Evaluate Forecasting Excellence The Value of Outliers in ... ›...

Evaluate Forecasting Excellence

The Value of Outliers in Consensus ForecastingA perspective on the portfolio strategy of the top 20 pharma companies

Executive Summary

How much consensus is there within consensus? When we describe

the future outlook of the pharmaceutical industry by leveraging

consensus estimates we rely upon the wisdom of the crowd (sell-

side analysts) but we are not questioning the level of agreement –

or disagreement– implicit within each consensus estimate.

In this analysis we have looked beyond consensus, focusing on the outliers within each distribution as they can

provide an insightful view of which companies have more upside potential and downside risk.

Drug companies forecast to have solid growth, at or above market average, such as Bristol-Myers Squibb and Roche,

exhibit portfolio strategies infused with high levels of risk which may either pay off in the long run, or could be heavily

affected by pipeline setbacks or larger than expected biosimilar impact. We have also flagged companies like

GlaxoSmithKline and Celgene which are poised for growth based on a high degree of certainty, with potential upside.

We ultimately believe that combining consensus forecasts with the analysis of outliers will enable companies

investing in pharmaceutical assets to:

1. better value companies and portfolios by balancing risk and upside,

2. drive innovation in the most promising areas through improved accuracy of pipeline decision-making.

While the use of outliers enriches consensus forecasts as a powerful toolkit for market and portfolio analysis,

additional efforts are required to provide a more comprehensive value of the full pharmaceutical pipeline. With this

in place better decisions for earlier stages of development may be possible, shifting decision making upstream for

improved portfolio management.

2 Copyright © 2017 Evaluate Ltd. All rights reserved.

EVALUATE FORECASTING EXCELLENCE

The Value of Outliers in Consensus Forecasting

Antonio IervolinoHead of ForecastingEvaluate Ltd

[email protected]

+44 7773 237412

Introduction

Several empirical studies1,2 have shown that pooling forecasts

increases accuracy, particularly in situations characterised by a

high degree of uncertainty. While some criticism has been recently

raised around consensus forecasts, they do represent an impartial

benchmark which provides a single, timely, quality forecast, reflecting

complex news flow and market events.

Research to date has not focused on the additional insight provided by the upper and lower bounds of consensus

known as outliers – these may often identify potential upside or downside events that could significantly affect the

outlook of a pharmaceutical company. Such a topic seems particularly relevant at the end of 2016 where two major

unexpected political events, Brexit and the outcome of the US presidential elections, have impacted the political and

macroeconomic environments.

In this paper, we have analysed the portfolio strategy of the top 20 pharmaceutical companies (as of November 2016)

by leveraging the maximum and minimum forecast within EvaluatePharma® consensus. The results have given us a

thought-provoking perspective on how the forces at play in the market and the balance among companies could shift

significantly by looking at how risk and upside potential are spread across a company portfolio.

We seek to provide value through this analysis and help trigger a wider discussion on improving forecasting accuracy to

drive more robust decision-making, allocate investment more efficiently and ultimately drive continuous innovation.

1 Clemen, Robert T. (1989). “Combining forecasts: A review and annotated bibliography”. International Journal of Forecasting.2 McNees, Stephen K. (November 1987). “Consensus Forecasts: Tyranny of the Majority”. New England Economic Review.





A portfolio strategy review of Top 20 Pharma: A story of two tales

When we published our annual EvaluatePharma® World Preview report in September 2016,

we showed a positive outlook through 2022 where scientific innovation combined with more

targeted R&D and commercial strategies were likely to lead to a period of sustained growth

(6.1% CAGR3 2015-224).

A few months later the scenario has not materially changed. If we look at the Top 20 pharmaceutical companies5,

EvaluatePharma® aggregated consensus forecasts predict a CAGR of 4.1% through 20226 (Figure 1). Within this

group, companies such as Shire, Celgene and BMS stand out where expected growth far exceeds the average of

this cohort. On the lower end of the spectrum, Gilead’s long-term revenue outlook is heavily affected by the rapid

maturing of its Hepatitis C franchise.

Figure 1 – Top 20 pharmaceutical companies, 2022 Sales ranking


10

20

30

40

60

50

Source: EvaluatePharma® November 20162022 Ranking, Top 20 Pharma – Total Company Sales

Rev

enue

s, $

Bn

20

15-2

2 C

AG

R

10%

15%

20%

0%

5%

-5%0

Change in Ranking 2022 vs. 2015

PFZ

- -

JJ

21 2 1

GSK

1

ABV

2

AZ

-

AMG

1

BAY

1

BMS CLG

5

NOVSANROC NVS

-1 -1

MCK

-1

GLD

-6

ELI TEV

-3

ALL

-1

SHR TKD

-2-1

3 CAGR: Compounded Average Growth Rate.4 Source: EvaluatePharma® World Preview 2016, September 2016.5 Total company sales, November 2016.6 The sales reported on figure 1 includes all prescription (Rx) sales and revenues from alliances / co-promotion / royalties.


10

20

* Upside and downside are not available for what EvaluatePharma® classifies as Other Products (which mainly include mature products, some generics). This explains the limited variability observed for Teva.

30

40

60

70

80

50

Source: EvaluatePharma® November 20162022 Ranking, Top 20 Pharma – Upside vs. Downside* Scenario Ranking

Rev

enue

s, $

Bn

0PFZ JJ GSK ABV AZ AMG BAY BMS CLG NOVSANROC NVS MCK GLD ELL TEV ALL SHR TKD

Consensus



What if, instead of using consensus forecasts, we run the analysis by using the full range of the forecast distribution

within consensus? In our World Preview 2016 report, we flagged some potential risks that could affect the growth

trajectory of the industry, in particular the mounting pressure on pricing on both sides of the Atlantic, the higher

scrutiny posed by national and sub-national decision-makers and a larger than expected impact of biosimilars (which

could indeed turn into a favourable event for some of the companies in this cohort).

Figure 2 shows the 2022 revenues for the top 20 pharmaceutical companies using an upside and downside

scenario – based on respectively the maximum and minimum values in the forecast pool used to generate the

consensus. While it is unlikely that all the events leading to the upper or lower values will materialise, such an

analysis does provide a perspective on the risk and uncertainty implicit within each company’s portfolio.

Figure 2 – 2022 sales ranking based on upside/downside scenarios

Roche and Merck provide two interesting case studies. The Swiss giant is forecast to grow at the market rate,

however the large spread between the upper and the lower bound emphasises the large risk intrinsic in Roche’s

portfolio. On one side, a biosimilar cloud is looming for the established business – although some of Roche’s legacy

brands will be tough nuts to crack. The combined value of Herceptin, Avastin and Rituxan is expected to drop by

$13bn in 2022 in the downside scenario from about $21bn in 2016. On the other side, the high expectations on

Tecentriq and Ocrevus will be challenged by the extremely competitive environment they will be facing respectively

in the immuno-oncology and crowded multiple sclerosis markets.


For Merck the upside from Keytruda is expected to offset the potential setbacks in a highly risky pipeline featuring

complex developments in Alzheimer’s (verubecestat) and in the anti-cholesterol space (the CETP inhibitor, anacetrapib),

as well as the fourth-to-market SGLT2 compound (ertugliflozin, metformin) in diabetes. And Keytruda, despite the

impressive clinical results released over the last few months, will be challenged by the launch of competitive products

and increasing scrutiny from payers. The other growth driver, Zepatier, despite a good start in the first three quarters of

2016 could encounter setbacks from formulary restrictions in 2017 (e.g. Express Scripts7) and the heavy competition plus

market access pressures of the Hepatitis C market.

For both companies, it is certainly not the same reassuring picture painted by looking exclusively at the consensus value.

Uncovering hidden upside

To further explore some of the insights hidden behind consensus, we have explored a metric – the upside/downside

ratio – to identify companies that could either outperform in the coming years or are at risk of significant setbacks. A value

higher than 100% indicates that the upside expected by the brokers is larger than the downside for a given company.

For example, Celgene exhibits the largest ratio within the cohort, 151%. This is calculated by dividing the 35% potential

upside intrinsic in its portfolio by the 23% downside attributable to adverse events (e.g. poor commercial performance

of marketed products, pipeline setbacks).

Figure 3a – Top 20 pharma 2015-22 growth vs. upside/downside ratio



2015-22 CAGR vs. Up/Down Ratio – Top 20 Pharma

2022 Upside/Downside Ratio

120% 140%

CLG

ABVTEV

ELI

JJ

AZAMG

TKD

MCK

GLD

NVS

PFZSAN

BAYNOV

ROC

SHR

BMS

20

15-2

2 C

AG

R

100%80% 160%

15%

10%

5%

0%

-5%

20%

Held Back?

Growth at Risk? Potential Over Performers?

Unexpressed Potential?

GSK

ALL

Source: EvaluatePharma® November 2016

7 Zepatier appeared in the formulary exclusion list for both Express Scripts and CVS Health (August 2016).




The analysis uncovers an interesting scenario for Bristol-Myers Squibb (Figure 3a). According to consensus, BMS

is forecast to grow at a 6.6% CAGR through to 2022. That reflects the value attributed to the bolt-on acquisition

strategy executed by BMS over the last 5 years resulting in a successful portfolio led by Opdivo, Yervoy and Eliquis.

What is also apparent is that BMS seems to have most of its eggs in one basket. In fact, a closer look reveals that

BMS also has the lowest upside/downside ratio in the Top 20 cohort (i.e. 88%), primarily driven by the risks of the

Opdivo-Yervoy franchise. And the pipeline, excluding the vast clinical development programme behind Opdivo, does

not seem to include any other meaningful asset at least based on consensus estimates.

In addition to the upside/downside ratio, another useful metric is the magnitude of the spread, i.e. the sum of

upside and downside values, which represent a good proxy of the level of portfolio uncertainty. In figure 3b, Gilead

Sciences, AbbVie and Eli Lilly show the highest spread in the cohort. The multitude of high impact events expected

for the first two companies could explain the lower than average P/E8 awarded by the market. Eli Lilly, whose P/E is

still above average, could see a correction in the future following the recent disappointment from solanezumab and

the challenges Taltz will be facing.

It is worth noting that the large spread exhibited by Gilead reflects the uncertainty surrounding its lifecycle

management strategy and the ability to offset the decline of Harvoni and Sovaldi (combined revenues of which have

dropped by 44% in Q3 2016 vs. the same quarter in 2015).

Figure 3b – Upside/downside spread vs. consensus forecast

8 The P/E (Price per share /Earning per share) is a ratio often used in company valuation and measures a company’s current share price relative to its earnings per share. Based on Morningstar values reported on November 26, the average P/E of the Top 20 pharmaceutical companies in our cohort is 27.30 (Gilead: 7.00, AbbVie: 16.40, Eli Lilly: 30.10).

Source: EvaluatePharma® November 2016Top 20 Pharma, Portfolio Variance vs. Consensus

-100% -50% 50% 100% 150%

Ratio Spread

-38% 39% PFZ 103% 77%

-42% 42% ROC 101% 84%

-40% 44% NVS 109% 84%

-42% 52% JJ 122% 94%

-33% 31% SAN 94% 64%

-43% 50% MCK 114% 93%

-26% 31% GSK 119% 58%

-48% 62% ABV 130% 110%

-43% 50% AZ 116% 93%

-31% 41% AMG 131% 72%

-25% 25% BAY 99% 50%

-39% 34% BMS 88% 73%

-72% 103% GLD 142% 175%

-23% 35% CLG 151% 59%

-49% 60% ELI 124% 109%

-34% 49% TEV 143% 83%

-41% 44% NOV 107% 86%

-30% 33% ALL 106% 62%

-35% 48% SHR 138% 83%

-28% 35% TKD 127% 63%

0%

% Downside vs. Consensus % Upside vs. Consensus




Which other companies could follow a different trajectory either upward or downward compared to the scenario

outlined by consensus forecasts? In figure 4, the top 20 pharmaceutical companies have been segmented into three

clusters based on 2016-22 CAGR: respectively the top 25%, the middle 50% and the bottom 25%. Within each cluster

we have highlighted the companies which are more likely to over-or underperform based on a combination of the

upside/downside ratio and the spread.

Figure 4 – Companies at risk vs. companies with further upside potential


Growth RankingCAGR 2015-22

1 Ranking based on combination of risk spread (upside + downside vs. consensus) and ratio between upside and downside (ratio >1 → upside > downside).2 Upside vs. consensus over downside vs. consensus (ratio>1 → upside>downside).3 Spread measures the uncertainty of consensus and is based on the sum of upside + downside vs. consensus.

ShireCelgene

BMSAbbVieEli Lilly

TevaNovo Nordisk

AllerganTakeda

J&JRocheBayer

AmgenAstraZeneca

Sanofi

GSKNovartis

PfizerMerck & Co

Gilead

Upside/Risk Ranking1

Uncertainty Spread, Up/Down RatioUp/Down Ratio2

Celgene

BMS

151%

88%

131%

101%

103%

119%

Spread3

59%

73%

72%

84%

77%

58%

Amgen

Roche

GSK

Pfizer

>5.7%

>3.2%

High Growth Cluster

Further Upside Potential

Source of 2022 Revenue

Upside/Downside – Portfolio Mix

Established(Revlimid)

59%

Pipeline(Ozanimod,Mongersen)

16%

RecentlyLaunched

(Otezla,Pomalyst)

24%

% Upside vs.Consensus

% Downside vs.Consensus

27%

25%

82%

Established

Recently Launched

Pipeline

Total Portfolio

-19%

-14%

-54%

At Risk Potential



Established(Eliquis,Yervoy,Orencia)

52%

Pipeline(Prostvac,ARX618,

BMS-931699)

1%

RecentlyLaunched(Opdivo,Empliciti)

46%



29%

38%

130%

Established

Recently Launched

Pipeline

Total Portfolio

-44%

-32%

-100%

CELGENE BMS

Source: EvaluatePharma® data; The analysis focuses on the products where consensus forecast is available.

35% -23% 34%

PORTFOLIO SPREAD PORTFOLIO SPREAD59% 73%

-39%



High growth club: Celgene vs. BMS

Celgene and BMS fall in the same high growth cluster, but if we focus our attention to the upside/downside ratio and

the size of the spread, the two companies sit on polar opposites.

We have already discussed the question marks behind BMS’s growth trajectory. On the contrary Celgene seems

poised for additional upside based on analysis shown in Figure 5.

The upside/downside ratio within each segment of Celgene’s portfolio, including established9, recently launched

and pipeline products, is positive, fuelled by the positive news on ozanimod (both in multiple sclerosis and ulcerative

colitis) as well as the high hopes pinned on mongersen and the rest of the oncology pipeline. Additionally value

creation is fairly balanced across the portfolio where Revlimid still takes the lion’s share.

Figure 5 – Celgene vs. BMS

9 Established products have been defined as all medicines launched before January 2013.




Medium Growth Cluster




Established(Enbrel,Prolia,

Kyprolis)

73%

Pipeline(Amjevita,AMG 334)

11%

RecentlyLaunched(Repatha)

16%



34%

55%

68%

Established

Recently Launched

Pipeline

Total Portfolio

-27%

-30%

-56%

At Risk Potential



Established(Perjeta,Avastin,

Herceptin,Rituxan)

57%

Pipeline(Ocrevus,

Emicizumab)

19%

RecentlyLaunched(Kadcyla,Gazyva,

Tecentriq)

24%



33%

32%

83%

Established

Recently Launched

Pipeline

Total Portfolio

-32%

-45%

-67%

AMGEN ROCHE


41% -31% 42%


-42%

Medium growth club: Amgen vs. Roche

In the mid-tier, we have compared two of biopharma’s biggest names, Amgen and Roche. Both companies are

forecast to grow by around 4% through to 2022 (respectively, 3.6% and 4.2%). In the previous section we already

highlighted why we believe that Roche, despite the promising pipeline and the recent launches, has one of the

riskiest portfolios among the top 10.

Figure 6 – Amgen vs. Roche

On the other end of the spectrum, Figure 6 illustrates how Amgen has the potential to exceed consensus estimates.

While the portfolio is heavily weighted around established brands, hence exposed to the biosimilar risk threatening Enbrel,

the denosumab franchise (Prolia, Xgeva) and Kyprolis are set to continue driving growth. Among the recently launched

products Repatha seems to have an edge over Praluent in the PCSK9 race following the release of the Glagov data which

showed that Amgen’s compound caused a reduction in atheroma volume vs. the standard of care (i.e. statins). Additional

upside is expected from the pipeline driven by the promising biosimilar franchise (Amjevita, a biosimilar version of AbbVie’s

Humira) and the launch of AMG 334, one of the four CGRP receptor blockers approaching the migraine market, which

has just released positive top-line results from the STRIVE Phase 3 trial. It must be noted that brokers have been so far

fairly bullish in estimating the impact of this novel class and the presence of fairly cheap and mostly generic options

already available in the migraine market should cause some concerns as there are many parallels with the challenging

environment that Praluent and Repatha, and to a certain extent Entresto, had to face at launch.




Low Growth Cluster




Established(Advair,Pediarix,Synflorix)

48%

Pipeline(Shingrix,

FF/UMEC/VI)

9%

RecentlyLaunched

(Triumeq, Tivicay,Breo Ellipta,

Nucala)

43%



16%

38%

83%

Established

Recently Launched

Pipeline

Total Portfolio

-18%

-27%

-70%

At Risk Potential



Established(Prevnar 13,

Lyrica,Xeljanz)

71%

Pipeline(Eucrisa,

Avelumab)

13%

RecentlyLaunched(Ibrance,

Trumenba)

16%



28%

38%

102%

Established

Recently Launched

Pipeline

Total Portfolio

-28%

-50%

-77%

GSK PFIZER


31% -26% 39%


-38%

Low growth club: GSK vs. Pfizer

The lower tier group contains some of the largest pharma conglomerates, such as Pfizer, GSK, Novartis and Merck & Co.

If we look at the combination of spread and upside/downside ratio, GSK appears to have the best profile (Figure 4). The

analysts seem positive about the ability of GSK to maximise the lifecycle management of its respiratory franchise10, with

sales from new launches forecast to grow the value of the portfolio from $7.1bn in 2015 to $8.3bn in 2022 according to

consensus estimates. The key upsides are expected from Nucala and Breo Ellipta where the best case scenario is worth

close to $5bn in 2022, thus exceeding the $4.5bn Advair is expected to lose in 2022 under worst case assumptions.

Figure 7 – GSK vs. Pfizer

10 It includes Advair, Breo Ellipta, Anoro Ellipta, Ventolin, Incruse Ellipta, Nucala and the triple combination fluticasone, vilanterol, umeclidinium.

But the real boost is expected from the newly established HIV franchise where Triumeq and Tivicay could move

beyond the current consensus of $7bn (combined value) to $9.5bn in the highest scenario. While some of the

upsides from the new therapies especially Nucala and Triumeq could be challenged by upcoming competition, the

analysis does seem to reward the focused strategy implemented by GSK completed by the asset swap deal with

Novartis in early 2015.




2015-22 CAGR vs. Up/Down Ratio

20

15-2

2 C

AG

R

% of Ph2 & Ph3 Products Without Consensus

80%50% 100%

15%

10%

5%

-%

-5%

20%

60% 70% 90%

SHR

ABV

SAN AZ

NVSMCK

ROC

ELI

GLD

PFZ

ALL NOV

JJ

TKDTEV

CLG

AMG

High Growth / Low ‘Unvalued’ Potential High Growth / High ‘Unvalued’ Potential

Low Growth / Low ‘Unvalued’ Potential Low Growth / High ‘Unvalued’ Potential

BMS

GSK

Source: EvaluatePharma® November 2016

Within the same group, Pfizer is forecast to grow at the same rate of GSK. However an in-depth analysis of its

portfolio paints a slightly different picture. Figure 7 highlights an inherent risk within the recently launched products

with downside significantly higher than the upside, in particular Ibrance where strong uptake could be threatened by

the launch of competing products, such as Novartis’s ribociclib (Lilly’s abemaciclib is probably less of a concern). And

while Pfizer’s pipeline seems to show potential upside, the spread is very large and is heavily dependent upon the

outcomes of two products, crisaborole and the anti-PD-L1 monoclonal antibody avelumab and the biosimilar pipeline

inherited through the acquisition of Hospira.

Are we leveraging the full picture?

The analyses presented are entirely based on the sell-side analysts’ forecasts pooled by EvaluatePharma®. Does

that provide a full picture of the value of a company pipeline or are there potential gaps that should be further

investigated by future streams of research?

In Figure 8 we have analysed the number of compounds currently in phase 2 and 3 for which a consensus forecast is

not yet available. It appears that around 70% of products currently in development do not have a consensus forecast

attached to them.

Figure 8 – Share of Ph2 and Ph3 products without consensus




Conclusion

We began by highlighting the strong empirical evidence in the value of aggregating different

forecasts11. We firmly believe that consensus forecasts represent a robust tool to map market

trends, reliably sizing therapy areas and formulate assumptions on the strategic direction of

the industry. The analysis aims to encourage forecasters to combine the value of consensus

with its distribution and not to underestimate the presence and value of understanding

outliers. In some instances, the presence of outliers could help spot key signals that will help

make better predictions on future trends as well as conduct more thorough portfolio analysis

by weighting both risks and potential sources of upside.

Lastly, in Evaluate’s long-term ambition to improve the ability of the pharmaceutical industry to deploy capital more

efficiently, we urge researchers in this field to focus more attention on predictions of earlier stage compounds where

the availability of robust information is still lacking and where billions of dollars are at stake that could potentially be

re-directed to fund more promising streams of clinical research.

Evaluate remains committed to playing a key role in maximising the value and utility of consensus, especially for early

stage compounds, in the pharmaceutical industry through the continuous enhancements of its products and services.

These gaps may hide some potential diamonds in the rough, which could turn into relevant assets in the long run.

Among the top 20 pharmaceutical companies, it is worth noticing that BMS which we identified as a fast growing

company with a high component of risk built into its portfolio actually exhibits the largest unvalued pipeline in the

cohort. This final analysis also further reinforces the positioning of Celgene whose pipeline may contain additional

assets to further boost the already impressive growth trajectory anticipated by the analysts’ consensus.

11 From “The signal and the noise” by Nate Silver, 2012.


Additional complimentary copies of this report can be downloaded at: www.evaluategroup.com/Forecastingexcellence1

Evaluate Headquarters Evaluate Ltd. 1 1 -29 Fashion Street London E1 6PX United Kingdom T +44 (0)20 7377 0800 F +44 (0)20 7539 1801

Evaluate North America EvaluatePharma USA Inc. 15 Broad Street, Suite 401 Boston, MA 02109, USA T +1 617 573 9450 F +1 617 573 9542

Evaluate Japan Evaluate Japan KK Akasaka Garden City 4F 4-15-1 Akasaka, Minato-ku Tokyo 107-0052, Japan T +81 (0)80 1164 4754

www.evaluate.com

Evaluate is the trusted provider of commercial intelligence including product sales and consensus forecasts to 2022 for commercial teams and their advisors within the global life science industry. We help our clients make high value decisions through superior quality, timely, must-have data and insights, combined with personalised, expert client support.

EvaluatePharma® delivers exclusive consensus sales forecasts and trusted commercial insight into biotech and pharmaceutical performance.

@EvaluatePharma

EvaluateMedTech® sets a new standard in commercial analysis and consensus forecasts of the global medical device and diagnostic industry.

@EvaluateMedTech

EvaluateClinical Trials® delivers unique clinical trial intelligence expertly curated to efficiently analyse the global clinical trial landscape.

@EPClinicalTrial

EP Vantage an award winning editorial team, provides daily commentary and analysis with fresh perspectives and insight into current and future industry trends.

@EPVantage

Evaluate Custom Services provides customised solutions to help you access, analyse and manage the information you need to support effective decision-making.

The Evaluate services enable the life science community to make sound business decisions about value and opportunity.

Media contact and for general questions on the report:Christine Lindgren, Vice President Marketing, EvaluatePharma USA Inc.

[email protected] T +1 617 866 3906

https://twitter.com/evaluatepharma

https://twitter.com/@EvaluateMedTech

https://twitter.com/@EPClinicalTrial

https://twitter.com/@EPVantage

Evaluate Forecasting Excellence The Value of Outliers in ... ›...

Documents

Transcript of Evaluate Forecasting Excellence The Value of Outliers in ... ›...