Are conglomerate firms highly productive? Or do they suffer from internal quarrels and squander investment opportunities? Somewhat astonishingly, the answer to both questions is yes, depending on the literature one reads. On the one hand, a well-established observation in the trade literature is that conglomerate firms are more productive than single-product firms. They account for two-thirds of US exporters and for over 90% of export value and manufacturing sales (Bernard et al. 2018). On the other hand, these facts appear to be at odds with findings in the finance literature: multi-segment firms trade at a discount and have a lower Tobin’s Q than single-product firms, because internal capital markets misallocate funds across divisions within firms (Ozbas and Scharfstein 2009).
In a new study, we reconcile these two conflicting views and develop a novel theory of misallocation within firms due to managers’ empire building (Doerr et al. 2020). We introduce an internal capital market into a two-factor model of multi-segment firms. The combination of both features allows us to investigate the interplay between internal capital markets and international trade, and open the black-box of multi-product firms. The key insights we gain are two-fold. First, we find that informational frictions between headquarters and divisional managers can lead to the over-reporting of divisions’ costs and hence an inefficient allocation of capital across divisions within a firm. This misallocation results in the conglomerate discount. Second, our model shows that international trade reduces capital misallocation through competition. Tougher competition lowers the cost level at which firms and their divisions can survive in the market and reduces managers’ scope of misreporting. Consequently, it reduces capital misallocation within firms and the conglomerate discount.
In this way, our model explains why more open markets impose discipline on the allocation of capital within firms, and why exporters exhibit a lower conglomerate discount than non-exporters (a fact that we establish, see Figure 1). We test our model’s key predictions with data on US companies. By exploiting exogenous variation in import competition from China at the sectoral level, we establish that the conglomerate discount declines more in industries that are subject to a stronger increase in import competition. The decline in the discount is caused by a fall in marginal costs and a rise in allocated capital to the best segments, confirming the model’s predictions. Crucially – and in line with our model – allocation improves by more in firms that are subject to more severe informational frictions, which suggests that information asymmetries are an important source of within-firm capital misallocation.
Figure 1 The conglomerate discount over time, by exporter status
Notes: This figure plots the ratio of average Q of multi-segment firms over average Q of single-segment firms (the conglomerate discount) for each year in our sample. We split the sample into exporting (blue solid line) and non-exporting (black dashed line) firms.
International trade and the internal capital market
We begin our analysis by developing a theoretical model that embeds an internal capital market into a model of multi-product firms with monopolistic competition. In the model, managers of multi-product firms compete for funds within their firms. To model the internal capital market, we combine the concepts of ‘winner picking’ by Stein (1997) and ‘over-investment’ by Rajan et al. (2000) and Scharfstein and Stein (2000). However, there are informational frictions between the headquarters and the divisional managers that lead to a distorted allocation of capital across divisions. The headquarters does not know the true marginal costs of its divisions, while the divisional managers do. These frictions allow managers, who have a desire to run bigger divisions, to misreport their true marginal costs to receive more capital. Importantly, managers in better divisions have more room for misreporting, while the worst divisions will not be financed. This results in a distorted allocation of capital: headquarters over-allocate capital to the best divisions, which reduces firms’ return on assets and depresses their Tobin’s Q, resulting in a conglomerate discount.
We then introduce competition through international trade into our model and investigate how it affects the allocation of capital within firms. To do so, we embed the internal capital market into a two-factor version of the monopolistic competition model of multi-product firms in Mayer et al. (2014). Competition lowers the cost level at which firms and their divisions can survive in the market. Managers use this cut-off cost level as a benchmark when deciding by how much to misreport costs – if their divisions appear to inefficient, they might not be allocated any capital. Competition hence has a disciplining effect and reduces the scope for over-reporting – and this effect is particularly pronounced in the best divisions, i.e. those in which managers have the greatest scope for overreporting. Our model thus predicts that competition leads to a re-allocation of capital within multi-segment firms, and that the allocation improves the most in firms with more severe informational frictions. Better allocation of capital consequently increases firms’ profitability and Q, and hence reduces the conglomerate discount.
Empirical evidence: The China shock
We test the key predictions of our model in the empirical part of the paper. Based on our model, we derive four central predictions. Import competition leads to: (i) a fall in the cut-off cost and hence lower over-reporting, (ii) a fall in relative marginal costs, (iii) an increase in allocated assets for the best segments in the same industry, and (iv) the induced fall in marginal costs and increase in allocated assets imply that competition reduces the conglomerate discount. To test these predictions, we exploit the increase in import competition from China as a source of exogenous variation in industry-level competition (Autor et al. 2013). Using detailed data on US manufacturing firms at the segment level from 1999–2007, we confirm the model’s key predictions: we first show that the conglomerate discount declines more in industries that are subject to a stronger increase in import competition, as can been seen in Figure 2. It plots the average conglomerate discount over time from 1999–2007, where we split the sample into industries with a strong increase in competition (blue line, high competition), and industries with a modest increase in competition (black-dashed line, low competition). While the conglomerate discount decreases for industries with a strong rise in competition – that is, there is an increase in multi-segment firms’ Q relative to single-segment firms – there is no change for industries that saw little-to-no change in competition. Our regressions results indicate that a one standard deviation increase in import penetration lowers the average conglomerate discount by 32%, or almost twice the mean change in the conglomerate discount over the sample period.
Figure 2 The conglomerate discount over time, by Chinese import competition
Notes: This figure plots the ratio of average Q of multi-segment firms over average Q of single-segment firms (the conglomerate discount) for the time period covered in our ‘China shock’ sample. We split the sample into firms within industries with an above-median increase in import penetration from China (blue solid line) and those with a below-median increase (black dashed line) firms.
We then investigate the underling channels through which competition reduces the conglomerate discount. We first show that the decline in conglomerate discount is caused by a fall in marginal costs of and a rise in allocated capital to the best segments – confirming the model’s predictions. In our model, the underlying friction that gives rise to misallocation is asymmetric information. We thus expect the disciplining effects of import competition on segment marginal costs and assets to be particularly strong within firms that suffer more from informational asymmetries. Therefore, we use data on CEO tenure provided by BoardEx. We classify firms into those with high and low informational frictions by the average time that the CEOs spent on the board. The longer the time a CEO has served on the board of a company, the better she knows its segments and the lower the scope for over-reporting. We find that the disciplining effects of competition (i.e. a decrease in marginal costs and an increase in allocated assets to the best segments) to be particularly strong in firms subject to higher informational frictions. A one standard deviation increase in import competition from China reduces the marginal costs of better segments by approximately 5%. The same increase in imports increases assets allocated to the best segments by around 25%.
Based on our model, we can use these estimates to infer the change in the over-reporting factor (i.e. by how much divisional managers adjust overstating true costs). We find that for a one standard deviation increase in imports from China the over-reporting factor declines by 15% in the best segments, relative to the worst segments. Taken together, our empirical results provide strong support for our main predictions. They suggest that competition causally reduces within-firm misallocation of capital and reduces the conglomerate discount by around 30%.
Our model offers a novel theory of misallocation within firms (rather than between firms) and shows that international trade can discipline firms’ internal capital markets through tougher competition. Our analysis further provides strong empirical evidence for the model’s central predictions: tougher competition reduces capital misallocation across divisions, especially in firms with more severe informational frictions.
In our paper, we hence reconcile the two conflicting findings on the efficiency of conglomerate firms in the finance and trade literature.1 We thereby relate to literature on capital misallocation. Several papers identify financial constraints as a reason why the marginal products of capital are not equated across firms.2 However, Kehrig and Vincent (2017) note that a sizeable share of overall misallocation occurs within, rather than between, firms. In our paper, misallocation of capital within firms arises due to an information asymmetry between headquarters and managers. Importantly, this misallocation within firms arises even in the absence of misallocation of capital across firms. We further provide a link from changes in competition to the allocation of capital within firms and can thereby explain why exporting conglomerates do not suffer from a discount.
Autor, D H, D Dorn and G H Hanson (2013), “The China syndrome: Local labor market effects of import competition in the United States”, American Economic Review 103(6): 2121–2168.
Bernard, A B, J B Jensen, S J Redding and P K Schott (2018), “Global Firms”, Journal of Economic Literature 56(2): 565–619.
Caliendo, L and E Rossi-Hansberg (2012), “The Impact of Trade on Organization and Productivity”, The Quarterly Journal of Economics 127(3): 1393–1467.
Doerr, S (2018), “Collateral, reallocation, and aggregate productivity: Evidence from the US housing boom”, Working Paper.
Doerr, S, D Marin, D Suverato and T Verdier (2020), “Mis-allocation Within Firms: Internal Finance and International Trade”, CEPR Discussion Paper no. 144802.
Gopinath, G, S Kalemli-Ozcan, L Karabarbounis and C Villegas-Sanchez (2017), “Capital allocation and productivity in south Europe”, The Quarterly Journal of Economics 132(4): 1915–1967.
Hsieh, C T and P J Klenow (2009), “Misallocation and Manufacturing TFP in China and India”, The Quarterly Journal of Economics 124(4): 1403-1448.
Kehrig, M and N Vincent (2017), “Do Firms Mitigate or Magnify Capital Misallocation? Evidence from Plant-Level Data”, CESifo Working Paper Series No. 6401.
Marin, D and T Verdier (2008), “Power Inside The Firm and The Market: A General Equilibrium Approach”, Journal of the European Economic Association 6(4): 752–788.
Marin, D and T Verdier (2014), “Corporate Hierarchies and International Trade. Theory and Evidence”, Journal of International Economics 94(2): 295-310.
Mayer, T, M J Melitz and G I P Ottaviano (2014), “Market Size, Competition, and the Product Mix of Exporters”, American Economic Review 104(2): 495– 536.
Mueller, H M (2016), “Reallocation of Capital and Labor within Firms”, Swiss Journal of Economics and Statistics (SJES) 152(4): 289–303.
Ozbas, O and D S Scharfstein (2009), “Evidence on the Dark Side of Internal Capital Markets”, Review of Financial Studies 23(2): 581–599.
Rajan, R, H Servaes and L Zingales (2000), “The Cost of Diversity: The Diversification Discount and Inefficient Investment”, The Journal of Finance 55(1): 35–80.
Scharfstein, D S and J C Stein (2000), “The Dark Side of Internal Capital Markets: Divisional Rent-Seeking and Inefficient Investment”, Journal of Finance 55(6): 2537–2564.
Stein, J C (1997), “Internal Capital Markets and the Competition for Corporate Resources”, Journal of Finance 52(1): 111–33.
1 See Mueller (2016) for a review of the literature on internal capital markets. Workhorse models in the trade literature usually abstract from financial issues and assume a firms’ cost structure to be exogenous. For exceptions, see Marin and Verdier (2008, 2014) and Caliendo and Rossi-Hansberg (2012).
2 See Hsieh and Klenow (2009), Gopinath et al. (2017), or Doerr (2018).