16
The Price-Concentration Relationship in Early Residential Solar Third Party Markets Jacquelyn Pless INET, Oxford Martin School, University of Oxford June 21, 2017 2017 IAEE Summer Conference, Singapore 1 / 13

The Price-Concentration Relationship in Early Residential ... · The Price-Concentration Relationship in Early Residential Solar ... U.S. Solar PV Installations, 2000-2015 Source:

  • Upload
    lylien

  • View
    215

  • Download
    0

Embed Size (px)

Citation preview

The Price-Concentration Relationship in EarlyResidential Solar Third Party Markets

Jacquelyn Pless

INET, Oxford Martin School, University of Oxford

June 21, 20172017 IAEE Summer Conference, Singapore

1 / 13

Solar PV Market Growth

U.S. Solar PV Installations, 2000-2015

Source: GTM Research/SEIA U.S. Solar Market Insight Report

2 / 13

Third Party Ownership (TPO) Business Model

Source: NREL

3 / 13

Increasing Market Share for TPO

Third Party Owned Systems (Percent of New Solar Installations)

Source: U.S. Solar Market Insight Report, Q3 2012

4 / 13

Are Solar Leasing Companies Stealing From Consumers?

5 / 13

Research Question

• Residential PV and TPO market growth prompts newquestions about how the nature of competition affects prices

• The relationship between price and market competition istheoretically ambiguous

• Increases in concentration could enable firms to increase markups ordecrease prices as a strategy to obtain customers

• Existing literature examines determinants of solar prices andcorrelations (Nemet et al., 2016)

• This research: How does market (installer) concentrationimpact solar PV TPO prices?

• Focus on the leasing market• Take an instrumental variables approach• Preview of results: prices are lower in more concentrated markets

6 / 13

Research Question

• Residential PV and TPO market growth prompts newquestions about how the nature of competition affects prices

• The relationship between price and market competition istheoretically ambiguous

• Increases in concentration could enable firms to increase markups ordecrease prices as a strategy to obtain customers

• Existing literature examines determinants of solar prices andcorrelations (Nemet et al., 2016)

• This research: How does market (installer) concentrationimpact solar PV TPO prices?

• Focus on the leasing market• Take an instrumental variables approach• Preview of results: prices are lower in more concentrated markets

6 / 13

Data• Developed database of solar installations in SDG&E in

California, matching solar PV system level data to TPOcontracts

• Residential solar PV installations—combined address-level solar systemdata from ERP and CSI from 2007 to Q1 2013

• Constructed measures of competitiveness and/orconcentration of solar installers within a market

• Market is defined as census tract and time periods arequarter-years + preceding two quarters

• Eight measures of competitiveness: HHI, no. of installers,“monopoly” in two ways, etc.

• Construct TPO contract prices• Obtained TPO contracts for SDGE• Derive the net present cost of each contract

• Other controls: household and census tract leveldemographics, housing density, etc.

7 / 13

Summary Statistics of Market Concentration and SystemPrice Variables

8 / 13

Econometric Framework and Identification Strategy

Pi = α + β1Concjt + Xiθ + γk + δl + µm + εi

Pi price per watt of system i (net present cost)Concjt installer concentration in census tract j at time tXi matrix of control variablesγj module model fixed effectsδk installer fixed effectsµt year by quarter time fixed effectsεi disturbance term

• Market competition is endogenous• Unobserved shocks to cost and demand can affect prices, so market

structure could be correlated with unobservables• Output of firms affects both prices and concentration• Simultaneity bias if setting prices aggressively low deters entry or drives

out other firms

• Instrumental variables approach• Isolate exogenous variation in market concentration by using the

percentage of single-family homes within a Census tract that areowner-occupied as an IV

9 / 13

Econometric Framework and Identification Strategy

Pi = α + β1Concjt + Xiθ + γk + δl + µm + εi

Pi price per watt of system i (net present cost)Concjt installer concentration in census tract j at time tXi matrix of control variablesγj module model fixed effectsδk installer fixed effectsµt year by quarter time fixed effectsεi disturbance term

• Market competition is endogenous• Unobserved shocks to cost and demand can affect prices, so market

structure could be correlated with unobservables• Output of firms affects both prices and concentration• Simultaneity bias if setting prices aggressively low deters entry or drives

out other firms

• Instrumental variables approach• Isolate exogenous variation in market concentration by using the

percentage of single-family homes within a Census tract that areowner-occupied as an IV

9 / 13

Econometric Framework and Identification Strategy

Pi = α + β1Concjt + Xiθ + γk + δl + µm + εi

Pi price per watt of system i (net present cost)Concjt installer concentration in census tract j at time tXi matrix of control variablesγj module model fixed effectsδk installer fixed effectsµt year by quarter time fixed effectsεi disturbance term

• Market competition is endogenous• Unobserved shocks to cost and demand can affect prices, so market

structure could be correlated with unobservables• Output of firms affects both prices and concentration• Simultaneity bias if setting prices aggressively low deters entry or drives

out other firms

• Instrumental variables approach• Isolate exogenous variation in market concentration by using the

percentage of single-family homes within a Census tract that areowner-occupied as an IV

9 / 13

Main Results 1: Using HHI as Measure of Concentration

10 / 13

Main Results 2: “Monopoly” if HHI <0.25 as Measure ofConcentration

11 / 13

IV v. Non-IV Results for HHI and “Monopoly” Measures

12 / 13

Discussion & Conclusions

• Main result: Lower prices are associated with moreconcentrated (less competitive) markets

• Robust across numerous measures of market competition

• A few potential explanations for results• Entry deterrence strategy to ensure larger market share• Firms differentiate product offerings so as to not compete strictly on price

• Possibly unique to very early stage markets; future workfocusing on more mature markets needed

13 / 13