Vol. 121 th

20 Jan’ 2017

WHAT WE ARE READING – Vol.121  Distribution of Equity Returns in India Returns being higher than risk fee rate over 7 & 10 years  Taxation Matters Taxation Philosophy - and vision: Low, simple, stable and transparent  Global Growth Outlook in 2017 Faster Growth in USA is positive for the global economy  President Xi Speech at Davos 2017 Beijing’s’ desire to play a greater global role as United States turn inwards  Nomura’s Richard Koo share his views Trump election highlights conflict between society & market  French Elections Principal differences in policies between major presidential candidates  Netflix wants the world to binge Streaming Service has plans to expand globally & its begins in Sao Paulo  Aadhaar Linked Payment System Disruptor in Digital Transactions  Innovators should study the rise & fall of the Venetian Empire A lesson for Entrepreneurs

Strategy December 22, 2016

STRATEGY INSIGHT

The 1 year investment horizon offers the worst risk adjusted returns Using data for the last 24 years, we find that returns for the BSE100 are the highest over a 10 and 7 year investment horizon (in descending order of returns). The 3 & 5 year horizons give the lowest returns (see exhibit A in the right hand margin) while the 1 year investment horizon gives middling returns. However, the volatility of returns (using the coefficient of variation) for the BSE100 is the highest for the 1 year investment horizon at a staggering 4.4x the risk involved over a 10 year horizon (see exhibit B). Combining the risk and return perspective (using the Sharpe ratio) suggests that the 1 and 3 year investment horizons offer the worst risk adjusted returns (see exhibit C). The 10 year investment horizon offers the highest probability of beating the risk free rate The probability of generating returns in excess of 8% is the greatest over the 10 year horizon. The 1 year horizon, on the other hand, is doubly damaging – not only is the probability of the beating the risk free rate relatively low (54%), the probability of generating equity returns of -7% or less stands at an eye-watering 24% over a 1 year horizon. Once the investment horizon increases to 7 years or more, the probability of beating the risk free rate picks up meaningfully with the 10 year horizon offering the highest probability (see exhibits below). Investment Implications Given that equity returns in India are not normally distributed and given that the probability of equity returns being higher than the risk free rate is the highest over the 7 and 10 year time horizon, our Coffee Can Portfolio which focuses on investing for a decade in high quality franchises makes eminent sense (click here our note dated 17 Nov 2016). Every single Coffee Can Portfolio created in the last 16 years has beaten the risk free rate and the BSE100. The BSE100’s returns over a 10 year investment horizon are most likely to beat the risk free rate

No. of observations

Avg : 13%

80

-1SD

+1SD

120

Avg :17%

-1SD

10 Yr investment horizon 1 Yr investment horizon

+1SD

10 Yr horizon has a fat right tail

Median BSE100 returns (in %)

10% 10%

11% 9%

9%

3 Yr

5 Yr

5%

0% 1Yr

7 Yr

10 Yr

Source: Bloomberg, Ambit Capital Research. Period under consideration is from April 1991 – December 2016. The investment horizons are calculated on a rolling basis.

Exhibit B: Volatility of returns is the highest over the 1 year horizon 3 2.2 2 1.3

1.0

1

0.7

0.5

0 1Yr

3 Yr

5 Yr

7 Yr

10 Yr

Investment horizon Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016.

Exhibit C: The 10 year horizon dominates all other time periods on a risk adjusted returns basis 1.00 0.80

0.69

0.60 0.40

0.37 0.22

0.22

0.23

3 Yr

5 Yr

0.20 0.00 1Yr

7 Yr

10 Yr

Investment horizon Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016

Research Analysts

40 1 Yr horizon has a fat left tail 0 -50%

14%

15%

Investment horizon

Coefficient of variation of BSE100 returns

Using data on the BSE100 from April 1991 – December 2016, we find that median equity returns in India over a rolling 10 and 7 year investment horizons exceed returns from all other time horizons. Furthermore, these two investment horizons offer the highest risk adjusted returns and the highest probability of beating the risk free return of 8%. The one year investment horizon on the other hand offers middling returns (at 9.9%) with a high degree of volatility. In light of these findings, Ambit’s Coffee Can Portfolio, which invests in outstanding franchises for a decadal period, becomes even more attractive.

Exhibit A: Median equity returns for BSE100 are the highest over the 7 and 10 yr horizon

Sharpe Ratio

The peculiar distribution of equity returns in India

Ritika Mankar Mukherjee, CFA +91 22 3043 3175 0%

50%

100%

150%

BSE100 returns (in %) Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016. Avg refers to the average. The red dotted lines represent the average, average +1standard deviation, and average -1standard deviation for the 1 year investment horizon and the black dotted lines represent the average, average +1standard deviation, and average -1standard deviation for the 10 year investment horizon.

[email protected] Aditi Singh +91 22 3043 3284 [email protected]

Ambit Capital and / or its affiliates do and seek to do business including investment banking with companies covered in its research reports. As a result, investors should be aware that Ambit Capital may have a conflict of interest that could affect the objectivity of this report. Investors should not consider this report as the only factor in making their investment decision.

Strategy

Section 1: Equity returns in India are anything but ‘normal’ “Distributions of daily and monthly stock returns are rather symmetric about their means, but the tails are fatter (i.e. there are more outliers) than would be expected with normal distribution. The message for investors is: expect extreme returns; negative as well as positive.” – Eugene Fama and Kenneth French, Dimensional Fund Advisor’s website Conventional asset allocation frameworks typically make a range of assumptions about the “normality” of asset returns. An asset allocation based on the normal distribution typically understates the frequency and magnitude of negative events and hence the impact on total returns. In this note, we demonstrate how the notion of normality of equity returns does not hold true in the Indian context – across 1,3,5, 7, and 10 year investment horizons and how the 7 year and longer investment horizon is the best return horizon for an investor looking to beat the risk free rate with the highest probability. An analysis of BSE100 returns data in India from April 1991 – December 2016 suggests that contrary to popular belief, equity returns in India are nowhere close to being normally distributed. In specific, this analysis yields three critical findings namely that:  The 10 year and 7 year investment horizon offer not just the highest median returns (14% and 11% respectively) but also the highest risk adjusted returns (with a Sharpe ratio of 0.69 and 0.37 respectively) along with the most elevated probabilities of beating the risk free return of 8% (76% and 66% respectively).  The 1 year investment horizon offers middling returns with the median at 9.9%. However, the 1 year horizon entails disproportionately high risk (4.4x the risk involved over a 10 year investment horizon). Furthermore, the returns are far from being normally distributed and have a fat left tail.

Conventional asset allocation frameworks assume that equity returns are normally distributed

Equity returns in India are nowhere close to being normally distributed



The 3 year and 5 year investment horizon entail the lowest probabilities of beating the risk free return of 8% (with the probability at 51% and 54% respectively). Furthermore, the median returns in this time frame are the lowest (at 9% for both) as compared to the 10 year and 7 year horizon that delivers a return of 14.1% and 11% respectively. In the section below we provide details with respect to each of these findings. Insight#1: The returns perspective: 10 year and 7 year investment horizon offers the highest returns An examination of equity returns in India suggests that median returns for the BSE100 are the highest for a 10 and 7 year investment horizon (in descending order of returns). In contrast, the return outcomes over 3 and 5 years are the lowest (see exhibit below).

The 10 and 7 year investment horizon offer the highest returns

BSE100 returns (in %)

Exhibit 1: Equity returns for the BSE100 are the highest over a 10 year and 7 year investment horizon 20% 15% 10%

Median

17.2%

12.5% 9.9%

9.0%

Average

11.6%

11.0%

11.9%

14.1% 12.6%

9.1%

5% 0% 1Yr

3 Yr

5 Yr

7 Yr

10 Yr

Investment horizon Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016. The investment horizons are calculated on a rolling basis. For instance the median 1 year return is the median returns generated by considering 297 one year periods including Apr 91 - Mar 92, Apr 92 - Mar 93 and so on.

December 22, 2016

Ambit Capital Pvt. Ltd.

Page 2

Strategy It is worth noting that ‘median returns’ is a superior measure of central tendency as compared to ‘average returns’ since the median is defined to adjust for outliers/extreme values in the data. Additionally,  The median return for the 1 year investment horizon is middling at 9.9%. However, the 1 year investment horizon has the widest range of equity returns which ranges from a peak of 230% delivered in April 1991 – March 1992 to a trough of -57% (delivered over January 2008 – December 2008).  The range of equity returns narrows over the 3 year investment horizon from a peak of 61% delivered in May 2003 – April 2006 to the trough of -22% in March 2000 – February 2003. The median return for this investment horizon is 9%.  The range narrows considerably over the 5 year investment horizon. The early noughties saw some of the highest returns with the peak of 46.9% recorded in the period spanning November 2002 – October 2007 and the trough of -5.5% recorded over March 1994 – February 1999. The median return for this investment horizon is 9.1%.  The range of returns narrows further in the 7 year rolling period. The May 2003 – April 2010 periods saw the highest returns being delivered at 30% and the trough of -6.1% recorded over October 1994 – September 2001. The median return is the second highest for this investment horizon at 11%.  The 10 year rolling period offers the tightest range of returns with only one instance of negative returns (April 1992 – March 2002 at -1%). The peak return of 21% was earned in the periods spanning January 1998 – December 2007, February 1998 – January 2008, and May 2003 – April 2013. This is the period with the highest median return at 14%. Insight#2: The risk perspective: The 1 year investment horizon is the riskiest with risk levels being 4.4x that of the 10 year horizon An examination of the volatility of equity returns in India (using the coefficient of variation) suggests that the volatility of returns for the BSE100 is the lowest for the 10 year and 7 year investment horizon. It is worth noting that coefficient of variation is a superior measure of risk as compared to standard deviation as it is size agnostic and standardizes the risk so as to facilitate comparisons across time horizons. The 1 year time horizon appears be a particularly high risk bet as it tends to deliver middling median returns (third highest at 10% i.e. 0.7 times the median returns earned over a 10 year horizon) but entails risk (as measured by the coefficient of variation) which is 4.4x the risk involved over a 10 year holding period (see exhibit below).

The range of equity returns narrows as the investment horizon widens

The 1 year investment horizon has the highest volatility of returns while the 10 year has the least

Standard deviation and coefficient of variation of BSE100 returns

Exhibit 2: The 1 year investment horizon is the riskiest with risk levels being 4.4x that of the 10 year horizon Standard deviation 2.5

Coefficient of variation

2.2

2.0 1.5

1.3 1.0

1.0 0.5

0.7 0.38

0.5 0.16

0.11

0.08

0.06

0.0 1Yr

3 Yr

5 Yr

7 Yr

10 Yr

Investment horizon Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016.

December 22, 2016

Ambit Capital Pvt. Ltd.

Page 3

Strategy Additionally, 

The 10 year investment horizon clearly dominates all other investment horizons from the perspective of risk. The 3 year, 5 year, and 7 year investment horizon entail risk which is 2.8x, 2.2x and 1.5x times more than that compared to a 10 year period where risk is measure using the coefficient of variation.



Insight#3: The risk adjusted return perspective: The 10 year investment horizon offer the highest Sharpe/Sortino Ratio In a bid to combine the risk and return perspective we use the Sharpe and Sortino ratio which is defined as the average return earned in excess of the risk free rate per unit of volatility. The measure of risk is total volatility or standard deviation for Sharpe ratio and downside volatility for Sortino. So theoretically speaking, a portfolio invested in zero risk securities such as Government bonds should have a Sharpe/Sortino ratio of exactly zero. Thus the greater the Sharpe/Sortino ratio, the more attractive the investment is likely to be from a total and downside risk-adjusted return perspective respectively.

The 10 year investment horizon is the best placed on a risk adjusted perspective while the 1 and 3 year investment horizons are the worst placed

Quantifying the risk free rate in India We define the risk free rate for India as the median 10 year government bond yield which works out to be ~8% across time horizons. Furthermore, the risk free rate displays substantially low volatility in India (but not zero volatility) as measured by the coefficient of variation (exhibits below). Exhibit 3: India’s risk free rate as measured by the median 10 year government yield amounts to 8% Standard Coefficient of Investment Horizon Mean Median Deviation Variation 1 year 9.1% 8.3% 0.02 0.23 3 year

9.0%

8.2%

0.02

0.22

5 year

8.9%

8.2%

0.02

0.20

7 year

8.8%

8.0%

0.02

0.19

10 year

8.7%

8.1%

0.01

0.15

Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016. Note: SBI 1 year deposit rate is used from 1991-1998 to proxy for the 10 year government bond yield as data is not available for that period.

The risk free rate in India has been lower than the equity returns rate period (see exhibit below). Exhibit 5: Although the volatility of the risk free rate is not zero, it is fairly low

15%

2.5 Coefficient of variation

Median Returns (in %)

Exhibit 4: The median risk free rate is lower than equity returns across all time horizons

13% 11% 9% 7% 5% 1Yr

3 Yr

5 Yr

7 Yr

10 Yr

2.0 1.5

1.3 1.0

1.0

0.7 0.23

0.5

0.5 0.22

0.20

0.19

0.15

0.0 1Yr

3 Yr

5 Yr

7 Yr

10 Yr

Investment horizon

BSE100 median returns Median 10 Yr govt. bond yields

BSE100

Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016.

December 22, 2016

2.2

10 Yr government bond yield

Source: Bloomberg, Ambit Capital Research. consideration is from April 1991 – December 2016.

Ambit Capital Pvt. Ltd.

Note:

Period

under

Page 4

Strategy Given that the risk free rate tends to be constant across various investment horizons, the equity risk premium i.e. or the excess returns over and above the risk free rate tends to be the highest for the 10 year investment horizon followed by the 7 year and 1 year horizons (see exhibit below). Exhibit 6: The equity risk premium is the highest for the 1 year investment horizon followed by 10 and 7 year horizons

Equity risk premia

8% 6%

6% 4% 2%

The equity risk premium tends to be the highest for the 10 year investment horizon

3% 2% 1%

1%

3 Yr

5 Yr Investment horizon

0% 1Yr

7 Yr

10 Yr

Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016.

The risk adjusted return perspective: The 10 year investment horizon clearly dominates all other time periods An examination of the Sharpe ratio of equity returns in India suggests that the 10 year investment horizon dominates all other time periods with a material lead over the 7,5,3, and 1 year investment horizons (in descending order) (see exhibit below). Exhibit 7: The 10 year investment horizon dominates all other time periods on a total risk adjusted return basis 1.00

Sharpe Ratio

0.80

0.69

0.60 0.37

0.40 0.22

0.22

1Yr

3 Yr

0.23

0.20 0.00 5 Yr

7 Yr

10 Yr

Investment horizon Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016.

Additionally,  The 1 year and 3 year investment horizon stand out as the period offering the worst Sharpe ratio or total risk adjusted return.  The 5 year, 7 year and 10 year investment horizon entail a Sharpe ratio which is 1.1x, 1.7x and 3.2x times more respectively than that compared to a 1 year period. Since the Sharpe Ratio has been criticized for not capturing the risk adjusted returns appropriately when the underlying distribution is non-normal, we also analyze the “Sortino Ratio” for the different time horizons. The “Sortino Ratio” adjusts for risk by only factoring in the downside or negative volatility rather than the total volatility used in calculating the Sharpe Ratio. The Sortino Ratio takes the upside volatility out of consideration and uses only the downside standard deviation.

December 22, 2016

Ambit Capital Pvt. Ltd.

“Sortino Ratio” gives the downward risk adjusted return perspective

Page 5

Strategy An examination of the Sortino ratio of equity returns in India also suggests that the 10 year investment horizon dominates all other time periods with an even more material lead over the 5,7,3, and 1 year investment horizons (in descending order) (see exhibit below). Exhibit 8: The 10 year investment horizon outperforms even on a downside risk adjusted basis

Sortino Ratio

8

6.6

6 4 2

0.6

0.7

1Yr

3 Yr

1.9

1.8

5 Yr

7 Yr

0 10 Yr

Investment Horizon Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016.

Additionally,  The 1 year and 3 year investment horizon stand out as the period offering the worst Sortino ratio, or the downward risk adjusted return.  The 7 year, 5 year and 10 year investment horizon entail a Sortino ratio which is 3.1x, 3.2x and 11x times more respectively than that compared to a 1 year period.

Insight#4: The 10 and 7 year horizons offer the best probability of beating the risk free rate An examination of the probability of beating the risk free rate suggests that this probability is the highest for the 7 and 10 year investment horizons. In fact, the 10 year time horizon appears be the best bet for an investment manager who is keen to minimize the chances of not beating the risk free rate (see exhibit below).

The 10 year investment horizon offers the highest probability of beating the risk free rate

Exhibit 9: Probability of beating the risk free rate is the highest in 1,7, and 10 year periods Investment Horizon 1 year

Less than -22% 12%

Equity Returns Interval (BSE100, in %) -22% -7% 8% 23% to -7% to 8% to 23% to 38% 12% 22% 18% 10%

More than 38% 26%

Total no. of observations

Probability of returns exceeding 8%

297

54%

3 year 5 year

0% 0%

5% 0%

44% 46%

28% 40%

12% 10%

11% 4%

273 249

51% 54%

7 year 10 year

0% 0%

0% 0%

34% 24%

52% 76%

13% 0%

0% 0%

225 189

66% 76%

Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016.

Additionally,  The 1 year investment horizon offers the third highest probability of beating the risk free rate at 54% (i.e. 18%+10%+26%). However, this high probability entails 2the risk of a high downside as well as the probability of making equity returns of -7% or less stands at an elevated level of 24%.

December 22, 2016

Ambit Capital Pvt. Ltd.

Page 6

Strategy Exhibit 10: The 1 year horizon returns entail high downside as well as upside risks

No. of observations

60

The 1 year investment horizon typically tends to have high volatility of returns

40

20

0 -100%

-50%

0%

50%

100%

150%

200%

250%

BSE100 returns (in %) Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016. The red dotted lines represent the average, average +1standard deviation, and average -1standard deviation.



The probability of beating the risk free rate decreases as the investment horizon increases to 3 years (to 51%) and returns to 54% for the 5 year horizon (exhibits below).

Exhibit 11: The probability of beating the risk free rate is 51% for a 3 year investment horizon

Exhibit 12: The probability of beating the risk free rate is 54% for a 5 year investment horizon 150 No. of observations

No. of observations

150

100

50

0 -40%

0%

40%

100

50

0 -40%

80%

BSE100 returns (in %) Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016. The red dotted lines represent the average, average +1standard deviation, and average -1standard deviation.



40%

80%

Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016. The red dotted lines represent the average, average +1standard deviation, and average 1standard deviation.

Once the investment horizon increases to 7 years and beyond, the probability of beating the risk free rate picks up. For the 7 and 10 year investment periods, the same is 66% and 76% respectively. Interestingly, an investor is virtually guaranteed a return between 8%-23% if they are willing to remain invested for a 10 year long period. The probability of making a return less than -7% over both the 7 and 10 year investment horizons is 0% (exhibits below).

December 22, 2016

0%

BSE100 returns (in %)

Ambit Capital Pvt. Ltd.

Probability of beating the risk free rate picks up with a 7 year and above investment horizon

Page 7

Strategy Exhibit 13: The 10 year investment horizon is best placed to beat the risk free rate

No. of observations

Average:13% -1SD

+1SD 120

10 Yr horizon has a fat right tail

80

Average:17%

-1SD

+1SD

10 Yr investment horizon 1 Yr investment horizon

40 1 Yr horizon has a fat left tail

0 -50%

-30%

-10%

10%

30%

50%

70%

90%

110%

130%

150%

BSE100 returns (in %) Source: Bloomberg, Ambit Capital Research. Note: Period under consideration is from April 1991 – December 2016. The red dotted lines represent the average, average +1standard deviation, and average -1standard deviation for the 1 year investment horizon and the black dotted lines represent the average, average +1standard deviation, and average -1standard deviation for the 10 year investment horizon. The x axis is cut at 150% - note the maximum value is

December 22, 2016

Ambit Capital Pvt. Ltd.

Page 8

Strategy

India

OVERVIEW: BROAD AND MODERATE We expect the government to adopt a moderate taxation system over the next few years in line with its stated vision of lower tax rates and broader tax base. We believe moderate tax rates will help expand the tax base of income tax payers while the implementation of GST will expand the tax base of indirect tax payers. The recent demonetization exercise and potential unearthing of unaccounted income by income tax authorities may alter the apathy and antipathy of large sections of the society towards payment of taxes. A moderate taxation system is a fairly straightforward issue We believe the combination of (1) a moderate taxation system (‘carrot’) and (2) the government’s recent demonetization measure (‘stick’) will likely result in broadening of India’s paltry tax base. We expect the two measures to result in several positive changes from a taxation perspective. We note that there is reasonable clarity on the philosophy of the government with respect to direct and indirect taxes. The government has already stated the need for a wider tax base and moderate tax rates. We see several positive outcomes of a moderate taxation system.  Behavioral change among citizens and tax payers. In our view, reduction in corporate and individual income tax rates in the forthcoming union budget and commitment on behalf of the government to move towards moderate tax rates will motivate the right outcomes in the form of (1) better compliance with the tax laws of the country, (2) higher tax payment by companies and individuals and (3) elevated tax revenues for the government. Of course, the fear of government action against unreported income and unaccounted wealth may also force the right behavior. We believe this is an opportune time for the government to reduce income tax rates as the demonetization measure may have already forced a rethink among tax evaders, especially small entities in manufacturing and services in the informal sector and selfemployed professionals in the services sector. We note that individuals did suffer large losses (15-50%) in converting their old currency into new currency although most of the demonetized currency of ₹15.44 tn appears to have come back into the banking system. Lower tax rates may entice tax payers to switch to reporting their true incomes and payment of ‘full’ taxes. Concurrently, the fear of further government action against (1) accountholders whose bank accounts show a steep increase in deposits post the government’s demonetization measure, (2) accountholders whose bank accounts show a sharp increase in deposits disproportionate to their reported or declared income and (3) accountholders who may have ‘rented’ their bank accounts to others to facilitate deposits of demonetized currency may force (1) accountholders to disclose their unreported income under the ongoing income declaration scheme and (2) taxpayers to report their true incomes in the future. We note that a large section of the informal economy under-reports its revenues, taxable income and taxes. In our view, (1) general apathy towards taxation given the underappreciation of the importance of taxes in the economic and social development of the country, (2) sheer arrogance among tax evaders about their ability to avoid payment of taxes; the success of the government’s efforts in catching undisclosed income in bank accounts will be critical in this regard and (3) outright antipathy to taxes given limited visibility of the benefits from payment of taxes contribute to the non-compliant behavior. Of course, a section of society simply does not believe in paying taxes as a matter of right or belief stemming out of utter disregard of the law.

KOTAK INSTITUTIONAL EQUITIES RESEARCH

3

India

Strategy

 New ‘social contract’ between citizens and government. In our view, moderate tax rates can rejuvenate the ‘social contract’ between the citizens and the government. We believe that (1) a lack of trust between the government and taxpayers and (2) unrequited expectations are the key reasons for the paltry tax base in the country. The ‘social contract’ between the citizens and the government in terms of delivery of good governance by the government in exchange for taxes paid by citizens has weakened considerably over time. There seems to be a big gap between the expectations of taxpayers from the government and delivery of public services by the government to the taxpayers. Many taxpayers or potential taxpayers see the payment of taxes as a burden as they believe (rightly or wrongly) that the government has failed to deliver the desired level of social benefits to them in exchange for payment of taxes. The government sees payment of taxes as a means to provide basic services and meet certain social objectives. It is interesting to note that individual tax rates in India are significantly lower versus tax rates in the US and Western Europe and similar to tax rates in emerging market countries (see Exhibit 1 for individual tax rates in several developed and developing countries). Of course, citizens in developed countries receive high-quality public services (education, healthcare, infrastructure such as electricity, roads and water & sewage) for the high individual taxes paid by them. Indian citizens have to rely on private service providers for their needs if they can afford them or make do with poor-quality services provided by government agencies and utilities. Exhibit 1: Individual tax rates in India are lower than those in developed market countries but similar to emerging market countries Highest individual tax rate in developed and emerging economies, 2016 (%) 60

56.0 Individual tax rate (%)

49.0

50

45.0

45.0

45.0

45.0

45.0 39.6

40

38.0

35.5

35.0

35.0

32.0

30.0

30

28.0

27.5 22.0

20

15.0

13.0

10

Russia

Hong Kong

Singapore

Brazil

Malaysia

Indonesia

Philippines

Thailand

Mexico

India

Korea

US

UK

Taiwan

Germany

China

Australia

France

Japan

0

Notes: (a) For France, Korea, Russia and Thailand, data is for 2015. Source: KPMG, Kotak Institutional Equities

Thus, we would ascribe the general disgruntlement among Indian taxpayers (the ones with salaries in the organized sector) despite the relatively low individual tax rates in India versus other countries to (1) the poor quality of public services provided by governments (central, state, local) and (2) large evasion of taxes by large sections of the population in the informal economy, which puts the entire burden of public services on the narrow base of law-abiding taxpayers.

4

KOTAK INSTITUTIONAL EQUITIES RESEARCH

Strategy

India

As an aside, it is a moot point whether the quality of public services will improve in India simply from the government charging higher taxes. We are not even remotely suggesting that (far from it). In our view, public services need to be priced appropriately and taxation should not have much of a role in the pricing of public services, in our view. India requires significant administrative and financial reforms of its public service providers and public utilities. Of course, some amount of subsidies (direct payments to households below a certain level or universal basic income as is being debated currently in many countries) is appropriate and can be paid out from the government’s taxation revenues.  Greater economic vitality from lower corporate tax rates. We believe lower corporate tax rates will (1) improve cash flows of the larger companies who broadly pay full taxes (at the marginal corporate tax rate); this could result in higher investments over a period of time and (2) increase tax revenues of the government from companies who have hitherto under-reported their incomes. We believe the government can take a calculated ‘gamble’ at this stage as investment cycle is quite weak anyway and companies in the informal sector appear keen to make the transition to the formal (reported) economy. We note that the implementation of GST over the next few months will anyway make it difficult for companies to avoid detection of their true revenues and incomes. We note corporate tax rates in India are quite high relative to those in developed and developing countries (see Exhibit 2). More importantly, the taxation system has become very complex with a large number of exemptions, which reduces the effective tax rate to around 20-25% despite the marginal tax rate of 34.6%. The government already recognizes this and has already stated its intentions to reduce the corporate tax rate by 5% by FY2020 and remove exemptions gradually from April 1, 2017. We expect the government to probably reduce tax rates by 1.5-2.5% in the forthcoming union budget. We note that several tax exemptions for companies will cease from April 1, 2017 and more will cease over a period of time and it would be natural to reduce the marginal corporate tax rate simultaneously. Exhibit 2: Corporate tax rates in India are quite high relative to developed and emerging economies Corporate tax rate in developed and emerging economies, 2016 (%) 50 Corporate tax rate (%) 34.0 30.9

30.0

30.0

30.0

29.7

30

25.0

25.0

24.2

24.0 20.0

20.0

20.0

20

17.0

17.0

16.5

Hong Kong

34.4

Taiwan

34.6

Singapore

40

38.9

10

Thailand

Russia

UK

Malaysia

Korea

Indonesia

China

Germany

Philippines

Mexico

Australia

Japan

Brazil

France

India

US

0

Source: OECD, Deloitte, Tax foundation, Kotak Institutional Equities

KOTAK INSTITUTIONAL EQUITIES RESEARCH

5

Strategy

India

INDIRECT TAXES: FOR ONE AND ALL We believe the implementation of GST will likely broaden the tax base. It would be difficult for smaller entities in manufacturing and services, which under-report excise duties and service taxes currently, to stay out of the GST chain. The system of credits and debits under GST makes it incumbent on every entity in the GST chain to ensure payment of GST by the entities preceding them in the chain. GST should be a big improvement over the current indirect taxation system We expect GST to result in a significant increase in indirect tax revenues over time for two reasons.  Larger base of taxpayers through lower exemption limits. We note that the proposed GST system will reduce the minimum limit for exemption from payment of GST for manufacturing units to ₹2 mn of revenues (₹1 mn for north-east and hilly states) from the current ₹15 mn of revenues for payment of excise duty. This will result in inclusion of several smaller entities into the GST system. The government has already announced a schedule for registration of units under GST (see Exhibit 16 for details). Exhibit 16: The government has already announced schedule for registration of units under GST GST enrolment schedule for existing and new indirect tax payers Schedule Puducherry (68.5%) and Sikkim (46.1%)

Enrolment start date 08-Nov-16

Chhattisgarh (75.9%), Dadra Nagar Haveli (28.2%), Daman & Diu (49.3%), Goa (47.5%), Gujarat (79.2%), Maharashtra (62.1%)

14-Nov-16

Arunachal Pradesh (8%), Assam (3%), Bihar (49.9%), Jharkhand (57.8%), Madhya Pradesh (79.4%), Manipur (10.1%), Meghalaya (10.9%), Mizoram (29.3%), Nagaland (46.7%), Odisha (29.3%), Tripura (14.5%), West Bengal ( 60.9%)

30-Nov-16

Chandigarh (58%), Delhi (55.2%), Haryana (40.5%), Himachal Pradesh (43.4%), Jammu and Kashmir (0%), Punjab (37.6%), Rajasthan (72.6%), Uttarakhand (17.3%), Uttar Pradesh (47.2%),

16-Dec-16

Andhra Pradesh (53.5%), Karnataka (74.9%), Kerala (32.2%), Tamil Nadu (42.1%), Telangana (56.4%) Existing central excise and service tax registrants All other registrations

01-Jan-17 Jan 1-31, 2017 Feb 1-Mar 20, 2017

Notes: (a) Numbers in parenthesis indicate enrolment rate in the state.

Source: GST portal, Media reports, Kotak Institutional Equities

On the other hand, the proposed GST system will increase the minimum limit for exemption from payment of GST for services entities to ₹2 mn of revenues (₹1 mn for north-east and hilly states) from the current ₹1 mn of revenues for payment of service tax. However, we suspect that there is a lot of leakage of service tax revenues anyway noting that the proportion of service tax to overall tax revenues is quite small (see Exhibit 17), especially in the context of the large share of the services sector in India’s GDP.

KOTAK INSTITUTIONAL EQUITIES RESEARCH

17

India

Strategy

Exhibit 17: The share of service tax is very low to overall tax revenues Service tax to total tax, March fiscal year-ends, 2000-18E (%) Service Tax/Total Tax (%)

16

14.5 14.8

14

12.8

12

11.0 10.0

10 7.9 8

8.5

9.4 9.0

6.3

6

4.7 3.1

4 2

15.0

13.6 13.5

1.8 1.9 1.2 1.4

2018E

2017E

2016P

2015

2014

2013

2012

2011

2010

2009

2008

2007

2006

2005

2004

2003

2002

2001

2000

0

Source: CEIC, Kotak Institutional Equities

Of course, higher taxation rates on services under GST (we assume the standard rate of 18% will apply on most services although certain essential services may attract a lower rate of 12%) versus the current rate of 15% will result in a significant increase in ‘service’ tax under GST although the distinction will no longer apply under GST.  Larger base of taxpayers from greater difficulty in tax avoidance. We note that the implementation of GST will make it significantly difficult for entities in manufacturing and services to under-report their revenues. The system of GST makes it incumbent on entities in the GST chain to ensure payment of taxes by the preceding entities in the GST chain. Under the GST system, a ‘successor’ entity in the GST chain will not be able to claim (offset) the taxes paid by it on procurement of goods and services from a ‘predecessor’ entity in the chain against goods and services sold by it (take a credit of taxes on inputs, in other words) in turn unless the ‘predecessor’ entity has made the payment in the first place. Exhibits 18 and 19 are hypothetical examples of the working of the GST chain and illustrate the system of tax credits and debits in the GST chain.

18

KOTAK INSTITUTIONAL EQUITIES RESEARCH

Strategy

India

Exhibit 18: Dual GST within state: A working example

SGST Paid = Rs10

State

SGST Paid = Rs10 (Rs20 - Rs 10 Input Tax Credit)

Timber

SGST @10%

Furniture A

CGST Paid = Rs10

SGST Paid = Rs10 (Rs30 - Rs 20 Input Tax Credit)

Furniture B

CGST Paid = Rs10 (Rs20 - Rs 10 Input Tax Credit)

Final C

CGST Paid = Rs10 (Rs30 - Rs 20 Input Tax Credit)

Centre CGST Tax Invoice A

Tax Invoice B

Tax Invoice C

Cost of Goods = Rs100 SGST @ 10% = Rs 10 SGST @ 10% = Rs 10

Cost of Goods = Rs200 SGST @ 10% = Rs 20 SGST @ 10% = Rs 20

Cost of Goods = Rs300 SGST @ 10% = Rs 30 SGST @ 10% = Rs 30 Received Size

Source: Ministry of Finance, Kotak Institutional Equities

KOTAK INSTITUTIONAL EQUITIES RESEARCH

19

India

Strategy

Exhibit 19: IGST model: A working example

State X

State Y State Border

State X SGST @10% SGST Paid = Rs10

Timber

Tax Invoice B

Cost of Goods = Rs200 IGST @ 20% = Rs 40 -----------------------------

Furniture A

Furniture B

Final C

Payment of IGST IGST payable = Rs40 Less CGST ITC = Rs10 Less SGST (G) ITC = Rs10 -------------------------------

CGST Paid = Rs10

State Y SGST @10%

Payment of SGST SGST (M) payable = Rs30 Less IGST ITC = Rs10 -------------------------------

Tax Invoice A

Payment of CGST CGST payable = Rs 30 Less CGST ITC = Rs 0.0 Less IGST ITC = Rs 30 -----------------------------------

Tax Invoice C Centre CGST @10%

Cost of Goods = Rs100 SGST @ 10% = Rs 10 CGST @ 10% = Rs 10

Cost of Goods = Rs300 SGST (M) @ 10% = Rs 30 CGST @ 10% = Rs 30

Source: Ministry of Finance, Kotak Institutional Equities

We believe a large number of smaller entities in manufacturing and services under-report their production and revenues and thus, under-report their indirect taxes apart from direct taxes such as corporate income tax.  Manufacturing. This is particularly true for small-scale manufacturing units who underreport their production and sales volumes and sell a meaningful portion of their produce for cash and purchase the corresponding portion of raw materials in cash. In fact, it is not uncommon to see entities under-reporting their production to the extent of 50-75% in order to save on excise duties and also, on income taxes. We expect a significant increase in taxation revenues under either of the scenarios of (1) successful migration of a significant portion of the manufacturing entities in the unorganized (unreported) sector to the organized (reported) sector or (2) significant loss of market share by manufacturing units in the unorganized sector to larger units in the organized sector. We note that the share of the unorganized sector in several sectors in the broader manufacturing sector is quite large (see Exhibit 20). Thus it would be natural to expect a significant increase in tax revenues as and when the unorganized sector players become part of the formal economy and start reporting their true revenues and incomes. It is also possible that many of the smaller manufacturing units will fail to compete with their larger counterparts in the organized sector once they provide for higher taxes as part of their costs. They would lose market share to the organized sector players who anyway pay full taxes and will pay more taxes on their higher production volumes/revenues (arising from market share gains from unorganized sector players). 20

KOTAK INSTITUTIONAL EQUITIES RESEARCH

Strategy

India

Exhibit 20: Unorganized sector's share is quite high in several sectors Sectors which have high share of unorganized players

Category Auto ancillaries Aftermarket - batteries

Unorganized share (%) ~40

Consumer discretionary Adhesives & sealants Ceramic/tiles Footwear Paints

~30-35 ~50 ~60-70 ~35

Plywood

~60-70

QSRs

~60-70

Sanitaryware

~50

Consumer staples Biscuits

~35-40

Coconut hair oils

~35

Detergents

~30

Edbile oils

~40-45

Hair colours

~30-35

Other hair oils

~30

Packaged foods

~50

Tea

~40

Hospitals Healthcare

~70

Others Appliances

~30-40

Cable industry

~50

Fans

~20

Masterbatches

~75

Textiles

~50

Source: Kotak Institutional Equities estimates

 Services. We believe a large number of business entities in various services and selfemployed professionals in the services sectors likely under-report their revenues and underpay service taxes other than individual income taxes. It would be interesting to see if they will continue to under-report their revenues post the experience of demonetization and implementation of GST. As discussed in the overview section, (1) effective detection of unexplained deposits in bank accounts by the Income Tax department post the recent demonetization and (2) imposition of strict penalties on individuals indulging in such fraudulent practices will go a long way in changing the economic behavior of businesses and households who have had neither fear of the law nor respect for the law and have blithely under-reported their true incomes.

KOTAK INSTITUTIONAL EQUITIES RESEARCH

21

India

Strategy

CAPITAL GAINS: HIGHER FOR ALL? We do not rule out the possibility of government reviewing its taxation policy on capital gains for equities. The government changed the taxation policy for debt by increasing the holding period for debt in the FY2015 union budget. It has also changed tax agreements with countries to tax investments from low-tax jurisdictions. The increasingly stringent treatment of capital gains tax in general makes us wary of potential negative developments in the forthcoming union budget. Philosophy of taxation on capital gains The philosophy of taxation on capital gains largely depends on the strategic objectives of a country and its government. Most countries have a lower rate of taxation on capital gains versus the taxation rate on income (corporate or individual) as can be seen in Exhibit 21. Governments have presumably accorded higher value (in a manner of speaking) to income from capital (investment) versus income from labor (income) for the governments to have lower taxes on capital versus on labor. Governments perhaps want to rewards investors for the risk they take in creating businesses and generating employment. Exhibit 21: Most countries have a lower rate of taxation on capital gains versus the personal or corporate tax Capital gain tax rate across economies Capital gains tax (%)

Comment

Varies; highest is 14.5%

Individual's tax rate on 50% of realized capital gains

Developed economies Canada France Germany Hong Kong Japan Singapore

19% (+15.5% + (2% to 6%))

In addition to 19% capital gains tax, 15.5% social charges is applied. Supplementary tax of 2% to 6% is also applied

25% 0% 20%/39% 0%

UK

10%/20%

US

0%/15%/20%/28%

Under Hong Kong tax law, "profits arising from the sale of capital assets" are outside the scope of charge for profits tax Long-term: 15% income tax + 5% inhabitant tax; short term: 30%+9% There is no capital gains tax in Singapore Tax rate depends on the size of gain. Additional 8% if the asset is a residential property Depending on tax bracket, type of asset. Short-term capital gains are taxed as ordinary income

Emerging economies Brazil

15%/17.5%/20%/22.5%

China

20%

India Indonesia Korea Malaysia Mexico Philippines

0%/20% 25%/20% 20% 0%/5%/15%/20%/30% 10%/35% 5%/6%/10%

Depending on the gain size 20% for individuals; no distinction of capital gains for Chinese companies; 10% for foreign company Long-term tax: 0% for listed shares, 20% for other assets. Short-term: 15% (+surcharge and cess) Long-term capital gains are taxed as ordinary income, and losses are tax deductible Minimum of 10% of sales price/ 20% of gains Only gains from selling property or shares in property company is taxed depending on the holding period For residents, capital gains are considered normal income. For foreigners, 10% for shares and 35% on net proceeds of other assets 6% for real property; 5% and 10% for shares depending on the value

Russia

0%/13%

For residents, 0% if held more than 5 years, else 13%. 30% for foreigners

Taiwan

0%/35%

0% for gains from shares of listed companies and mutual funds, others are treated as ordinary income

Thailand

0% - 35%

Capital gains are exempted if the shares sold are of a public traded company; otherwise, gains are subject to progressive personal tax rates

Source: Government websites, Kotak Institutional Equities

We are not getting into the debate about the need for capital gains tax in the case of companies as capital gains can be construed as an additional tax (double taxation) on shareholders or owners of capital—the company pays tax on its income (corporate tax) and shareholders pay tax on capital gains as and when they sell the shares (at a higher price). We put forward a few thoughts on the different treatment of tax on capital versus tax on labor.  Should capital or labor be taxed differently or similarly? It is possible that governments increase taxes on capital as a political and social response to rising inequality. Governments may use higher tax on capital as a way to increase expenditure on social welfare for the underclass in society. Thus, we would not rule out further government intervention in the form of ‘distribution’ politics and policies.

22

KOTAK INSTITUTIONAL EQUITIES RESEARCH

Strategy

India

It appears that the holders of capital have benefited disproportionately from the established economic system among the three major economic participants in capitalism—capitalists (through profits), government (through taxes) and labor (through wages). The share of shareholders (or holders of capital) of the value of economic output in a country will presumably rise further in the future as greater automation and mechanization will likely reduce the demand for and wages of certain types of labor in particular and the prices (or wages) of labor in general. We see some historical merit in the policies of governments to tax capital at concessional rates compared to labor. The economic argument for preferential treatment of capital perhaps reflects the fact that (1) there is ‘higher’ risk associated with investment of capital versus deployment of labor and (2) governments want investment, which in turn will generate employment. However, this argument was perhaps more valid when property rights were weak and the risk of failure of investments quite high. With the establishment of (1) the ‘rule of law’ in most countries and (2) full-fledged property rights, we are not sure the historical argument still holds. Also, we are not sure whether this argument holds at a philosophical level and whether governments should be treating various inputs of production differently. Such an argument will perhaps not stand the test of principle of equity since risk needs to be viewed in the context of returns (on a pre-tax basis and not on a post-tax basis). The basic purpose of taxation by governments is to simply provide certain services to all citizens and all sections of society and not as a tool to encourage or discourage different forms of economic activity. In fact, the different taxation rates on capital and labor may have also contributed to the rising inequality globally, which has become a big political and social issue in many countries.  Should there be tax on capital gains if the capital gains are reinvested in the ‘same’ or ‘similar’ assets? If the essential argument for lower tax rates on capital versus labor is reward for greater risk, we believe the argument can be extended to ‘reinvestment’ of capital gains in the ‘same’ or ‘similar’ asset. For example, reinvestment of capital gains on one stock by an investor in another stock is tantamount to ‘new’ risk being taken by the investor. Thus, irrespective of the holding period of the asset, it would be logical not to tax the capital gains on the original investment if the capital gains on the original investment are invested in the ‘same’ asset.  Should all investors in the same asset be taxed the same way? Extending the argument about lower taxation rates on capital gains being a reward for taking risk, we raise the question whether all investors in an asset face the same risk at any given point in time. It is obvious that all investors are subject to the same risk at any given point in time. However, we wonder whether the benefits of preferential treatment of capital gains should be restricted to the original investors who perhaps took the greatest amount of risk at the time of starting a company. Such a taxation policy would retain the logic of providing preferential tax treatment for risk-taking while treating income from investment and labor equally. In other words, any capital gains by passive investors (non-founder investors) can be treated as normal income for investors and taxed appropriately.  Should capital gains on equity and debt be taxed the same way or differently? We see no economic or financial logic for different tax treatment for income or gains on debt and equity. We note that both the (1) short-term and long-term capital gains tax rates and (2) the holding period for classification of capital gains as short-term and long-term for debt and equity in general and debt and equity mutual funds in particular are different in India. Exhibit 22 shows the short-term and long-term capital gains on debt and equity and debt and equity mutual funds and the holding period for classification as short-term and long-term capital gains.

KOTAK INSTITUTIONAL EQUITIES RESEARCH

23

7KH+RXVH9LHZ2XWORRN  7KHJOREDOHFRQRP\DQGPDUNHWVHQWHURQFRQVLGHUDEO\ILUPHUIRRWLQJWKDQODVW 7DEOHRIFRQWHQWV ƒ 2YHUYLHZ \HDU*URZWKPRPHQWXPKDVSLFNHGXSLQUHFHQWPRQWKVDQGULVNDVVHWVSDUWLFXODUO\ 2XWORRN  LQGHYHORSHGPDUNHWVKDYHFRQWLQXHGWKHUDOO\VSDUNHGE\7UXPS¶VVXUSULVHYLFWRU\ ƒ LQKHDGOLQHV 7UXPS¶VHOHFWLRQZLOOIXQGDPHQWDOO\UHRUGHUWKHHFRQRPLFILQDQFLDODQGVHFXULW\ $5HYLHZRI ƒ LQFKDUWV DUUDQJHPHQWVRIWKHSRVW::HUD:HH[SHFWWKHQHZDGPLQLVWUDWLRQWRUHPDLQWUXH  ƒ WRWDOUHWXUQV WRDQ$PHULFD)LUVWDSSURDFKUHHYDOXDWLQJH[LVWLQJDUUDQJHPHQWVXQGHUWKHOHQV ³:KDW VLQLWIRUWKH86"´:KLOHWKHUHLVFRQFHUQDERXWDSURWHFWLRQLVWWXUQE\WKH86 7KHPHVIRU ƒƒ 7KHVWDUWRIQRUPDOLVDWLRQ 3ROLF\XQFHUWDLQW\ ZHGRQRWH[SHFWDGLVUXSWLRQRIJOREDOWUDGH ZHGRQRWH[SHFWDGLVUXSWLRQRIJOREDOWUDGH  ƒ 'HJOREDOLVDWLRQ ƒ *HRSROLWLFV ,QVWHDG7UXPS¶VHFRQRPLFSODQVKRXOGUHEDODQFHWKHSROLF\PL[DQGLVDJDPH ƒ *OREDORYHUYLHZ FKDQJHUIRUWKH867KLVPDNHVXVYHU\EXOOLVK86JURZWK)LVFDOVWLPXOXVDQGEURDG 0DFUR ƒ 86RXWORRN EDVHGGHUHJXODWLRQDUHH[SHFWHGWRMROWWKH86HFRQRP\WRZDUGDORQJWHUP RXWORRNIRU ƒ (XURSHRXWORRN HTXLOLEULXPRIKLJKHUJURZWKLQIODWLRQDQGLQWHUHVWUDWHV  ƒ &KLQDDQG(0RXWORRN $IDVWHUJURZLQJ86LVSRVLWLYHIRUWKHJOREDOHFRQRP\EXWWKHLPSDFWZLOOEHOLPLWHG $IDVWHUJURZLQJ86LVSRVLWLYHIRUWKH JOREDOHFRQRP\EXWWKHLPSDFWZLOOEHOLPLWHG &HQWUDO ƒ *OREDORYHUYLHZ EHIRUHDQGZHH[SHFWRQO\DPLOGSLFNXSIURP¶VSRVWFULVLVWURXJK EHIRUH DQGZHH[SHFW RQO\DPLOGSLFNXSIURP¶VSRVWFULVLVWURXJK EDQNVLQ ƒ )HG 3ROLWLFDOXQFHUWDLQW\DQGDVRIWHUFUHGLWLPSXOVHDUHOLNHO\WRZHLJKRQ(XURSHDQ 3ROLWLFDOXQFHUWDLQW\DQGDVRIWHUFUHGLWLPSXOVH DUHOLNHO\WRZHLJKRQ(XURSHDQ  ƒ (&% JURZWK,Q&KLQDJURZWKLVH[SHFWHGWRVORZVOLJKWO\(0JURZWKSURVSHFWVDUH ƒ 6XPPDU\RIPDUNHWYLHZV 0DUNHWVLQ EULJKWHUSDUWLFXODUO\LQ%UD]LODQG5XVVLDWKRXJKZHDUHFDXWLRXVQHDUWHUP EULJKWHUSDUWLFXODUO\LQ%UD]LODQG5XVVLDWKRXJKZHDUHFDXWLRXVQHDUWHUP ƒ (TXLW\UDWHV);FUHGLW  DQG(0YLHZV 7KHQRUPDOLVDWLRQRILQIODWLRQDQGSROLF\LVDWODVWFRPLQJFOHDUO\LQWKH86JUDGXDOO\ '% ƒ 'HXWVFKH%DQNIRUHFDVWV HOVHZKHUH:KLOHPRQHWDU\SROLF\ZLOOUHPDLQHDV\LWLVVKLIWLQJWKH)HGZLOOUDLVH :KLOHPRQHWDU\SROLF\ZLOOUHPDLQHDV\LWLVVKLIWLQJWKH)HGZLOOUDLVH IRUHFDVWV ƒ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¶V*OREDO6WUDWHJ\*URXSZKLFKDGYLVHVPDQDJHPHQWDQG FOLHQWVRQEURDGPDUNHWULVNVDQGJOREDOHFRQRPLFDQGILQDQFLDOGHYHORSPHQWV7KHYLHZVDQGIRUHFDVWVRIWKHJURXS ZKLFKFRQVLVWVRIVHQLRUUHVHDUFKVWDIIPD\RFFDVLRQDOO\GLIIHUIURPWKRVHGLVVHPLQDWHGE\WKHLUUHVHDUFKFROOHDJXHV 'HXWVFKH%DQN 5HVHDUFK

WKHKRXVHYLHZ#OLVWGEFRP KWWSKRXVHYLHZUHVHDUFKGEFRP 7KH+RXVH9LHZ±-DQXDU\

(GLWRUV0DUFRV$UDQD$GLW\D%KDYH 0DWWKHZ/X]]HWWL5DMQL7KDNXU 

*OREDOJURZWKLVH[SHFWHGWRSLFNXSLQGULYHQE\WKH86 DQGDOHVVQHJDWLYHRXWORRNLQ&((0($DQG/DW$P *OREDOJURZWKRXWORRN 86YHU\EXOOLVKJURZWKRXWORRN ƒ 7UXPS¶VSODQDJDPHFKDQJHU ƒ 7D[FXWVLQIUDVWUXFWXUHVSHQG [FXWVLQIUDVWUXFWXUHVSHQG GHUHJXODWLRQYHU\SRVLWLYH GHUHJXODWLRQYHU\SRVLWLYH ƒ 3HDNLPSDFWLQ++  

8.%UH[LWVKRFNWRPDWHULDOLVH ƒ 3RXQGZHDNQHVVGRHVOLWWOHIRU H[SRUWVEXWKXUWVKRXVHKROGV H[SRUWVEXWKXUWVKRXVHKROGV 3ROLF\XQFHUWDLQW\WRZHLJK ƒ 3ROLF\XQFHUWDLQW\WRZHLJK 6FRSHIRUSROLF\HDVLQJOLPLWHG ƒ 6FRSHIRUSROLF\HDVLQJOLPLWHG JLYHQLQIODWLRQGHEWG\QDPLFV JLYHQ JLYHQLQIODWLRQGHEWG\QDPLFV

(XUR]RQHFDXWLRXVRXWORRN ƒ 'HVSLWHVWURQJPRPHQWXPVHH VORZGRZQLQ+ ƒ 3ROLWLFDOXQFHUWDLQW\VRIWHUFUHGLW LPSXOVHWRZHLJKRQJURZWK LPSXOVHWRZHLJKRQJURZWK ƒ 0DWHULDOSRVLWLYHJURZWK 0DWHULDOSRVLWLYHJURZWKVSLOORYHU VSLOORYHU IURP86XQOLNHO\EHIRUH IURP IUR IURP86XQOLNHO\EHIRUH

ƒ ,PSURYHG\HWQRWRYHUO\ H[FLWLQJJOREDOJURZWKRXWORRN í 7XUQHGPRUHSRVLWLYHDIWHU 7UXPS¶VHOHFWLRQRQ H[SHFWDWLRQRIIDVWHU H[SHFWDWLRQRIIDVWHU86 ƒ *URZWKPRPHQWXPDWLWV VWURQJHVWLQWKHODVW\HDUV DFURVVDOOPDMRUHFRQRPLHV

$IWHUUHFHQWSLFNXSVWDUWVZLWK VWURQJHVWJURZWKPRPHQWXPLQ\HDUV  &RPSRVLWH30, PPD 

86

 (0WXUQLQJFDXWLRXVO\SRVLWLYH ƒ )DVWHUJOREDOJURZWKKLJKHU FRPPRGLWLHVDUHSRVLWLYH ƒ %XWVWURQJGROODUULVLQJ86UDWHV %XWVWURQJGROODUULVLQJ86UDWHV WRZHLJKVKRUWWHUP WRZHLJKVKRUWWHUP 'HJOREDOLVDWLRQSROLWLFDOQRLVH ƒ 'HJOREDOLVDWLRQSROLWLFDOQRLVH SUHVHQWGRZQVLGHULVNV SUHVHQWGRZQVLGHULVNV 'HXWVFKH%DQN 5HVHDUFK

&KLQDPDQDJHGVORZGRZQ &KLQDPDQDJHGVORZGRZQ *URZWKWRVORZEXWUHPDLQ ƒ *URZWK ƒ ([SHFWSROLF\HDVLQJWRKLW JURZWKWDUJHWUDWKHUWKDQWLJKWHQ WRFRQWDLQKRXVLQJEXEEOH ƒ 1RVKDUSVORZGRZQEXW); FUHGLWFRQFHUQVWRFRQWLQXH

WKHKRXVHYLHZ#OLVWGEFRP KWWSKRXVHYLHZUHVHDUFKGEFRP 7KH+RXVH9LHZ±-DQXDU\

-DSDQUHFRYHU\WRFRQWLQXH ƒ (FRQRPLFUHFRYHU\WRFRQWLQXH LQDQGEH\RQG±PRVWO\ GULYHQE\SULYDWHFRQVXPSWLRQ



(XUR]RQH (0

      

1RWH,60IRUWKH86 6RXUFH%ORRPEHUJ)LQDQFH/3'HXWVFKH%DQN5HVHDUFK

UHDO*'3JURZWK\R\       !

QD



:HDUHYHU\EXOOLVKRQ86PDFURDVWKHQHZDGPLQLVWUDWLRQ SXUVXHVDVXEVWDQWLDOSURJURZWKILVFDODQGUHJXODWRU\UHYDPS SXUVXHVDVXEVWDQWLDOSURJURZWKILVFDODQGUHJXODWRU\UHYDPS ƒ 7UXPSHOHFWLRQDJDPHFKDQJHIRUWKH86RXWORRN í 6WURQJEHOLHIWKDW7UXPSZLOOSXW86JURZWKILUVW í &DPSDLJQSURPLVHVWKDWJRDJDLQVWWKLVOLNHO\WR EHGHHPSKDVLVHG í 8QLILHG5HSXEOLFDQJRYHUQPHQWLQFUHDVHV FKDQFHVRISROLF\LPSOHPHQWDWLRQ ƒ 3URJURZWKSROLF\DJHQGD í 6WURQJGHUHJXODWLRQELDVHVSHFLDOO\LQHQHUJ\ ILQDQFHHQYLURQPHQWKHDOWKFDUHODERXUZHOIDUH í 3URJURZWKWD[UHIRUPLQFOXGLQJFRUSRUDWHDQG LQGLYLGXDOWD[FXWV í ,QIUDVWUXFWXUHVSHQGLQJSODQ í 6\PEROLFEXWQRWPDWHULDOSURWHFWLRQLVWPHDVXUHV ƒ 7KLVSROLF\PL[KDVWKHSRWHQWLDORIUHLJQLWLQJ SURGXFWLYLW\JURZWKDQGUDLVLQJ86JURZWKSRWHQWLDO KLJKHUXQFHUWDLQW\WKLVLV ƒ :KLOH7UXPSLQWURGXFHV :KLOH7UXPSLQWURGXFHVKLJKHUXQFHUWDLQW\WKLVLV EHWWHUWKDQWKHQHDUFHUWDLQW\RIWKHFRQWLQXDWLRQRI EHWWHUWKDQWKH QHDUFHUWDLQW\RIWKHFRQWLQXDWLRQ FHUWDLQW\RIWKHFRQWLQXDWLRQRI DPHGLRFUHVWDWXVTXR DPHGLRFUHVWDWXVTXR í 6HFXODUVWDJQDWLRQILQDQFLDOUHSUHVVLRQHYHU WLJKWHUUHJXODWLRQIDGLQJHFRQRPLFG\QDPLVP 'HXWVFKH%DQN 5HVHDUFK

WKHKRXVHYLHZ#OLVWGEFRP KWWSKRXVHYLHZUHVHDUFKGEFRP 7KH+RXVH9LHZ±-DQXDU\

2XUNH\DVVXPSWLRQVUHJDUGLQJ7UXPS¶VDGPLQLVWUDWLRQ $VVXPSWLRQ ,PSOLFDWLRQ ƒ )DVWHU86HFRQRPLF ƒ 3ROLF\LPSOHPHQWDWLRQ JURZWKLVILUVWSULRULW\ VNHZHGWRZDUGPRUH 86JURZWK ƒ &RPSURPLVHVWUDGHRIIV JURZWKIULHQGO\OHVV FRPHVILUVW YLHZHGYLDWKLVOHQV JURZWKXQIULHQGO\ ƒ 1RWLQKHUHQWO\DQWLWUDGH SROLFLHV ƒ 3UDJPDWLVPRYHU ƒ &KDOOHQJHDOOH[LVWLQJ %XVLQHVV SULQFLSOHVDQGGRJPD SULQFLSOHVDQGGRJPD UHODWLRQVLQFOIRUHLJQ OLNH ƒ %XVLQHVVOLNHDSSURDFK %XVLQHVVOLNHDSSURDFK ƒ +LJKHUXQFHUWDLQW\ DSSURDFK WRUXQQLQJJRYHUQPHQW XQSUHGLFWDELOLW\ 7UXPS¶VSURSRVHGFRUSRUDWHWD[FXWZRXOGWDNHWKH86IURPWKHOHDVW WRDPRQJWKHPRVWDWWUDFWLYHGHYHORSHGFRXQWULHV 86 )UDQFH ,WDO\ *HUPDQ\ -DSDQ &DQDGD 6.RUHD 6ZLW]HUODQG 8. 86

+HDGOLQH

2WKHU  7UXPS¶VSURSRVDO 









1RWH 2(&'FDOFXODWHVDFRPELQHGFRUSRUDWHWD[UDWHWRHQVXUHDSSOHVWRDSSOHVFRPSDULVRQV7UXPS¶V SURSRVDODVVXPHVKHDGOLQHFRUSRUDWHLQFRPHWD[UDWHRI6RXUFH2(&''HXWVFKH%DQN5HVHDUFK



:HH[SHFWJURZWKWRULVHWRZLWKSHDNLPSDFWLQ+ DQG+HVSHFLDOO\RQKLJKHUEXVLQHVVLQYHVWPHQW ƒ :HDUHPRUHEXOOLVKWKDQFRQVHQVXVRQ86JURZWK í *'3IRUHFDVWVUHYLVHGXSWRE\HQG LPSDFWH[SHFWHGLQ+DQG+ í 3HDN 3HDNLPSDFWH[SHFWHGLQ+DQG+ ƒ 0DLQGULYHURIJURZWKWREHEXVLQHVVVSHQGLQJ YHU\ZHDNWKLVF\FOHVLJQLILFDQWSHQWXS í &DSH[ &DSH[YHU\ZHDNWKLVF\FOHVLJQLILFDQWSHQWXS LQYHVWPHQWVSHQGLQJ LQYHVWPHQWVSHQGLQJ í ,QFUHDVHLQFDSLWDOVWRFNKDVSRWHQWLDORI UHYHUVLQJGRZQWUHQGLQSURGXFWLYLW\JURZWK ƒ 3RVLWLYHEXWPRUHPRGHVWLPSDFWIURPFRQVXPSWLRQ 3RVLWLYHEXWPRUHPRGHVWLPSDFWIURPFRQVXPSWLRQ í &RQVXPHUGHPDQGWKHPDLQHQJLQHIRUJURZWKLQ &RQVXPHUGHPDQGWKHPDLQHQJLQHIRUJURZWKLQ UHFHQW\HDUVVROHVVSHQWXSGHPDQG UHFHQW\HDUVVROHVVSHQWXSGHPDQG í %XWFRPSDULVRQ WRSHULRGDIWHU%XVKWD[FXWV %XWFRPSDULVRQWRSHULRGDIWHU%XVKWD[FXWV VXJJHVWVXSVLGHWRRXUFRQVXPSWLRQIRUHFDVWV VXJJHVWVXSVLGHWRRXUFRQVXPSWLRQIRUHFDVWV ƒ &RQILGHQFHERRVWWRSOD\DQLPSRUWDQWUROH ƒ 'HVSLWHDOOWKHDWWHQWLRQLQIUDVWUXFWXUHVSHQGLQJQRW DPDMRUGULYHUIRUJURZWK SRVVLEO\ ƒ 8QHPSOR\PHQWVKRXOGFRQWLQXHIDOOLQJ 8QHPSOR\PHQWVKRXOGFRQWLQXHIDOOLQJSRVVLEO\ GURSSLQJWRE\IURP  QRZ GURSSLQJWRE\IURPQRZ  ƒ 7KHUHDUHULVNVWRWKLVRXWORRN í )LUVWDQGIRUHPRVWLI7UXPSSULRULWLVHV )LUVWDQGIRUHPRVWLI7UXPSSULRULWLVHV SURWHFWLRQLVPRYHU86JURZWK SURWHFWLRQLVPRYHU86JURZWK í ([FHVVLYHGROODUVWUHQJWKLI)HGKDVWRKLNHIDVW 'HXWVFKH%DQN 5HVHDUFK

WKHKRXVHYLHZ#OLVWGEFRP KWWSKRXVHYLHZUHVHDUFKGEFRP 7KH+RXVH9LHZ±-DQXDU\

%XVLQHVVLQYHVWPHQWLVH[SHFWHGWREHWKHPDLQGULYHURIIDVWHU*'3 JURZWKLQWKHPHGLXPWHUP 





    

4 4 4 4 4 4 4 4 4 *RYWH[SHQGLWXUH 1HW([SRUWV 1RQUHVLGHQWLDO,QYHVWPHQW

2WKHUIL[HGLQYHVWPHQW 3YWFRQVXPSWLRQ 5HDO*'3

6RXUFH+DYHU$QDO\WLFV'HXWVFKH%DQN5HVHDUFK

2XUIRUHFDVWVRIUHDO3&(JURZWKDUHPRGHVWUHODWLYHWRWKH DFFHOHUDWLRQWKDWRFFXUUHGDIWHUWKH%XVKWD[FXWV 

TRT$5

 



DFWXDO 

IRUHFDVW