The Organization of Exports in a Growing Industry: Chilean Wine Exports to UK

1
The Organization of Exports in a Growing
IndustryChilean Wine Exports to UK

• Rocco Macchiavello
• Oxford (Nuffield) and CEPR
• October, 2nd 2007
• E-mail rocco.macchiavello_at_nuffield.ox.ac.uk

2
The Organization of Exports in a Growing
IndustryChilean Wine Exports to UK

• Rocco Macchiavello
• - Work in Progress -
• Oxford (Nuffield) and CEPR
• October, 2nd 2007
• E-mail rocco.macchiavello_at_nuffield.ox.ac.uk

3
Motivation (Broad Themes)

• This project is about
• Development Economics
• (Industrial Upgrading and Export Growth),
• International Economics
• (Microeconomics of Trade Costs)
• Organizational Economics
• (Economics of Relational Contracts)

4
Motivation (Development Economics)

• How do LDC break in new markets where quality is
  important?
• Structural transformation has historically
  occupied the driving seat on the road towards
  economic development
• Resources move out of traditional agriculture
  into modern sectors (e.g. manufacturing) as
  countries learn how to produce and export new
  things
• Many firms / industries in LDC countries are
  reaching world best standards in terms of
  manufacturing capabilities
• McKinsey (2001), Sutton (2006), World Bank
  (2007), Bloom et al. (2007)

5
Question(s)

• Looks at industrial upgrading and export growth
  through the lens of the organization of marketing
  channels in export markets
• Focus on the relationships between exporters and
  foreign agents entry point to the export
  market
• 1. Which characteristics of these relationships,
  if any, correlate with export performance?
• 2. Are those relationships valuable?
• 3. Which forces shape the formation and
  maintenance of these relationships?

6
Motivation (Organizational Economics)

• Economics of relational contracts in markets
  where quality is important
• Looking at exports
• - makes it easier to track transactions,
• - can use x-rate variation to identify shocks to
  surplus,
• - firms from LDC are unlikely to have access to
  informal networks to enforce contracts
• Can we empirically test for different models of
  reputation formation?
• (Exploiting exogenous exchange rate and whether
  shocks)

7
Motivation (Organizational Economics)

• Economics of relational contracts in markets
  where quality is important
• Looking at exports
• - makes it easier to track transactions,
• - can use x-rate variation to identify shocks to
  surplus,
• - firms from LDC are unlikely to have access to
  informal networks to enforce contracts
• Can we empirically test for different models of
  reputation formation?
• (Exploiting exogenous exchange rate and whether
  shocks)

8
Marketing Channels and Export

• Contacting distributors in the importing
  countries is a critical decision in the process
  of exporting
• once a company is organized to handle
  exporting, a proper channel of distribution needs
  to be carefully chosen for each market,
  potential entrants should investigate potential
  representatives or distributors carefully before
  entering into an agreement
• U.S. Department of Commerce
• (t)he greatest challenge for any exporter is to
  gain a compatible match of agent / distributor
• International Wine Marketing Handbook
• The most important step in setting up your
  business is finding the contacts for commercial
  distribution
• Home Based Business Opportunities

9
Marketing Channels and Export

• Long-term collaborative relationships with
  developed country buyers might be particularly
  valuable to developing country exporters
• important source of information about developed
  country markets conditions (e.g. product quality
  and delivery standard requirements)
• simply crossing borders considerably complicates
  transactions
• monitoring costs, enforcing contracts is
  harder, distance increases scope for opportunism
• Long lasting relationships mitigate the
  opportunism arising from lack of enforceable
  contracts and might facilitate the transmission
  of hard to codify information (e.g. Macaulay
  (1963), Klein and Laffer (1981))

10
Todays Talk

• Today I will only
• - describe certain features of a particular
  industry,

11
Todays Talk

• Today I will only
• - describe certain features of a particular
  industry,
• - offer a tentative explanation,

12
Todays Talk

• Today I will only
• - describe certain features of a particular
  industry,
• - offer a tentative explanation,
• - mention directions for future work

13
Related Literature

• Reputation and Contracts (part 1)
• Banerjee and Duflo (2002)
• Firms heterogeneity, Contracts and International
  Trade
• Antras (2003), Antras and Helpman (2004),
  Spencer (2005), Araujo and Ornelas (2007)
• Economics of Export
• - export and trade costs Roberts and Tybout
  (1997) Anderson and Van Wincoop (2004) Goldberg
  and Knetter (1997) Besedes and Prusa (2006)
  Bernard et al. (2007)
• - choice of invoice currency Goldberg and Tille
  (2006) Friberg and Wilander (2007) Donnenfeld et
  al. (1991, 1995, 2003)
• Reputation and Contracts (part 2)
• MacLeod and Malcomson (1988) Levine (2002, 2003)
  Baker et al. (2002, 2004), Ramey and Watson
  (1997) Rauch and Watson (2003) McMillan and
  Woodruff (2003) Johnson, McMillan and Woodruff
  (2002)
• International Marketing Channels (Business
  Literature )
• - Very large, it stresses the importance of
  organizing appropriate marketing channels

14
Why Chilean Wine in the UK?
15
Why Chilean Wine Exports to U.K.?

• Wine industry as interesting case of
  globalization at work,
• Chiles wines have done very well in recent
  years
• - industrial upgrading to (almost) worlds best
  practices,
• - medals in international concourses (Data to
  come )
• 3. UK is most competitive wine market in the
  world (historical reasons no domestic
  production recent market trends),
• 4. Production and distribution of wine are
  complex, human physical capital intensive
  economic activities

16
Historical Background

• Stable consumption at the world level vs. major
  supply side restructuring
• ? New world producers (Australia, South Africa,
  New Zealand, California and Chile)
• - technological innovations (obtain consistent
  quality in warmer climate),
• - upgrading in producers capabilities ("flying
  winemakers" across regions), and
• - radical changes in marketing practices
• Share of wine production internationally traded
  10 in the '60s, 15 in the '90s and 30 now.
• Chile to UK trade
• - Demand ? Resale Price Maintenance Act (mid
  60s), ? demand for wine (late '80s)
• - Supply two century history as producer of
  regular wines, macroeconomic conditions right and
  good investment climate
• - Quickly, very high entry rates Chile now
  market share is around 6 (rank 7th)

17
Chilean Wine Production and Exports
18
Exporting Wineries in the UK
19
Industrys Description
20
Wine Trade in a Nutshell, 1

• How is wine produced, imported and distributed
  from Chile to UK?
• It is said that the quality of the wine is made
  in the vineyards. Once picked, grapes need to be
  processed as quickly as possible.
• Some growers crush their own grapes, ferment the
  juice, age and bottle the wine and then sell it
  using their own names.
• Most likely (in the old world) growers bring the
  grapes to a cooperative to be processed, or will
  sell the juice to a merchant (negociant), perhaps
  through the services of a local broker.
• Regardless of the organization of production
  (vertically integrated winery, cooperative, or
  negociant), it is usually at this stage that
  foreign buyers encounters the production chain.

21
Wine Trade in a Nutshell, 2

• Two kinds of buyers
• - broker, for large clients (supermarkets, high
  street chains) or for producers. Wine is sold
  "ex-cellar under commission (normally 2 to 5
  of the sale).
• - shipper looking for wines to add to a
  portfolio. Wines are shipped to destination, held
  in stock and distributed to smaller clients on a
  "duty-paid" basis. More expensive and with
  greater variation in margins.
• 2. Key distinction
• broker does not own the wine at any time. The
  shipper does. Shippers bear financial risk, but
  there is more competition within their portfolio.
  
• 3. Two categories are not mutually exclusive.
  Numerous importers act as brokers and shippers
  even for different brand/client of the same
  winery.
• 4. Agency problem marketing effort, hold-up once
  a brand has been developed
• 5. The choice of partner and which kind of
  contractual arrangement is crucial for the
  success of the exporter.
• (t)he greatest challenge for any exporter is to
  gain a compatible match of agent / distributor
  (Handbook of International Wine Marketing)

22
Contracts Characteristics
Source Authors Interviews with Export
Directors. Non Random Sample

• Very informal relationships either no written
  contract at all, or very loose.
• If written contract, almost only bilateral
  excusivity
• 3. Conversations suggests that there are
  cross-country differences e.g. exports to USA
  Russia higher use of written contracts.
• 4. is by far the most common invoice currency
  i) Lower transaction costs, ii) History, iii)
  Bargaining Power
• 5. Practitioners believe that invoicing in keep
  gives better control over price point

23
Value Chain Decomposition

1. Costs of Trade (difference between Retail and
FOB) are large FOB is Lower than 40 of retail
price 2. Even for bottled wine, Transportation
Handling Costs Excise are around 30 of FOB
prices and do not exceed 15 of retail price
24
Data

• Original Dataset, 1997-2006 period, Various
  sources
• Relationships
• - I have matched each winery in each year with
  its agent(s) in the U.K. (Source Harpers Trade
  Directory, Wines of Chile, web)
• Export of wines to the U.K.
• - volume and FOB value (Source Chilean Customs)
• Winery Characteristics
• - Focus on Location and Vineyards (Source
  Compendio Vitivinícola de Chile)
• Agents Characteristics
• - Focus on Age, Size, and Geographical
  Composition of their Wine Portfolios (Source
  Harpers Trade Directory)

25
Distribution of Relationships Lengths
26
Number of Partners
27
Descriptive Statistics
28
Do Relationships Characteristics Correlate with
Export Performance?
29
Export Performance
30
Export Performance
31
Export Performance
32
Mechanics Within Agent Competition

• Agents differ w.r.t. geographical composition of
  their portfolios
• A depreciation of South African Rand implies
  higher margins on South African wines marginal
  cost of effort in marketing Chilean wines goes
  up.
• Combine exchange rate fluctuations in New World
  Producers with agents share of business in each
  of these countries
• Consider Australia, California, New Zealand and
  South Africa.
• Main Assumptions
• Extensive margin adjustment is difficult,
• Incomplete pass-through

33
Export Performance
Agents portfolio exchange rate
34
Export Performance
35
Export Performance
Agents differ w.r.t. geographical composition of
their portfolios ? South African Rand, ? margins
on SA wines, ? marginal cost of effort in
marketing Chilean wines. X-rate in N.W.
Producers agents share of business in
Australia, California, New Zealand and South
Africa Main Assumptions Extensive margin
adjustment is difficult, Incomplete pass-through
36
Agents Fixed Effects
Note estimated FE are normalized around the mean
37
Agents Fixed Effects
Note estimated FE are normalized around the mean
38
Export Performance Agents FE
39
Summing Up

• Conditional on winery year FE, relationship Xs
  correlate with export performance
• FOB Prices are higher
• - in older relationships after the first year
• - in relationships started with more surplus
• - when the (opportunity) cost of marketing
  effort is lower
• Learning, Bargaining or Quality?
• Agents FE are highly heterogeneous some agents
  are good, others are bad
• how much of it is due to Sorting?

40
Which forces shape formation and maintenance of
these relationships?
41
Formation and Maintenance of Relationships

• 1. (Almost) No evidence of sorting between
  wineries agents in the market
• 2. Duration analysis sheds light on the
  economics of these relationships
• - description
• - exchange rate as shocks to surplus,
• - complete contracts framework,
• - evidence, not consistent with it,
• - mutually bilateral arrangements

42
Is There Sorting?
43
Is There Sorting?
44
(Relational) Contracting Theory

• Assume perfect contracting
• by definition only level of surplus matters for
  the decision to continue the relationship,
  parties agree to maximize value of relationship
• Imperfect contracting
• marketing effort (hard to monitor hold-up)
• Limits to non-linear pricing,
• why arent LT contracts signed?
• effort not written in a contract relational
  contracts
• if one party reneges on implicit contract,
  relat. breaks down

45
Duration Analysis
46
Relationships as assets

• Domestic importer and a foreign winery.
• Winery produces 1 unit of wine per period at cost
  1 in terms of foreign currency
• Per period surplus
• Exchange rate dynamics
• m is the quality of the match / relationship

47
Relationships as assets

• Quality of the match is m, random draw from G(m)
• Decision start a new relationship or keep on
  looking.
• New relationship starts next period sunk cost
• Exchange rate proxy for per-period surplus

48
Relationships as assets

• Value of the asset relationship with match m
• Value of the option to start a relationship

49
Discussion

• 1. the unobservable m can be proxied with xrate
  at the time the relationship was started
• 2. can update m conditional on staying
  together, the toughest xrate is summary
  statistics for m
• 3. Normalizing xrate allows to control for year
  FE ( heterogeneous TE)
• Caveat changes in xrates might matter for other
  reasons
• - proxy for uncertainty (? uncertainty ? wait
  and see)
• - market thickness (many break-ups in previous
  period)

50
Exchange rate, (Nominal and Sustainable)
t
51
Exchange rate, (Nominal and Sustainable)
s
s
t
52
Exchange rate, (Nominal and Sustainable)
Sustainability Effect
Sustainability Region
s
s
t
53
Exchange rate, (Nominal and Sustainable)
Selection Effect
Sustainability Region
s
s
t
54
Exchange Rate and Entry
55
Perfect Contracting
56
Imperfect Contracting Evidence
57
Entry of New Wineries
58
Imperfect Contracting

• Mutual bilateral exclusivity is consistent with
  imperfect contracting (Barnheim and Whinston
  (1998))
• explicit contractual exclusion (as distinguished
  from a retailer's unilateral decision to carry
  only one product) do not arise unless common
  representation involves externalities among the
  manufacturers that cause inefficiency in
  incentive contracting
• exclusive dealing if costs of contractual
  externalities are large relative to the benefits
  of variety
• Benefits of variety might be low, if consumer do
  not perceive that Chilean producers can offer
  variety (e.g. grapes, origin,) exclusive
  dealings might be efficient
• Exports directors are concerned about
  competition of other Chilean wineries
• Hard to speak with importers to reassure them,
  exclusivity might be a way to commit not to
  switch agent

59
A story consistent with the evidence

• Long term relationships with good agents are
  valuable,
• However
• - they are costly to build,
• - there are relatively few good agents
• Contractual frictions and relatively small market
  share imply that only bilaterally mutual
  relationships can be sustained therefore
• ? limited sorting wineries are initially unknown
  and forced to look within a restricted pool of
  unmarried agents,
• (If) good agents are taken by early entrants,
• Later entrants are stuck in short lived
  relationships with bad agents (unlikely that
  worse firms entered the market later on )

60
Implications

• Industrial upgrading
• 1) appropriate marketing channels to foster
  exports success,
• 2) LT relationships with good agents firms
  "marketing" capabilities,
• 3) Policies role of Large Firms
• - early entrant create the market, but barriers
  to late entry,
• - horizontal cooperation industry Pol. Ec.
  market share as a public good
• - access to credit to pursue FDI lack of
  skills
• Trade Costs
• - trade costs are heterogeneous across firms,
• Organizational Economics
• - Observed changes in organizational practices
  associated with globalization
• - Tension between survival of LT relationships
  and the process of globalization
• - Increased competitive pressures destroy
  relationships,
• - Value of long lasting relationships is not
  reduced by competition

61
Future Work / Results

• New Zealand, Australia and Californian wineries
  are more likely to share same agent,
• Controlling for quality (measured in concourses),
  Chilean wines have lower retail prices than other
  NWP
• - I still do not know how much of this difference
  can be explained by agents fixed effects
• - I am trying to get finer export data to match
  FOB prices at the bottle level with medals in
  concourse
• More work needs to be done
• i) track evolution in the distribution of
  surplus,
• ii) exploit heterogeneity across wineries
• iii) exploit random variation in whether

62
Thanks !!