely to translate into a rising Beta for Indian IT. As industry
cycles get shorter relative to the past, as the cyclicality of Indian IT companies
rise, and as Indian IT increasingly competes with and mirrors Accenture, they
could tend to approach Accenture in their business rhythm. Accenture is a
higher Beta stock (its Beta is ~1.1) than its Indian peers (Accenture’s Beta has
always been higher due to its higher cyclicality thanks to its consulting exposure
but this difference is narrowing over time as Indian IT’s Beta rises). Within
Indian IT, it is Infosys’s Beta that exhibits the sharpest rise (unsurprisingly so).
 Over what time frame should we compute the Beta? Theory suggests that 3-5
years’ rolling period is appropriate as this period is long enough to contain a
good part of the tech business cycle. But taking it as 5 years would lead us to a
pre-Lehman period, which is not representative of the world today, in our view.
Generally, a more volatile world economy post-Lehman should result in more
volatile or sharper business cycles, which means that Beta be correspondingly
higher. At the same time, a 1-2 year period seems rather narrow especially if this
period is too small to contain the business cycle or a substantial part of the cycle.
 A higher Beta has a bearing on valuation multiples, all other things
remaining the same. As opposed to DCF/Gordon Growth valuation approaches
where the cost of equity is an explicit input in discounting future cash flows,
investors who use multiple-based approaches (such as P/E, EV/EBITDA), must
recognize that a higher implied Beta translates into a lower multiple (all other
operating factors such as company’s earnings/cash flow & ROE profiles, or
market-based factors such as market risk premium, risk free rate remaining the
same). A rise in Beta by 0.2-0.3 impacts cost of equity by 1.4-2.0%, which in
turn adversely impacts the valuation multiple by as much as ~20%.

Table 1: Current Betas in the sector for varying periods (1-year to 5-year) ending 20th April, 2013
Time period for Beta Accenture Cognizant TCS Infosys
1 year 1.03 1.38 0.77 0.81
2 years 1.18 1.30 0.75 0.87
3 years 1.09 1.25 0.79 0.88
4 years 0.93 1.19 0.71 0.71
5 years 0.87 1.18 0.68 0.62
Source: Bloomberg, J.P. Morgan estimates
Table 2: Historical Betas for varying time periods ending Dec 31, 2007 (pre-Lehman) – Indian IT
companies’ Beta were lower than they are today by ~0.15-0.3
Time period Years Accenture Cognizant TCS Infosys
1 year 2006-07 0.92 1.72 0.30 0.46
2 years 2005-07 0.97 1.63 0.55 0.65
3 years 2004-07 0.93 1.77 0.63 0.72
4 years 2003-07 0.94 1.92 0.62 0.64
Source: Bloomberg, J.P. Morgan estimates