Anticompetitive Effects of Common Ownership

The combined holdings of BlackRock, Vanguard, and State Street make them the largest investor of 88% of all firms in the S&P 500.

“Fichtner, Heemskerk, and Garcia-Bernardo (2016) calculate that the combined holdings of BlackRock, Vanguard, and State Street make them the largest investor of 88% of all firms in the S&P 500.”

• cites Fichtner, Heemskerk, and Garcia-Bernardo (2016) sourceSupported

The average MHHI delta in the airline industry generated by common ownership in 2014Q4 implies increases in concentration more than 10 times higher than the threshold that would likely generate antitrust concerns according to the guidelines.

“Thus, the average MHHI delta in the airline industry generated by common ownership in 2014Q4 implies increases in concentration that are more than 10 times higher than the threshold that would likely generate antitrust concerns according to the guidelines, if consequentially applied to MHHI.”

• AssumptionThe authors state this as an assumption: Control is proportional in the fraction of votes held.

Going from the 25th to the 75th percentile of routes by MHHI delta increases prices by 4.3%.

“Similarly, going from the 25th to the 75th percentile increases prices by 4.3%.”

Going from the 10th to the 90th percentile of routes by MHHI delta indicates an 8.2% increase in fares.

“Going from the 10th to the 90th percentile of routes by MHHI delta indicates an 8.2% increase in fares.”

The stark increase in MHHI delta in 2009 coincides with BlackRock's acquisition of Barclays Global Investors.

“The stark increase in MHHI delta in 2009 coincides with BlackRock's acquisition of Barclays Global Investors that we will later use in one of our identification strategies.”

The same fund family is often the single largest shareholder of several firms in the same industry.

“the same fund family is often the single largest shareholder of several firms in the same industry, with similarly diversified investors following suit.”

The correlation between MHHI and HHI is 0.87.

“The correlation between the MHHI and the HHI is 0.87.”

Variation in MHHI delta driven by changes in ownership by shareholders with little effective control does not correlate with price changes.

“We also show in several placebo tests that variation in MHHI delta driven by changes in ownership by shareholders with little effective control does not correlate with price changes.”

An instrumented version of the regression produces higher estimates of β.

“If this theory is correct, an instrumented version of the above regression should produce higher estimates of β, which is a hypothesis we later test and find support for.”

The anti-competitive incentives implied by common ownership concentration alone are more than ten times larger than what the FTC/DOJ 2010 horizontal merger guidelines presume likely to enhance market power.

“are more than ten times larger than what the FTC/DOJ 2010 horizontal merger guidelines presume”

• AssumptionThe authors state this as an assumption: Control is proportional in the fraction of votes held.

There is a large and significant positive effect of MHHI delta on average fares across all specifications.

“We find a large and significant positive effect of MHHI delta on average fares across all specifications.”

Weighting by average passengers in the market over time, the average MHHI delta in 2014 is 2,044.

“Weighting by average passengers in the market over time, the average MHHI delta in 2014 is 2,044.”

• AssumptionThe authors state this as an assumption: Control is proportional in the fraction of votes held.

The correlation between MHHI delta and HHI is -0.69.

“The correlation between MHHI delta and HHI is -0.69.”

Across the entire sample, about 5% of routes have an MHHI delta close to zero, meaning there is no common ownership.

“Across the entire sample, about 5% of routes have an MHHI delta of close to zero -that is, there is no common ownership.”

Changes in passenger volume are negatively related to changes in common ownership.

“We also find that changes in passenger volume are negatively related to changes in common ownership”

The BlackRock acquisition of BGI itself increased average ticket prices by about half a percent.

“Multiplying these estimates with the average route-level increase in common ownership due to the consolidation event indicates that the acquisition itself increased average ticket prices by about half a percent.”

• AssumptionThe authors state this as an assumption: Control is proportional in the fraction of votes held.

In specification (3), adding controls for market characteristics leaves the coefficients of both HHI and MHHI delta positive and statistically and economically significant, albeit slightly attenuated.

“In specification (3), we add controls for market characteristics.The coefficients of both the HHI and the MHHI delta remain positive and statistically and economically significant, albeit slightly attenuated relative to the first specification.”

Ticket prices are approximately 3-7% higher on the average US airline route than would be the case under separate ownership.

“Using fixed-effect panel regressions, we find that ticket prices are approximately 3-7% higher on the average US airline route than would be the case under separate ownership.”

• AssumptionThe authors state this as an assumption: Control is proportional in the fraction of votes held.

An increase in MHHI delta from 0 to 2,000 is associated with an increase in average fares of 4%.

“The coefficient of 0.192 in the first specification with only time-and market-carrier fixed effects implies that an increase in the MHHI delta from 0 to 2,000 (approximately the weighted average level of MHHI delta in 2014Q4) is associated with an increase in average fares of 4%.”

A linear fit indicates a positive raw correlation between airfares and MHHI delta across markets.

“A linear fit indicates a positive raw correlation between airfares and MHHI delta across markets.”

No significant effects obtain when control is assumed by shareholders that cannot reasonably be expected to have control, such as very small and short-term shareholders.

“The fact that no significant effects obtain when we assume control by shareholders that cannot reasonably expected to have control (such as very small and short-term shareholders)”

• AssumptionThe authors state this as an assumption: Control is proportional in the fraction of votes held.

Product prices may be 10-12% higher due to common ownership, based on difference-in-differences using BlackRock's acquisition of BGI.

“They indicate that product prices may be 10-12% higher due to common ownership.”

• AssumptionThe authors state this as an assumption: Control is proportional in the fraction of votes held.

In the 2014Q4 MHHI delta distribution, the 10th percentile is at 109 HHI points, the 25th at 1,421, the median at 2,684, the 75th at 3,642, and the 90th percentile at 4,184 HHI points.

“In the 2014Q4 distribution, the 10th percentile is at 109 HHI points, the 25th at 1,421, the median at 2,684, the 75th at 3,642, and the 90th percentile is at 4,184 HHI points.”

The magnitude of common ownership concentration dwarfs the time-series variation in HHI.

“The magnitude of common ownership concentration furthermore dwarfs the time-series variation in HHI.”

The results with Banzhaf indices are similar in magnitude to the baseline.

“The results are similar in magnitude to the baseline, consistent with the idea that the proportional control assumption is not driving the baseline results.”

Common ownership by short-horizon shareholders yields coefficients of varying sign and weak (if any) significance.

“We find that only common ownership by long-horizon shareholders has a significant positive effect on prices, while common ownership by short-horizon shareholders yields coefficients of varying sign and weak (if any) significance.”

The implied change in MHHI delta is larger than 100 HHI points in more than 2,000 routes.

“The mean and median across routes of the implied change is 91 HHI points; the implied change is larger than 100 HHI points in more than 2,000 routes; the largest implied increase is 281 HHI points.”

If sampling only market-carriers in bankruptcy, there is no effect of MHHI delta.

“Table C.3 shows that if we sample only market-carriers in bankruptcy, there is no effect of MHHI delta (but also no effect of HHI, maybe because of the reduced sample size).”

Dividing the coefficient from the quantity regression specification (1), -0.665, by the coefficient from the price regression specification (4), 0.325, implies an elasticity of demand in the average route of -2.05.

“Dividing the coefficient from the quantity regressions specification (1), -0.665, by the coefficient from the price regressions specification (4), 0.325, implies an elasticity of demand in the average route of -2.05.”

The mean and median across routes of the implied change in MHHI delta is 91 HHI points.

“The mean and median across routes of the implied change is 91 HHI points; the implied change is larger than 100 HHI points in more than 2,000 routes; the largest implied increase is 281 HHI points.”

The largest implied increase in MHHI delta is 281 HHI points.

“The mean and median across routes of the implied change is 91 HHI points; the implied change is larger than 100 HHI points in more than 2,000 routes; the largest implied increase is 281 HHI points.”

Excluding quarters in which one of the major airlines was in bankruptcy yields estimates similar to the main specification.

“To more directly investigate the impact of bankruptcies on our estimates, in Table 4, specification (1), we exclude quarters in which one of the major airlines was in bankruptcy from the sample, retaining only the periods 2001Q1-2002Q2, 2007Q2-2011Q3, and 2014Q1-2014Q4.The estimates are similar to the main specification.”

We find a significant effect for routes within a range of HHI between around 2,500 and 6,500.

“Consistent with the above hypotheses, we find a significant effect for routes within a range of HHI between around 2,500 and 6,500.”

Only common ownership by shareholders ranked first and second has a positive and highly statistically significant effect on ticket prices.

“Only common ownership by shareholders ranked first and second has a positive and highly statistically significant effect on ticket prices.”

Only common ownership by long-horizon shareholders has a significant positive effect on prices.

“We find that only common ownership by long-horizon shareholders has a significant positive effect on prices, while common ownership by short-horizon shareholders yields coefficients of varying sign and weak (if any) significance.”

There is a positive and highly statistically significant effect of multi-market contact on ticket prices.

“Table 4, specification (7) shows that there is a positive and highly statistically significant effect of multi-market contact on ticket prices, measured as average route contact as in in Evans and Kessides (1994); see Ciliberto and Williams (2014) for a structural version.”

• cites Evans and Kessides (1994) sourceNot checked — no full text of the source
• cites Ciliberto and Williams (2014) sourceNot checked — no full text of the source

Using the discrete treatment variable as an instrument, the estimated effect using the post-period 2010-2014Q1 is 0.462.

“Table C.2 shows similar results using the discrete treatment variable as an instrument.The estimated effect using the post-period 2010-2014Q1 is 0.462.”

Using city pairs instead of airport pairs yields similar and somewhat stronger results.

“For example, city pairs may constitute a better market definition than airport pairs.Table 4, specification (3) shows that the results are similar, and in fact somewhat stronger.”

The effect of the MHHI delta on prices increases with market size.

“Indeed, the effect of the MHHI delta on prices increases with market size.”

MHHI delta calculated assigning zero control to top 10 shareholders and using only lower-ranked shareholders has no significant effect on ticket prices.

“Instead, as shown in Table C.6, we find that the MHHI delta calculated in this way has no significant effect on ticket prices.”

• AssumptionThe authors state this as an assumption: In the placebo test, only shareholders ranked below the top 10 are assumed to control the firm, with zero control assigned to the top 10 shareholders.

We have documented a statistical link between common ownership and higher prices.

“We have documented a statistical link between common ownership and higher prices.”

Using specifications (2) and (5) from the quantity and price regressions, the implied elasticity is -1.95.

“Using specifications (2) and ( 5) from the quantity and price regressions, respectively, the implied elasticity is -1.95.”

Controlling for a tenth-order polynomial in HHI does not significantly change the MHHI delta coefficient.

“Also, the functional form assumed by equation ( 2) is unlikely to drive the results: controlling for a tenth-order polynomial in HHI does not significantly change the MHHI delta coefficient (spec.(4)).”

Including all shareholders present in the Thomson database has a minimal effect on the estimated coefficients.

“Consistent with this idea, including all shareholders present in the Thomson database has a minimal effect on the estimated coefficients.”

• AssumptionThe authors state this as an assumption: Smaller shareholders have no significant say in corporate strategy.

The MHHI delta coefficient is insignificant in 2006 and 2007.

“The MHHI delta coefficient is insignificant in 2006 and 2007, possibly because both Delta Air Lines and Northwest were in bankruptcy during this time.”

The t-stats for CRCO are lower than those in the baseline specification for MHHI delta.

“Table 4 specification (6) shows that the effect of the carrier-route-level common ownership measure is positive and statistically significant at the 1% level, but the t-stats are lower than those in the baseline specification for the MHHI delta, as expected.”

Common ownership and control by shareholders ranked 3 to 10 has a small and insignificant effect on ticket prices.

“Common ownership and control by shareholders ranked 3 to 10 has a small and insignificant effect on ticket prices.”

Specifications (3) and (6) imply an elasticity of -1.05.

“Specifications (3) and ( 6) imply an elasticity of -1.05.”

In the first-stage IV regressions using continuous treatment, the implied change in MHHI delta is positive and statistically significant in all post-period specifications (2010Q1, 2011Q1, 2012Q1, 2013Q1, 2014Q1, and 2010-2014 Q1).

“Implied Change in MHHI delta 6.416*** 3.104*** 5.379*** 4.101*** 5.161*** 4.832*** (0.360) (0.450) (0.477) (0.472) (0.667) (0.432)”

• AssumptionRests on an assumption: The pre-period 2009Q1 is before the Barclays BGI acquisition by BlackRock was announced.
• AssumptionRests on an assumption: The counterfactual MHHI delta combining the holdings of Barclays and BlackRock represents the change in common ownership caused by the acquisition.

In panel regressions for bankruptcy markets, MHHI delta has a positive and statistically significant effect on log average fare.

“MHHI delta 0.0383 0.156*** 0.134*** 0.177*** 0.188*** 0.252*** (0.0711) (0.0404) (0.0350) (0.0578) (0.0412) (0.0488)”

In panel regressions for non-bankrupt carriers, MHHI delta has a positive and statistically significant effect on log average fare.

“MHHI delta 0.0383 0.156*** 0.134*** 0.177*** 0.188*** 0.252*** (0.0711) (0.0404) (0.0350) (0.0578) (0.0412) (0.0488)”

Route-level capacity decisions are a frequent topic of conversation in public earnings calls of U.S. airlines.

“We find that route-level capacity decisions are a frequent topic of conversation.”

Common ownership concentration for the average route is more than ten times larger than what would be presumed likely to enhance market power in a traditional merger, according to the US Antitrust Agencies' Horizontal Merger Guidelines.

“Common ownership concentration for the average route is more than ten times larger than what would be presumed”

Carrier-level MHHI delta has a positive and statistically significant effect on log average fare.

“Carrier-Level MHHI delta 0.983*** 0.910*** 0.996*** (0.205) (0.185) (0.184)”

Market-level MHHI delta has a positive and statistically significant effect on log average fare.

“MHHI delta 0.140*** 0.172*** 0.0947*** 0.199*** 0.212*** 0.140*** (0.0425) (0.0374) (0.0337) (0.0460) (0.0394) (0.0385)”

In the US airline industry, a modified index of market concentration that accounts for common ownership indicates levels of market concentration far exceeding those indicated by the conventional measure of market concentration.

“In particular, in the US airline industry, a modified index of market concentration that takes into account to which extent competitors are commonly owned by the same investors indicates levels of market concentration that far exceed those indicated by the conventional measure of market concentration.”

Average carrier-level MHHI delta of competitors has no statistically significant effect on log average fare.

“Average Carrier-Level MHHI delta of Competitors 0.0740 0.172 0.0576 (0.153) (0.151) (0.160)”

In panel regressions for non-bankruptcy markets, MHHI delta has a positive and statistically significant effect on log average fare.

“MHHI delta 0.0383 0.156*** 0.134*** 0.177*** 0.188*** 0.252*** (0.0711) (0.0404) (0.0350) (0.0578) (0.0412) (0.0488)”

In panel regressions, MHHI delta has a negative and statistically significant effect on log market passengers.

“MHHI delta -0.665*** -0.607*** -0.213*** (0.0876) (0.0824) (0.0527)”

The absence of significant effects for shareholders unlikely to have control is inconsistent with the hypothesis that endogeneity of market shares drives the main results.

“is inconsistent with the hypothesis that the endogeneity of market shares drives the main results.”

On average, common ownership adds about as much concentration as going from four roughly equal-sized carriers to two equal-sized carriers would add.

“On average common ownership adds about as much concentration as going from four roughly equal-sized carriers to two equal-sized carriers would add.”

Since airline stocks constituted only a small fraction of the merging parties' portfolios, it is unlikely that this variation is driven by expected changes in US airline ticket prices.

“Since airline stocks constituted only a small fraction of the merging parties' portfolios, it is unlikely that this variation is driven by expected changes in US airline ticket prices.”

These findings are consistent with a model of rational attention allocation by investors and airlines to markets where the bottom-line impact of increased prices is greater.

“These findings are consistent with a model of rational attention allocation by investors and airlines to markets where the bottom-line impact of increased prices is greater.”

The omission to explicitly demand or incentivize tougher competition between portfolio firms may allow managers to enjoy a quiet life and thus cause an equilibrium with reduced competition and sustained high margins.

“the omission to explicitly demand or incentivize tougher competition between portfolio firms may allow managers to enjoy a”

• cites Hicks, 1935 sourceNot checked — no full text of the source
• cites Bertrand and Mullainathan, 2003 sourceNot checked — no full text of the source

• AssumptionThe authors state this as an assumption: Wells Fargo's CEO understands that it is not in his largest shareholders' best interest to compete aggressively for market share against Bank of America, even without explicit communication.
• AssumptionRests on an assumption: Top management teams of the largest US airlines learned about major shifts in the ownership structure of the industry even without being personally informed by Warren Buffett.
• AssumptionRests on an assumption: Bank of America and JP Morgan are well-informed about these interests.

Common owners of competitors need not explicitly communicate their anticompetitive interests to management for the documented outcomes to materialize.

“common owners of competitors need not explicitly communicate their anticompetitive interests to management for the documented outcomes to materialize.”

• AssumptionThe authors state this as an assumption: Wells Fargo's CEO understands that it is not in his largest shareholders' best interest to compete aggressively for market share against Bank of America, even without explicit communication.
• AssumptionRests on an assumption: Top management teams of the largest US airlines learned about major shifts in the ownership structure of the industry even without being personally informed by Warren Buffett.
• AssumptionRests on an assumption: Bank of America and JP Morgan are well-informed about these interests.

The endogeneity of market shares does not drive the results.

“This finding is consistent with the hypothesis that the endogeneity of market shares does not drive our results.”

Common ownership links can have the effect of blurring formal firm boundaries.

“Our results suggest that common ownership links can have the effect of blurring formal firm boundaries.”

The incentives for anti-competitive behavior implied by current levels of common ownership, as measured by MHHI delta, are an order of magnitude larger than the implications for market power recognized by conventional measures on the same scale.

“In sum, the incentives for anti-competitive behavior implied by current levels of common ownership, as measured by the MHHI delta, are an order of magnitude larger than the implications for market power recognized by conventional measures that are measured on the same scale.”

The negative relation between passenger volume and common ownership indicates that the price effects are not driven by increased demand that institutional shareholders correctly foresee.

“indicating that the price effects are not driven by increased demand that institutional shareholders correctly foresee (a reverse-causality argument): if increases in common ownership were caused by increased demand, changes in common ownership should be correlated with higher, not lower quantity.”

The paper is the first to empirically identify an effect of common ownership on product prices.

“To our knowledge our paper is the first to empirically identify an effect of common ownership on product prices”

The difference-in-differences identification strategy uses only variation in common ownership across routes implied by the hypothetical combination of the two parties' portfolios as of the quarter before the announcement of the acquisition.

“This identification strategy uses only variation in common ownership across routes that is implied by the hypothetical combination of the two parties' portfolios as of the quarter before the announcement of the acquisition.”

The estimates from the difference-in-differences strategy are arguably less affected by endogeneity of ownership and market shares.

“While estimated using much less variation than the panel regressions, the estimates from this strategy are arguably less affected by endogeneity of ownership and market shares.”

An investor holding equal-sized stakes in two competitors enjoys greater total portfolio profits when the firms set prices or quantities as if they were two divisions of a monopoly, instead of acting like two independent firms.

“That means an investor holding equal-sized stakes in both A and B enjoys greater total (i.e.portfolio) profits when the firms set prices or quantities as if they were two divisions of a monopoly, instead of acting like two independent firms.”

Fixed effects difference out many alternative interpretations at the firm-, route-, firm-route, or firm-time level, such as confounding effects of fuel or oil price changes.

“Fixed effects difference out many (but of course not all) alternative interpretations at the firm-, route-, firm-route, or firm-time level, such as confounding effects of fuel or oil price changes.”

Market shares are potentially endogenous in ways likely to negatively bias the MHHI delta coefficient.

“When interpreting the MHHI delta coefficient (β), one should keep mind that market shares (which enter both MHHI delta and HHI calculations) are potentially endogenous in ways that are likely to negatively bias the MHHI delta coefficient.”

Using the weighted average MHHI delta across markets of 2,044 in 2014Q4, the coefficient of around -.2 from the most saturated specification indicates about 4% less market passengers in the average route than under separate ownership.

“Using the weighted average MHHI delta across markets of 2,044 in 2014Q4 to gauge economic signficance, the coefficient of around -.2 from the most saturated specification indicates that there are about 4% less market passengers in the average route than there would be under separate ownership.”

The history of Barclays' attempt to sell iShares suggests the divestment decision was not primarily driven by considerations regarding how the iShares portfolio would combine with BlackRock's in terms of potential product market effects.

“The history of Barclays' attempt to sell iShares to investors other than BlackRock suggests the divestment decision was not primarily driven by considerations regarding how the iShares portfolio would combine with BlackRock's in terms of potential product market effects.”

Freeman (2016) finds an effect of common ownership on the longevity of customer-supplier relationships.

“Freeman (2016) finds an effect of common ownership on the longevity of customer-supplier relationships, which corroborates the assumption of the various theoretical models motivating our empirical analysis that firms internalize externalities imposed on commonly owned firms.”

• cites Freeman (2016) sourceNot checked — no full text of the source

Only economically meaningful ways of calculating common ownership concentration are linked to significant product market outcomes.

“Rather, only economically meaningful ways of calculating common ownership concentration are linked to significant product market outcomes.”

The sign of the effect based on variation in common ownership generated by the BGI acquisition is consistent with previous results from the panel regressions.

“Thus, the sign of the effect, based on variation in common ownership generated by the BGI acquisition is consistent with our previous results from the panel regressions.”

The results are inconsistent with the hypothesis that various forms of misspecification mechanically drive a spurious correlation between MHHI delta and ticket prices.

“These results are consistent with standard notions of corporate governance, but inconsistent with the hypothesis that various forms of misspecification mechanically drive a spurious correlation between MHHI delta and ticket prices.”

The results are consistent with the predictions of increased market power.

“Instead, the results are consistent with the predictions of increased market power.”

In a Cournot model, if a match between lazy principals and lazy agents becomes pervasive in an industry, industry output declines and margins increase.

“If a match between lazy principals and lazy agents becomes pervasive in an industry, industry output declines and margins increase in a Cournot model.”

The variation across portfolio firms in the degree to which the BlackRock-BGI changed their ownership structure translates into variation across airline routes, because different combinations of airlines compete in different routes.

“The variation across portfolio firms in the degree to which the BlackRock-BGI changed their ownership structure translates into variation across airline routes, because different combinations of airlines compete in different routes.”

The loss of power is consistent with the data favoring a Cournot model over Bertrand as the relevant model of competition in airlines.

“The loss of power is also consistent with the data favoring a Cournot model over Bertrand as the relevant model of competition in airlines, as previously argued by Brander and Zhang (1990) and the literature that followed.”

• cites Brander and Zhang (1990) sourceNot checked — no full text of the source

The proportional control assumption is not driving the baseline results.

“The results are similar in magnitude to the baseline, consistent with the idea that the proportional control assumption is not driving the baseline results.”

• AssumptionThe authors state this as an assumption: Shareholder control is proportional to the number of shares they own.

When various industry competitors are owned by concentrated activists that push their targets to compete aggressively, a competitive outcome obtains.

“When various industry competitors are owned by concentrated activists that push their targets to compete aggressively, a competitive outcome obtains.”

Our empirical tests do not rely on the assumption that airlines necessarily compete à la Cournot.

“This feature does not imply that our empirical tests rely on the assumption that airlines necessarily compete à la Cournot, as we now show.”

Campaigns by activist investors lead to increases in market share of the target at the expense of its rivals.

“By contrast, it is well documented that campaigns by activist investors, which typically concentrate their capital in one target firm per industry, lead to increases in market share of the target at the expense of its rivals (e.g.Aslan and Kumar, 2016).”

• cites Aslan and Kumar, 2016 sourceNot checked — no full text of the source

Multiplying the estimates with the average MHHI delta across routes implies ticket prices are about 10%-12% higher because of common ownership alone, compared to a counterfactual world in which firms are separately owned or ignore anti-competitive incentives.

“Multiplying the estimates with the average MHHI delta across routes would imply that ticket prices are about 10%-12% higher because of common ownership alone, compared to a counterfactual world in which firms are separately owned, or in which firms ignore the anti-competitive incentives of their shareholders.”

Omitting common ownership from regressions of price on industry concentration (HHI) leads to a negative omitted variable bias on the HHI coefficient.

“That paper also shows that omitting common ownership from regressions of price on industry concentration (HHI) leads to a negative omitted variable bias on the HHI coefficient.”

• cites Azar, Raina, and Schmalz (2016) sourceNot checked — no full text of the source

Gutiérrez and Philippon (2017) find that in a broad cross-section of US firms, common ownership is negatively related to firms' propensity to invest amid high profitability and Q.

“Gutiérrez and Philippon (2017) find that in a broad cross-section of US firms, common ownership is negatively related to firms' propensity to invest amid high profitability and Q.”

• cites Gutiérrez and Philippon (2017)Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

MHHI delta does not appear to have a mechanical relationship to prices.

“These findings serve as yet another placebo test: MHHI delta does not appear to have a mechanical relationship to prices.”

This range of elasticity estimates is similar to that reported in the existing literature (-1.37 to -2.01 (Berry and Jia, 2010); -1.4 (IATA, 2008)).

“This range of estimates is similar to that reported in the existing literature (-1.37 to -2.01 (Berry and Jia, 2010); -1.4 (IATA, 2008)).”

• cites Berry and Jia, 2010 sourceNot checked — no full text of the source
• cites IATA, 2008Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

This finding ensures that the main results are not driven by small markets.

“This finding not only corroborates the basic narrative above, but also ensures that the main results are not driven by a small markets.”

The majority of the effect of common ownership on competitive strategy may be implemented at the firm-pair level, or even at the firm level.

“The majority of the effect of common ownership on competitive strategy may be implemented at the firm-pair level, or even at the firm level.”

Large diversified mutual fund families are not expected to actively push for more aggressive product market behavior between portfolio firms.

“One should of course not expect that large diversified mutual fund families actively push for more aggressive product market behavior between portfolio firms, given that doing so would not only be costly, but also go against the incentive to maximize the value of the family's total portfolio.”

Powerful diversified shareholders have little incentive to intervene in that case.

“The powerful diversified shareholders continue to have little incentive to intervene in that case.”

Azar, Raina, and Schmalz (2016) show that the combination of common ownership and cross-ownership of US banks cause higher prices of retail deposit products.

“In particular, Azar, Raina, and Schmalz (2016) show that the combination of common ownership and cross-ownership of US banks cause higher prices of retail deposit products, using within-bank variation across branches over time and variation in county-level common ownership that results from index fund growth as an instrument.”

• cites Azar, Raina, and Schmalz (2016) sourceNot checked — no full text of the source

Because the paper employs carrier-route fixed effects, market power on specific routes exerted through frequent-flyer programs is differenced out.

“For example, because we employ carrier-routefixed effects, market power on specific routes exerted through frequent-flyer programs (Lederman, 2007) is differenced out in our regressions.”

• cites Lederman, 2007 sourceNot checked — no full text of the source

There are large anti-competitive incentives due to common ownership links at the market level.

“This paper presents evidence of large anti-competitive incentives due to common ownership links at the market level”

Consolidation in the asset management industry can adversely affect competition in the product markets of their portfolio companies.

“regulators should keep in mind that consolidation in the asset management industry can adversely affect competition in the product markets of their portfolio companies.”

The common ownership problem is unlikely to be solved by shutting down a particular channel.

“This insight is important, as it suggests that the common ownership problem is unlikely to be solved by shutting down a particular channel.”

Passive institutions have the power to shape managerial incentives.

“Yet, they have the power to shape managerial incentives.”

Anecdotal evidence exists that investors and portfolio firms discuss product market strategy, sometimes even at the market level.

“We conclude that anecdotal evidence exists that investors and portfolio firms discuss product market strategy, sometimes even at the market level.”

Exploiting variation in common ownership concentration generated by the merger of two large asset managers, product prices are 10-12% higher due to common ownership.

“We then exploit variation in common ownership concentration generated by the merger of two large asset managers that arguably occurred for reasons unrelated to expected route-level differences in US airline ticket prices, and find that product prices are 10-12% higher due to common ownership.”

• AssumptionRests on an assumption: The merger of two large asset managers occurred for reasons unrelated to expected route-level differences in US airline ticket prices.
• AssumptionRests on an assumption: The control variables included in the regressions are sufficient to address confounding from the listed events and other factors.

One should expect a lessening of competition, even when the diversified owners do nothing at all to actively reduce the competitiveness of their portfolio firms' product markets.

“One should thus expect a lessening of competition, even when the diversified owners do nothing at all to actively reduce the competitiveness of their portfolio firms' product markets.”

• AssumptionRests on an assumption: The past three decades have witnessed a shift from the low-common-ownership to the high-common-ownership equilibrium, because diversified institutions have increasingly crowded out concentrated owners as the most powerful shareholders of firms.

A lack of relative performance incentives gives managers reduced incentives to compete.

“A lack of relative performance incentives gives managers reduced incentives to compete.”

Antón, Ederer, Giné, and Schmalz (2016) show that common owners have incentives to reduce relative performance incentives, either actively or by omitting to support other shareholders' initiatives to implement relative performance incentives.

“Antón, Ederer, Giné, and Schmalz (2016) show that common owners have incentives to reduce relative performance incentives, either actively or by omitting to support other shareholders' initiatives to implement relative performance incentives.”

• cites Antón, Ederer, Giné, and Schmalz (2016) sourceNot checked — no full text of the source

The omission on behalf of large diversified mutual fund families to push portfolio firms to compete aggressively against each other can implement the outcomes we empirically document.

“In sum, the omission on behalf of large diversified mutual fund families to push portfolio firms to compete aggressively against each other can implement the outcomes we empirically document.”

Compensation contracts can be used to align managers' strategic incentives with those of shareholders.

“We conclude that compensation contracts can be used to align managers' strategic incentives with those of shareholders.”

The strategic, longer-term pricing patterns studied in the paper are implied by capacity pre-commitments.

“the strategic, longer-term pricing patterns we study are implied by capacity pre-commitments”

• AssumptionRests on an assumption: Capacity pre-commitments typically happen about one year before the flight.

The paper's results make an empirical case for the theoretical insight that shareholders do not agree on profit maximization as an objective when firms are not price takers.

“Our results can be viewed as challenging this assumption, and thus make an empirical case for taking seriously the theoretical insight (perhaps most clearly stated in Hart, 1979) that shareholders do not agree on profit maximization as an objective when firms are not price takers.”

• cites Hart, 1979 sourceNot checked — no full text of the source

Berkshire Hathaway's Co-CIO would reduce the Berkshire portfolio value if he used his role as a JP Morgan board member to propose a particularly aggressive competitive strategy against American Express, Bank of America, or Wells Fargo, in all of which Berkshire Hathaway is a major or the largest investor.

“For example, Berkshire Hathaway's Co-CIO would reduce the Berkshire portfolio value if he used his role as a JP Morgan board member to propose a particularly aggressive competitive strategy against American Express, Bank of America, or Wells Fargo, in all of which Berkshire Hathaway is a major or the largest investor (Buhayar, 2016).”

• cites Buhayar, 2016Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Financial choices can influence firms' product market strategy.

“It is well known that financial choices can influence firms' product market strategy (Brander and Lewis, 1986;Chevalier, 1995;Kovenock and Phillips, 1995;Phillips, 1995;Dasgupta and Titman, 1998).”

• cites Brander and Lewis, 1986Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

• cites Chevalier, 1995Not checked — no full text of the source
• cites Kovenock and Phillips, 1995Not checked — no full text of the source

When firms implement shareholders' incentives, and all shareholders (including those with significant control) are fully diversified, product market competition will tend towards monopolistic outcomes, with an associated deadweight loss for the economy.

“When firms implement shareholders' incentives, and all shareholders (including those with significant control) are fully diversified, product market competition will tend towards monopolistic outcomes, with an associated deadweight loss for the economy.”

Managers are unlikely to be isolated from common owners' anti-competitive incentives when regulators prohibit communication about subjects of competition, while conversations about pay structure or voting on board members continue to be permitted.

“For example, managers are unlikely to be isolated from common owners' anti-competitive incentives when regulators prohibit communication about subjects of competition, while conversations about pay structure or voting on board members continue to be permitted.”

The three goals of perfect shareholder diversification, firms' maximization of shareholder interests ('good governance'), and preservation of competitive product markets cannot be simultaneously achieved.

“The three goals of (i) perfect shareholder diversification, (ii) firms' maximization of shareholder interests (”

• cites Azar, 2012 sourceNot checked — no full text of the source

Actively managed funds can threaten management with selling the stock in case management does not adhere to their desired product market strategy, which may explicitly feature not entering competitors' markets.

“Actively managed funds can threaten management with selling the stock in case management does not adhere to their desired product market strategy, which may explicitly feature not entering competitors' markets (Reuters, 2008).”

• cites Reuters, 2008Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

Wealth-performance-sensitivities decline with common ownership.

“Fos and Tsoutsoura (2014); Aggarwal, Dahiya, and Prabhala (2015) show commonly owned, and show that wealth-performance-sensitivities decline with common ownership.”

• cites Fos and Tsoutsoura (2014) sourceNot checked — no full text of the source
• cites Aggarwal, Dahiya, and Prabhala (2015) sourceNot checked — no full text of the source

Liang (2016) independently corroborates that pay-performance-sensitivities decline with common ownership.

“Liang (2016) independently corroborates that pay-performance-sensitivities decline with common ownership, using firm-level variation, alternative functional forms, and a different identification strategy.”

• cites Liang (2016)Not matched

  Couldn’t be matched to a paper in Crossref, so it can’t be checked.

The feature of capacity pre-commitments and the previous literature suggest a Cournot model.

“This feature, as well as the previous literature(Brander and Zhang, 1990), suggest a Cournot model.”

• cites Brander and Zhang, 1990 sourceNot checked — no full text of the source

• AssumptionRests on an assumption: Capacity pre-commitments typically happen about one year before the flight.

Using fourteen years of market-firm-level quarterly panel data, airline ticket prices are 3-7% higher because of common ownership, compared to a counterfactual world in which firms are separately owned, or in which firms entirely ignore their owners' anti-competitive incentives caused by common ownership.

“Specifically, using fourteen years of market-firm-level quarterly panel data, we find that airline ticket prices are 3-7% higher because of common ownership, compared to a counterfactual world in which firms are separately owned, or in which firms entirely ignore their owners' anti-competitive incentives caused by common ownership.”

• AssumptionRests on an assumption: The counterfactual world in which firms are separately owned, or in which firms entirely ignore their owners' anti-competitive incentives caused by common ownership, is a valid comparison for estimating the effect of common ownership on airline ticket prices.
• AssumptionRests on an assumption: The control variables included in the regressions are sufficient to address confounding from the listed events and other factors.

It is unlikely that one single mechanism is solely responsible for translating shareholder incentives into firm behavior.

“Because there are many plausible channels through which shareholder incentives can translate into firm behavior, we find it unlikely that one single mechanism is solely responsible.”

The case illustrates the silent shift of power from concentrated to diversified investors.

“The case thus illustrates the silent shift of power from concentrated to diversified investors.”

None of the prior empirical studies addresses the question whether competition is affected by common ownership.

“None empirically addresses the question whether competition is affected.”

Engagement in corporate governance merely exacerbates the problem.

“Engagement in corporate governance merely exacerbates the problem.”

Any influence of ownership by large passive institutional investors on portfolio firm's capital structure or payout decisions can affect the product market equilibrium.

“Hence, any influence of ownership by large passive institutional investors on portfolio firm's capital structure or payout decisions can affect the product market equilibrium.”