How Distorting are Stock Market Valuations Really to Corporate Investment?

Transcription

1 Houdou BASSE MAMA Postdoctoral Research Associate School of Business, Economics and Social Sciences nstitute of Capital Markets & Management How Distorting are Stock Market Valuations Really to Corporate nvestment? National Taiwan University nternational Conference on Finance Session #3 [3:4-5:] Taipei, December 6, 22 Houdou Basse Mama

2 AGENDA Research Questions and Hypotheses Estimated Generic nvestment Equation Measuring Stock Price nformativeness & Misvaluation Results Conclusions Houdou Basse Mama 2

3 Research Questions RQ a. Should/Do firm managers follow the signals transmitted by the stock market when they make investment spending decisions? RQ b. Does stock price informativeness moderate or enhance managerial propensity to listen to the market? RQ 2. Do irrational gyrations in stock prices (misvaluations) color business fixed investment decisions? f they do, what is the transmission path? Houdou Basse Mama 3

4 The Learning Hypothesis: Two Contrasting Normative Views [RQ a b ] () Stock markets are just a sideshow Conventional wisdom reduces the stock market to a sideshow nsiders are privy to superior information about firm s genuine investment opportunities ( passive informant hypothesis ) o Morck et al. (99) nsiders are concerned about the firm s long term prospects when making investment decisions mplications Firm managers should simply ignore stock market valuations and invest until the marginal product of capital converges to the risk free rate (Bosworth, 975) Stock prices may embed a speculative bubble ( Animal spirits, rrational exuberance ) Critique: Shareholders are not typically locked in and there are new issues whose floatation price will directly affect investment A moderate view is that managers should respond to stock market valuations to the extent they affect firm s fundamentals Houdou Basse Mama 4

5 The Learning Hypothesis: Two Contrasting Normative Views [RQ a b ] (2) St. prices enhance investment decisions Stock markets generate and aggregate price sensitive serendipitous information collected from outside investors (Chen et al., 27; Subrahmanyam and Titman, 999) nformation contained in stock prices may relate to: demand side of firm s products or competition of the firm with rivals in the relevant product market Whether firms later consummate an M&A deal depends on market reaction to news mplications Firms should follow market valuations even though these might not coincide with own assessment of fundamentals (Fischer and Merton, 984; von Furstenberg, 977) Critique: Adverse signalling effects may prick a bubble (Myers and Majluf, 984); a pure wealth transfer Horizon is key (Blanchard et al., 993) Short horizon: follow the market Long term horizon: ignore the market Monetary Policy: Target the stock market Houdou Basse Mama 5

6 The Learning Hypothesis: Empirical Literature [RQ a b ] The related empirical evidence somewhat reflects the theoretical disagreement : + ± There is empirical evidence attesting to an independent (e.g., Jiang et al., 2; Chririnko and Schaller, 2; Chari and Henri, 28; Panageas, 25; Wongbangpo and Sharma, 22; Galeotti and Schiantarelli, 994; Barro, 99) and informative function of the stock market for firm investment decisions (e.g., Bakke and Whited, 2; Chen et al., 27; Luo, 25) and for the conduct of monetary policy (Rigobon and Sack, 23) Though the stock market is not a pure sideshow, it does not appear to be central either (Durham, 2; Samuel, 996; Blanchard et al., 993; Porter, 992; Morck et al., 99) n contrast, Wang et al. (29) find that the Chinese stock market does not affect firm investment through its information channel; the same result is reported by Li et al. (2) for Australia, and Bolbol and Omran (25) for five emerging Arab countries o This is because stock prices contain little information about future operating performance Houdou Basse Mama 6

7 Stock mispricing is driven by investor sentiment (Grundy and Li, 2): Houdou Basse Mama 7 C R P N C R P N N P C R P C R R R N u i u i ; ~ ~ ; ~ ~ 2 * * 2 * 2 * lim () (3) (2) nvestor Sentiment (S) and Stock Mispricing [RQ 2 ] (4) (5)

8 Ability of nvestor Sentiment to Color Corporate nvestment [RQ 2 ] (3) Theoretical Foundations Equity issuance channel (Stein, 996): Given overvaluation, managers of equity dependent firms will issue more shares at inflated prices [and invest the proceeds into new physical capital]: Pumping up the tulips requires distortions in real investment. f the share is undervalued, the firm will pass up the investment Catering theory (Polk and Sapienza, 29): make investment wanted by investors Pre commitment to real investment (DeLong et al., 989) Financial arbitrage in their own shares will hinder sentiment from affecting investment (Morck et al., 99 and Blanchard et al., 993) Empirical Evidence There is large evidence in support of the notion that managers make investment decisions that cater to investors sentiment (optimism): Grundy and Li, 2; Polk and Sapienza, 29; Dong et al., 27; Lamont and Stein, 26; Baker and Wurgler, 26; Gilchrist et al., 25; Jensen, 25; Goyal and Yamada, 24; Rhodes Kropf and Viswanathan, 24; Shleifer and Vishny, 23; Baker et al., 23; Chirinko and Schaller, 2; Chang et al., 27 s this good or bad for efficiency? (Shleifer, 2) Managers do not pursue an investment strategy that caters to current mispricing: Li et al., 2; Bakke and Whited, 2 Houdou Basse Mama 8

9 Hypotheses Hypothesis a. Managers follow the signals transmitted by the stock market when making corporate investment spending decisions. Hypothesis b. Stock price informativeness enhances managerial propensity to listen to the market. Hypothesis 2. Managers pursue investment strategies that cater to current investor sentiment driven mispricing shocks. Houdou Basse Mama 9

10 Expected Contributions of the Paper The extant empirical literature has hardly been concerned about endogeneity: [...] we simply follow a large earlier literature that has estimated investment cash flow regressions without taking into account the endogeneity problem. Wang et al. (29:55) Besides, test whether and to what extent public complaints about resource misallocation in Europe as a result of stock market valuations are exaggerated Disentangling the real investment effects of growth options and serendipitous information Measuring the distorting effects of putative stock misvaluations on real investment Policy implications: f stock market activity induces real negative externalities, What institutional changes are needed to mitigate the economic costs of mispricing? Cautionary investment of social security savings [4(k); Riester plans] in equities Emerging and developing economies would need to reconsider their endeavors to developing/reforming their capital markets Houdou Basse Mama

11 Estimable nvestment Equation (VE) The generic investment equation estimated (clustering at firm level) is: nv it i. fqit 2. fqit. NFOit 3. mqt. X it (6) Variable definition: nv: capital expenditures to capital ratio fq: fundamental component of the market to book ratio NFO: a proxy for stock price informativeness mq: non fundamental component of the market to book ratio X: a vector of control variables incl. time dummies Houdou Basse Mama

12 Measures of Stock Price nformativeness (SP) and Misvaluation Stock price informativeness: Logistic relative idiosyncratic volatility (extracted from a specific returngenerating process, usually the market model) Price impact (Amihud, 22; Thapa and Poshakwale, 2) nvestor sentiment driven misvaluation Non fundamental component in the market to book ratio (mq) (Rhodes Kropf, Robinson and Viswanathan, 25) Composite Security ssuance (Daniel and Titman, 26) Concerns: Are measurement errors at play here? mplications for estimator choice? Houdou Basse Mama 2

13 Some Variables: Summary Statistics (99 2) 4 European Countries Variable # Obs. Mean Std. Dev. 5th Pctl. 25th Pctl. 5th Pctl. 75th Pctl. 95th Pctl. Capital expenditures 3, Fundamental Q 3, Logistic relative idiosyncratic volatility 34, Non fundamental Q 3, Annual returns 34, nv it i. fqit 2. fqit. NFOit 3. mqt. X it (6) Houdou Basse Mama 3

14 Results (VE): Full Sample Main predictions (economic significance): Model 2 H a : () 8.25% H b : () 3.26% H 2 : ().23% Time varying effects Tech bubble: () Tech burst: (X) Subprime: (?) Other controls (not reported) Variables () (2) (3) (4) (5) capx capx capx dppe capx fq t-.39 *.64 ***.835 ***.67 ***.474 ** (.685) (.5) (.2) (.33) (.484) ψ t (.98) (.469) (.796) fq. ψ t-.223 ***.9 ***.92 *** (.) (.47) (.64) mq t-.783 ***.543 ***.659 ***.669 *** (.) (.6) (.4) (.35) mq.nsi t (.3628) (.2) (.3535) tech bubble.2 *** ***.78 *** (.56) (.3677) (.74) (.53) tech burst.49 ***.2.6 ***.2 ** (.) (.67) (.386) (.22) subprime **..5 (.339) (.59) (.3848) (.4685) mq.techbubble * -.7 (.2289) (.473) (.884) (.765) mq.techburst -. *** *** -.63 * (.5) (.8898) (.53) (.537) mq.subprime * Air t (.6262) (.25) (.33) (.633) (.62) fq.air t-.76 ** (.4) # obs J -statistic p -value (J-test) ? Houdou Basse Mama 4

15 Market Misvaluation led Equity ssuance and Cash Savings Variables Panel A: Equity ssuance (nsi) - OLS Panel B: Cash Savings (cash) - V Full Small Large Full Small Large fq (.7897) (.2243) (.662) mq.48 ***.74 ***.3 ***.42 ***.4.46 *** (.) (.) (.) (.) (.2266) (.5) tech bubble.5 *** **.37 ** * (.2) (.9689) (.24) (.37) (.749) (.534) mq.techbubble.78 *** (.) (.343) (.399) nsi.techbubble.264 ***.6586 *** (.93) (.3) (.263) capx -.98 * *** -.69 (.664) (.45) (.5558) nsi.449 ***.3934 ***.3799 *** (.) (.) (.) R p-value J-test # obs Houdou Basse Mama 5

16 Conclusions Managers vicariously learn from stock prices when making investment decisions. Specifically, managers propensity to learn increases in stock price informativeness. Moreover, the route of investor sentiment transmission to investment runs substantially through the catering channel. While opaque firms are more likely to cater, evidence for the equity issuance hypothesis is confined to large, transparent and constrained firms Small firms hoard part of the proceeds from sale of overvalued shares Rising bubbles are associated with vigorous investment episodes But unexpectedly neither the 2 23 recession nor the subprime crisis induced significant cuts in investment Houdou Basse Mama 6

17 Thank you! Houdou Basse Mama 7

18 Fig.: Aggregate Gross Fixed nvestment and the Stock Market 3,8 3,4 3 2,6 2,2,8,4 DJSTOXX Europe 6 Aggregate nvestment EURO Area Source: Raw data are from The World Bank (GFCF) and Datastream (DJSTOXX E 6). Both series are expressed in logs and rescaled to equal in 99. Houdou Basse Mama 8

19 nvestment Efficiency mplications coefficient on p -value investment Full sample *** (.) Panel A: Sort by firm size Small *** (.) Q *** (.) Q (.922) Large -.8 (.974) Panel B: Sort by price non-synchronicity Opaque -.24 (.7387) Q *** (.) Q *** (.) Transparent -.3 (.4) Panel C: Sort by KZ4 Unconstrained *** (.) Q *** (.) Q *** (.44) Constrained ** (.29) P P4 P4 P Houdou Basse Mama 9