India Equity Research Reports, IPO and Stock News Visit http://indiaer.blogspot.com/ for complete details �� ��
HCL-Technologies
Overweight
HCLT.BO, HCLT IN
A progressive approach so far but work still lies ahead
• It’s more than the Street’s traditional growth versus profitability debate:
The Street has largely focused on the tradeoff made by HCL Technologies
(HCLT) in favor of revenue growth at the expense of margins. However, we
think an important question not asked relates to the scalability of the business
portfolio that HCLT is constructing. Less profitable business if scalable can
improve profitability due to scale benefits.
• Has the AXON acquisition delivered so far? AXON was intended to: (a)
lead customer engagements downstream, (b) enhance the revenue/margin
profile, and (c) accelerate HCLT’s positioning in higher-value enterprise
solutions. Financials in the seven quarters following the AXON integration
indicate that synergies have been modest to date. But this is a longer-term
game and the better outlook on discretionary spending today should bode well
for AXON.
• Good hunters but farming could be better: HCLT has made good initial
wins/forays but we believe the company should develop larger client
relationships. In Dec 06, when Infosys was clocking HCLT’s quarterly run rate
(~US$800 mn), it had more than twice the number of US$40mn+ accounts as
HCLT does today (this ratio increases with client scale). The fuller-services
model results in business model scalability; HCL could sustainably narrow the
valuation gap relative to peers by increasing its full-services model to clients.
We think AXON will help here.
• Sep-10 margin performance has caused some concern but up to 50% of the
520 bps Y/Y margin drop is salvageable in due time: We think HCLT should
be able to operate at normalized EBIT margins of 15% in FY12E/FY13E after
the dust settles on margins during FY11. That said, some part of the margin
deterioration seen over the past four quarters may remain due to HCLT’s
business and geographic mix, higher attrition, and a relative lack of
institutionalized hiring apparatus geared toward freshers (rather than laterals).
• Valuations – don’t count on a narrowing of the valuation gap relative to
TCS/Infosys as margins settle in at a new (lower) normal: Conscious of the
new normal for EBIT margins at 15% versus 17-18% during FY05-08, we now
peg our one-year forward P/E valuation discount for HCLT to Infosys higher at
30% (versus 22% 5-year median discount) and at 40% on EV/EBITDA (versus
five-year median discount of 34%). Accordingly, we revise our Sep-11 price
target downwards to Rs465 (from Rs490). We believe HCLT is still an
evolving story. We retain our OW rating on the stock and advise buying the
stock on declines (at levels below Rs400) for 15% upside potential over 9-12
months.
Price target and valuation analysis
Our Sep-11 price target of Rs465 is based on a one-year forward P/E
multiple of 14x or EV/EBITDA multiple of 9.3x –reasonable, we
believe, with estimated 30% EPS CAGR FY10-12E. This embeds a
one-year forward valuation discount of 30% and 40% to Infosys on P/E
and EV/EBITDA, respectively, which we believe is fair and warranted,
given the margin profile and return ratios of HCLT. Our ascribed
discount is higher than the historical mean and we note that this
discount can widen if HCLT’s margin profile does not improve as we
expect.
Risk-free rate: 6.5%
Market risk premium: 8.0%
Beta: 0.87
Debt/equity: 10.0%
Cost of debt: 8.0%
Terminal “g”: 4%
Key risks to our price target are a slowdown in deal ramp-ups and
appreciation of the rupee against the US$. Taking on a higher-thannormal share of lower-margin and/or asset heavy deals can impact
operating margins and return ratios impacting HCLT’s valuation.
Introduction – Look beyond the negative widening spread
between growth at the top line and operating profit
Setting the context. Over the past five years or what we characterize as the full cycle
in Indian IT (CY06-CY10), HCLT has emerged as the fastest-growing company
among its Indian peer set (growth includes acquisitions as well which admittedly
distorts top-line growth comparisons). But unfortunately for HCLT, the negative
spread between EBIT growth and revenue growth is also a maximum for the
company (Figure 1) — a fact that has garnered significant attention on the Street.
This negative spread has widened even more in the past four quarters (Figure 2). We
believe that this trend points to specific aspects in HCLT’s business model worth
highlighting and the company’s inherent choices therein for the medium to long haul.
Exploring further in a nutshell. In this report, we look at the road ahead for HCLT
by dissecting the evolving direction of its business model and discuss its implications
for investors and investor returns. Over the years, HCLT has constructed a portfolio
of commendable strengths but in the process it has also acquired weaknesses, some
structural, that require deeper redress and others cyclical that are relatively amenable
to easier resolution. Interpreting this to financials, we believe that HCLT’s new
(lower) normal margins of 15% and lower conversion of profits to free cash flows is
likely to keep its valuation discount to larger peers higher than it has been in the past.
HCLT's thinking is progressive on several counts
We believe HCLT is progressive in its thinking on several counts and we like the
company for it.
• Newer models and technologies. It talked of infrastructure utility models,
cloud computing, new engagement models which are so much the
buzzwords today well before others started articulating it. It has set up an
incubation unit to focus on spawning commercial opportunities around
themes such as cloud computing and mobility computing.
• Thinking on non-linearity in BPO by going the platform way. It
acquired BPO platforms (Control Point Solutions and Liberata) with the aim
of making its BPO non-linear and differentiated (only TCS had that thinking
before HCLT with the acquisition of the operating platforms of the Pearl
BPO subsequently rolled up into Diligenta – its UK BPO subsidiary).
• Good hunters, in particular, in seeking out newer markets/domains. It
is good at penetrating new markets (e.g., good inroads into Asia-Pacific,
Middle East) and newer/less-penetrated verticals (such as
media/entertainment and pharma/healthcare). Notably, Rest of the World
(ROW) (ex-US and ex-Europe and now over 15% of revenues) has grown
twice as fast for HCLT as it has for peers.
In this context, we think it is a better hunter of new business than even some
larger peers such as Wipro. It makes forays into newer client relationships
on the back of its strength in infra management. Not surprisingly, revenues
from new clients as a % of overall revenues have been more than twice that
of larger peers over the past two years.
• Realizing that it is difficult to be a high-value transformative player
organically, we think HCLT made the right decision in acquiring AXON.
The fact that Infosys made the first bid for the target attests to the quality of
AXON, in our view.
• Differentiated contract structuring such as carving up turnkey
structures. HCLT is relatively adept at adopting new engagement models
with its clients. Differentiated contract structuring, some of which may
entail some risk, was necessary to change the terms in its favor to the extent
possible, given HCLT’s weak positioning in mainstream applications (as
opposed to R&D/product engineering and infra management) particularly in
dominant verticals such as BFSI, telecom and parts of manufacturing. Such
customized, sometimes bespoke structures work well in emerging markets.
and Asia-Pacific (the bespoke deal with the Singapore Stock exchange is a
case in point).
• Happy to invest in sales and marketing dollars to increase penetration.
HCLT’s aim to increase sales and marketing spend to address emerging
market opportunities is commendable but what we find surprising is that
this should bear on margins for a company that clocks over US$800 mn in
quarterly revenues. Cognizant, for example, invests on an ongoing basis
but has set margin expectations for long-term investors in an invariant,
narrow band (of 19-20% non-GAAP operating margins). It is unclear to
us why HCLT should attribute swings in its margins partly to SG&A
investments given its scale.
• A reviving core, which is engineering and R&D (~19%), but this must
lead to something bigger. HCLT’s traditional strength in engineering and
R&D offers embedded, mechanical and software product engineering
solutions to sub-verticals in manufacturing such as automotive, aerospace
and defense, consumer electronics, semiconductors, industrial
manufacturing. However, we believe that this is limited by the scalability of
relationships. HCLT is trying to parlay its strength in these accounts to the
enterprise side as well. This is an aspect that Wipro is also grappling with.
How can both R&D and IT be better cross-sold to technology and
manufacturing clients?
Yet, there are some structural issues
Some of the issues are structural which warrant concentrated attention while others
being cyclical represent low-hanging fruit.
Below we outline some areas where we think the company can improve upon.
Structural issue no. 1: Good hunters but poor farmers. Loses steam with
difficulty in ramping up relationships beyond critical thresholds
HCLT can bag deals on the basis of their innovative deal structuring and some might
say even by aggressive pricing. But we believe it falls short in scaling the
relationships beyond a decent threshold (US$30 mn+ annualized, Figure 3 which
shows that HCLT struggles to scale up client sizes to US$40 mn and beyond).
Simply put, we believe HCLT does well in building up new clients from scratch up
to US$30 mn but thereafter it struggles to ramp them up. In Table 2, we show it
compares well with Infosys on smaller engagements of less than US$10+ size but
divergence is sharp thereafter and only increases. The comparison has been done
with Infosys in 3QFY07 (Dec 06) when it was HCLT’s current size to normalize for
size.
Why could this be so? Better ability needed to sell multi-line services beyond infra
management. This was and is still a problem with Wipro for client account size
exceeding US$50 mn (with HCLT, the challenge begins at US$30 mn). HCLT finds
it relatively easier to sell new business to new/small clients than work much more
meaningfully with existing clients. This potentially affects HCLT’s ability to realize
economies of scale and hence, operating leverage as larger client sizes are more
likely to afford this. The weakness in BPO affects the multi-service paradigm selling
(Infosys estimates that about 25% of its BPO sales would be bundled along with
applications/infrastructure management; this proportion is increasing for Infosys).
In our view, on HCLT’s annualized revenue base of US$3 billion+, quality and
durability of growth can come by increasing the count of US$30mn, 40mn, 50
mn, 100 mn relationships, less so by working more extensively at lower levels.
This likely affects HCLT’s ability to realize economies of scale and hence,
operating leverage as larger client sizes will more easily give that.
Structural issue no. 2: Has AXON delivered so far after seven quarters of
integration?
Not fully yet going by two indicators:
1. Percentage of offshore revenues over the past seven quarters (since the first full
quarter post AXON) has remained stagnant at 72% (see Figure 4). One of the
principal intentions of AXON was to generate downstream business offshore
which we believe has not happened meaningfully yet (since Dec 08 when AXON
was fully integrated).
2. Has enterprise applications grown ahead of company average and peers post
AXON? The answer is not yet. AXON has not really been instrumental thus far
in enabling enterprise solutions to grow ahead of HCLT’s own company average
growth or of peers’ enterprise solutions division for six to seven quarters
Conclusion on why AXON has not delivered so far. In our view, this might have
to do with the lower downstream (offshore) capability that HCLT had on enterprise
solutions at the time of acquisition. This makes the company less able to extract
downstream benefits from the AXON acquisition. This is particularly so in the
verticals that AXON operates in (public services, energy & utilities) wherein
downstream capacity is limited. That said, we believe that this is a longer-term game
and there is time to extract benefits from AXON, particularly as the outlook for
discretionary spending continues to be bright.
Structural issue no. 3: Issues in BPO business affects winning integrated deals.
At peak, BPO accounted for almost 14% of revenues (in Q307), now down to 6%.
The company did well to ramp the BPO business to USD 200 mn+ by FY08 on the
basis of call-centre work and onsite acquisitions (such as customer contact centers in
Ireland and the US), but transition to non-voice and platform-based propositions has
been difficult.
It’s more than the current size of BPO. Current size in BPO may not hurt the
overall company revenue trajectory, given the significant scale-back already
experienced in BPO. However, BPO is increasingly becoming an important
component in multi-service relationships and HCLT’s BPO is still in turnaround
mode. However, losses in BPO have stabilized even if the company is not poised to
come to profitability in this division in quick time.
New head for BPO. On the senior personnel side, Rahul Singh, who headed eServe
(earlier Citi's captive BPO acquired by TCS) has come on board. We view this as a
positive.
Extremely high BPO attrition still represents a work in progress. Notably,
quarterly attrition (annualized) for in BPO stood at over 100% in the Sep-10 quarter,
indicating the fluid nature of the business. BPO is still a work in progress especially
on the people front.
Impact on margins. About a quarter of the Y/Y margin decline of 510 bps in the
Sep-10 quarter is attributable to BPO losses which have peaked (130 bps). We
believe a good part of this is recoverable in due course – we factor this in our
financial model.
Structural issue no. 4: HCLT has not institutionalized campus-recruitment
model especially for IT-Services.
HCLT’s per offshore employee cost is higher than that of Infosys by more than 20%.
This, among other factors, is reflected in the margins (of HCLT versus Infosys).
However, as HCLT institutionalizes a robust recruitment model in favor of freshers
(except in infra management) this should results in margin-improvement leverage.
In good years about three quarters (75%) of Infosys and TCS’ hiring is typically from
the campus; for HCLT, the proportion is only about 20-25%. This explains about
130-150 bps of the margin gap (the benefit given by the pyramid), not really that
significant toward margin gap closure given that the overall margin gap is
about 14-15 % points. So perhaps this is a small factor.
Incrementally tougher in a wage inflationary environment. But consequences of
dependence on laterals can get incrementally worse in a wage-inflationary
environment and HCLT could likely suffer in a market wherein demand for laterals
continues to be tighter than for freshers (and at which end there is maximum wage
inflation). In other words, we think HCLT is perhaps more vulnerable to wage
inflation, given its current recruitment model.
Thus, forced to hand out higher-than-industry average wage hikes. This is
occurring as the company doled out wage hikes higher than industry average which
dampened margins 300 bps in the Sep 2010 quarter (versus a normal impact of 200-
220 bps in the earlier years on account of wage hikes). We think that the recruitment
model of the company also had something to do with higher-than-industry average
wage inflation.
What is the normalized steady-state of margins is the
crucial question?
Margin performance has caused some concern but how much of the slippage is
salvageable? EBIT margin of 12.3% in the Sep 10 quarter (down 510 bps Y/Y) is
even below mid-cap Indian IT margins. Two factors may explain to what extent the
losses are salvageable:
• Losses in BPO explain about 25% of the 510 bps Y/Y drop in margins (i.e., 130
bps). This has stabilized.
• Salary hikes amounting to 300 bps of margin impact have been incurred in the
latest quarter (this also carries cost of higher offshore bench of about 60-70 bps
which must be excluded). Normally, salary hikes are roughly spaced out over two
quarters (June and Sep quarter) in the ratio 2:1, so HCLT did have an extra 100
bps impact on margins by concentrating its wage hikes in one quarter this time. If
we take out about 25 bps of this due to costs in BPO (included in one when we
consider operating losses in BPO), then HCLT has had an additional 75 bps
impact on account of margin impact on account of salary hikes given in one
quarter (as opposed to two). Including cost of bench (about 60-70 bps), this
amounts to about 135 bps of impact due to two factors that will normalize going
forward (utilization and front-ending of pay).
• That still leaves a gap of 250-260 bps Y/Y decline which is due to the structural
factors that we have earlier discussed, aggressive pricing and increased transition
costs on large deals.
So, while Infosys and TCS have maintained/improved their margins Y/Y, HCLT’s
margins have declined by 250bps Y/Y normalizing for the wage cycle, excess bench
and excluding BPO losses (believing that this will stabilize hereafter). Thus, we
believe ~ 50% of the Y/Y margin drop is due to systemic factors relating to the
performance of the business model.
Accordingly, for FY12/FY13, we have built in 15% EBIT margin representing
the new steady-state, normalized state (we believe that during FY11 the dust on
operating margins should settle). This is appreciably lower than 17-18% band
over FY05-08.
Resolution of cyclical issues is low-hanging fruit
• Slack in people utilization. With hiring robust in the past two quarters, offshore
utilization is about 400 bps lower than average. Back-to-normal trend utilization
could result in about 60-70 bps of margins improvement. This is in contrast to
larger peers (TCS/Infosys) that are running at fairly high utilization.
• High attrition despite the “employees first, customers second policy,”
however, this may moderate. Quarterly attrition annualized for HCLT stood
well above 30% for 4QFY10/1QFY11. In the latest Sep-10 quarter, HCLT’s
attrition (quarterly annualized) stood at a high 38% (versus 21% for TCS and
24% for Infosys). This is because of BPO where attrition in the Sep-10
quarter or 1QFY11 (quarterly annualized) stood at over 100%. Excluding
this, attrition in IT (software services + infra management), quarterly attrition
annualized stood at 22%, still high. This also impacts costs and utilization.
• This, we note, may have forced higher-than-industry wage hikes but with the
hikes already in effect, attrition may moderate hereon (also attributable to the
industry attrition crossing the pent-up stage).
Business model conclusion – it may make sense to slow
down and consolidate initial gains for greater scalability
We have discussed HCLT's several strengths amidst its weaknesses. On balance, we
believe that HCLT has displayed commendable progress in newer markets/verticals
on the back of its acknowledged strengths in infra management and technology
R&D.
We would like to see strengthening of HCLT’s internal processes for client mining,
account management, focus on building its full-services value proposition and rollout of that to its strategic clients.
Our view is that even if revenue growth in HCLT moderates slightly but the
company closes the negative spread between EBIT growth and revenue growth,
investors would be content.
Valuations: reversion to mean discount as a valuation
argument not likely
We believe the fortunes of the Indian IT industry are looking up and the commentary
of players for CY11 outlook is optimistic. That said, the industry may tend toward a
new normal and companies whose characteristics (e.g., growth, margins, etc.) change
also face a new normal in the context of the changing industry. In such a case, we see
the reversion to mean valuation as less of a likelihood (as newer secular factors have
emerged) and in the worst case an unreliable valuation argument.
With regards to HCLT, we believe one more round of margin disappointment would
lead to yet another round of EPS downgrades, which would amplify the current
valuation gap of 30% relative to Infosys. We believe that the company is sufficiently
forewarned on this aspect. During the recent crisis, the peak discount to Infosys of
HCLT’s rolling P/E touched 60% (during Feb and Marc 09 characterized by flight to
safety to Infosys). The five-year median discount of 12-month rolling P/E is about
22-23%. We take a more conservative approach rather than assume a quick and
complete reversion to mean discount of 20-22% especially if operating margins
settling in at the new normal post FY10 are appreciably lower than in the past.
Desired integration benefits of AXON should ideally have helped to narrow the
discount beyond the median as evidence would have been seen in HCLT’s
enhanced revenue and margin profile – this has not been the case so far.
We believe all of the operational issues mentioned herein should imply a valuation
gap of at least 30% even as revenue growth is still robust. Applying a target multiple
of 14x to one-year 12-month forward rolling forward PS of Rs32.7 applied to Sep-
11, we arrive at our price target (Sep-11 PT) of Rs465. We retain our OW on the
stock but would be more constructive on it at declines (below Rs400) for a 15%+
upside potential.
EV/EBITDA multiple-based valuation to normalize for historical earnings
volatility due to hedging/forex losses
HCLT’s earnings performance in the past was marred by swings owing to forex
losses (on hedges) and interest on debt. Now with HCLT substantially cutting back
its hedging and partly repaying its debt, we see a cleaner flow-through from EBITDA
to PAT and a better correlation between the two. To understand the true discount to a
top-tier player (e.g., Infosys) on a historical basis, we plot HCLT’s EV/EBITDA
(rolling 12-month forward) discount to that of Infosys. The median discount for the
last five years stood at 34% but we ascribe a higher discount of 40% in our target
valuations adjusting for the new normal. This gives us a Sep-11 PT of Rs475.
Figure 8 lays out the five-year median discount of 12-month rolling HCLT’s EV to
EBITDA to that of Infosys.
HCLT's current valuations are slightly below its five-year full-cycle median
Figure 9 shows how HCLT’s current valuations compare with its five-year full-cycle
median. Currently, HCLT’s valuations are only about 7% below its full-cycle fiveyear median. Looking at the recent trend in HCLT’s stock price, we believe that the
market is counting on a reversion to mean discount theme.
(Note: Only TCS has broken away from its 5-year median valuation and rightly so, in
our view, given the structural improvement in its growth/margin characteristics.)
No comments:
Post a Comment