Download presentation
Presentation is loading. Please wait.
Published byΣοφοκλής Βουρδουμπάς Modified over 5 years ago
1
Speculation, Price Indeterminacy and Money Supply in Financial Markets: An Experimental Study
Shyam Sunder (Joint Work with Shinichi Hirota, Juergen Huber, and Thomas Stoeckl) Shidler Lecture, School of Accountancy, Shidler College of Business, University of Hawaii at Manoa Honolulu, April 18, 2019
2
The purpose of this paper
Explore Why prices may deviate from fundamental values in financial markets. Focus on Investors’ short trading horizons (relative to maturity) and the difficulty of backward induction Supply of liquidity. Conduct Laboratory experiment .
3
Main Findings Prices tend to deviate from fundamental values (bubbles, indeterminacy) when investors have horizons shorter than the maturity of securities they trade. Prices are also more volatile under such circumstances Mispricing is likely to be strategic, and not non-rational Mispricing increases with the number of transfers that remain from present until maturity Speculative trading pushes prices upward (downward) when liquidity is high (low). .
4
Inference Difficulty of forming higher order beliefs about future cash flows appears to be a key factor Difficulty of backward induction through higher order beliefs to fundamental present values Why liquidity should affect prices remains an open issue. Needs further research. Expected to use up all their money? Play money syndrome? Happened under Greenspan’s easy money policy also
5
Previous Research on Bubbles
Rational Bubbles: Blanchard and Watson (1982), Tirole (1985); Infinite Maturity Deviations arise from future investors’ noisy beliefs or irrationality: Abreu and Brunnermeier (2003), Allen, et al. (2006), Delong, et al. (1990a, 1990b), Dow and Gorton (1994), Froot, et al. (1992), Scheinkman and Xiong (2003) Both utilize RE hypothesis (assuming that at least the current investors form rational expectations). However, empirical support for RE hypothesis is weak. Field surveys on financial markets (Greenwood and Shleifer 2014, Vissing-Jorgensen 2003, Frankel and Froot 1987, 1990, Taylor and Allen 1992). Experimental data (Marimon and Sunder 1993, Hommes, et al and Hommes 2011).
6
Background Bubbles and price volatility in financial markets are often attributed to short-term investors’ speculative trading. However, variations in decision horizons of investors do not enter finance theory. Even in a market dominated by short-horizon investors, their backward induction is supposed to generate prices close to the fundamental values. .
7
Two critical assumptions in finance theory
All future generations of investors form rational expectations of prices. Rational expectation is common knowledge By recursive process, Pt = Ft is derivable through backward induction.
8
In practice, backward induction may not hold.
Some generations of investors may not form rational expectations. Even if all generations of investors do, rational expectation may not be common knowledge. Under such conditions, investors cannot backward induct from first and higher order expectations to the present value of securities. Prices are no longer anchored to the fundamental values and become indeterminate. In laboratory experiments we explore whether investors’ difficulty in forming rational expectations is a cause of price indeterminacy in financial market.
9
Fundamental Value vs. Price for a simple, single dividend security
(1) Long-term Investor’s Valuation: (2) Short-term Investor’s Valuation: (3) Pt is not necessarily equal to Ft
10
For Pt to be equal to Ft Rational Expectation of P t+k
Homogeneous Investors The Law of Iterated Expectations By recursive process, Pt = Ft is derivable by the backward induction.
11
Difficulty of Backward Induction
Backward Induction may fail. Infinite maturity (rational bubbles) Blanchard and Watson (1982), Tirole (1985) Infinite number of trading opportunities Allen and Gorton (1993) Heterogeneous Information Froot, Scharfsten, and Stein (1992), Allen, Morris, and Shin (2002) Rationality may not be common knowledge Delong et al. (1990a)(1990b), Dow and Gorton (1994)
12
Price Bubble sans Dividend Anchors
There are cases where short-term investors have difficulty in backward induction. Stock prices (Pt ) form deviate from fundamentals ( Ft ) No longer anchored by future dividends
13
In an Earlier Experimental Study Hirota, Shinichi and Shyam Sunder
In an Earlier Experimental Study Hirota, Shinichi and Shyam Sunder. “Price Bubbles sans Dividend Anchors: Evidence from Laboratory Stock Markets,” Journal of Economic Dynamics and Control 31, no. 6 (June 2007): What happens when short-term investors have difficulty in the backward induction? Two kinds of the lab markets (1) Long-term Horizon Session (2) Short-term Horizon Session Bubbles tend to arise in (2), but not in (1)
14
Long-term Horizon Session
Period 1 Period 15 (Trade) D Single terminal dividend at the end of period 15. An investor’s time horizon is equal to the security’s maturity. Prediction: Pt = D
15
Short-term Horizon Session
Period 1 Period x Period 30 (Trade) Ex (Px+1) D Single terminal dividend at the end of period 30. The session will “likely” be terminated earlier. If terminated earlier, the stock is liquidated at the following period predicted price. An investor’s time horizon is shorter than the maturity and it is difficult to backward induct. Prediction: Pt D
16
Figure 4: Stock Prices and Efficiency of Allocations for Session 4
(Exogenous Terminal Payoff Session)
17
Figure 5: Stock Prices and Efficiency of Allocations for Session 5
(Exogenous Terminal Payoff Session)
18
Figure 6: Stock Prices for Session 6
(Exogenous Terminal Payoff Session)
19
Figure 7: Stock Prices and Efficiency of Allocations for Session 7
(Exogenous Terminal Payoff Session)
20
In long-horizon sessions
Long-horizon Investors play a crucial role in assuring efficient pricing. Their arbitrage brings prices to the fundamentals. Speculative trades do not seem to destabilize prices. 39.0% of transactions were speculative trades. By contrast, in short horizon treatments:
21
Figure 8: Stock Prices and Efficiency of Allocations for Session 1
(Endogenous Terminal Payoff Session)
22
Figure 9: Stock Prices and Efficiency of Allocations for Session 2
(Endogenous Terminal Payoff Session)
23
Figure 10: Stock Prices and Efficiency of Allocations for Session 8
(Endogenous Terminal Payoff Session)
24
Figure 11: Stock Prices and Efficiency of Allocations for Session 9
(Endogenous Terminal Payoff Session)
25
Figure 12: Stock Prices for Session 10
(Endogenous Terminal Payoff Session)
26
Figure 13: Stock Prices for Session 11
(Endogenous Terminal Payoff Session)
27
Discussion (short-horizon sessions)
Price levels and paths are indeterminate. Level Small Bubble (Session 1) Large Bubble (2, 8, 9, 10) Negative Bubble (11) Path Stable Bubble (1, 11, 2 ?) Rational Bubble Growing Bubble (8, 9, 10) Amplification Mechanism, Positive Feedback
28
Result In the long-horizon sessions, price expectations are consistent with backward induction. In the short-horizon sessions, price expectations are consistent with forward induction.
29
However, Objections to Design of the Short-Horizon Sessions
Period 1 Period x Period 30 (Trade) Ex (Px+1) D Single terminal dividend at the end of period 30. The session will “likely” be terminated earlier. If terminated earlier, the stock is liquidated at the following period predicted price. Environment not fully specified In the current work, we use a fully specified overlapping generations structure
30
Present Laboratory Experiment
A market has 16 periods of trading (120 sec). At all times two overlapping generations are present in the market (5 traders/generation). 1, 2, 4, and 8 entering generation(s). Investors’ holding periods decrease in the number of entering generations. Assets pay a single, certain, common knowledge certain dividend (50) only to the last generation. Last generation: dividend-collecting investors Previous generations: speculating investors Low/High Liquidity treatments Speculating investors will consider the valuation of future generations according to standard finance theory.
31
Markets with Overlapping Generations of Traders
Every period has two overlapping generations of five traders each in the market Only one initial generation is endowed with assets (single common knowledge dividend of 50 paid at maturity—end of period 16) All other generations enter with cash, can buy assets from the “old” generation, and sell them when they become “old” to exit the market with cash Individuals may re-enter after sitting out the market for one or more (random number) of generations (in T4 and T8 only) Each session is repeated six times (independently with different subjects) Equilibrium transaction volume per session: 160
32
Table 1: Overlapping Generations Experimental Design (Four Treatments: T1, T2, T4, and T8)
Period Subjects 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 T1 G0 G1 T2 G2 T4 G3 G4 T8 G5 G6 G7 G8
33
Laboratory Experiment
Endowments Initial generation (G0): assets; “Entering” generations: cash Specific parameters differ across treatments to control for shorter holding periods (keep “workload” constant). High and Low liquidity markets. Trading mechanism: countinous double auction, single unit trading, no short sales or margin buying. Assets not sold by the leaving generation are re-distributed randomly among entering traders. Entering subjects are recruited from a pool of inactive participants (either 15 or 18 sub./session).
35
Table 2: Treatment Overview
Liquidity Low (C/A ratio=2) High (C/A ratio=10) Number of entering generations 1 T1L T1H 2 T2L T2H 4 T4L T4H 8 T8L T8H Six replications for each treatment. 828 bachelors and masters students from different fields.
36
Continuous double auction markets
Trader: Information about your task (trader), period you leave the market, current Share and Taler holdings. Predictors: Information about your task (predictior) and your forecast. BID: enter the price you are willing to pay for one unit. Trade does not take place until another participant accepts your bid!!! Current Market Price (of Stock) ASK: seller’s analogue to BID - see above. List of all BIDS: from all traders - your own bids are written in blue. The bid with blue background is always the most attractive, yielding the highest revenues for the seller. List of all ASKS: from all traders - your own asks are written in blue. The ask with blue background is always the most attractive, because it is the cheapest for the buyer. Price-Chart of current period SELL: You sell one unit, given the price with the blue background. BUY: You buy one unit, given the price with the blue background.
37
Hypotheses Hypothesis 1: Deviations of prices from the fundamental value are the same during periods when only speculating investors are present, as compared to periods when dividend-collecting investors are present in the market. Hypothesis 2: For a security of a given maturity, the deviation of prices from the fundamental value is not affected by the length of investors’ holding periods. Hypothesis 3: Prices will be the same irrespective of the total amount of cash in the economy. H1: following REE, there should be no difference in mispring between periods with speculating investors or div.collecting investors present. H2: investigate the effect of the length of the holding period. According to finance theory, the holding period should have no influence on pricing. (Treatmentvergleich) H3: Liquidity should not matter in the pricing of a security (its value remains unaffected).
38
Experimental Results
40
Red line: Fundamental value
G0 G1 Red line: Fundamental value Thin grey line: Prices in individual sessions Bold line with circles: mean prices across 6 sessions We find that in T1L markets with long-horizon investors (G1), prices are mostly close to fundamentals (50) throughout the session.
41
G0 G1 G2
42
G0 G1 G2 G3 G4
43
G0 G1 G2 G3 G4 G5 G6 G7 G8
44
Summary of results (low Liquidity)
Low liquidity treatments
45
G0 G1
46
G0 G1 G2
47
G0 G1 G2 G3 G4
48
G0 G1 G2 G3 G4 G5 G6 G7 G8
49
Summary of results (high Liquidity)
50
Figure 2: High Liquidity Treatments
51
Hypotheses Hypothesis 1: Deviations of prices from the fundamental value are the same during periods when only speculating investors are present compared to periods when dividend-collecting investors are present in the market. Hypothesis 2: For a security of a given maturity, the deviation of prices from the fundamental value is not affected by the length of investors’ holding periods. Hypothesis 3: Prices will be the same irrespective of the total amount of cash in the market.
52
Table 4: Formulae for market efficiency measures
53
Results: Hypothesis 1 Periods with dividend- collecting investors (2) Periods with only speculating investors Difference (2)-(1) High-liquidity session (H) 0.401 (177) 1.024 (204) 0.623*** Low-liquidity session (L) 0.140 (178) 0.502 (203) 0.362*** Period-RAD generally higher in H. Period-RAD ( ) is larger during periods when only speculating investors are present – reject H1. REE does not hold in our markets.
54
Results: Hypothesis 1 To gain some further insights on pricing behavior of speculating investors we investigate whether with fewer generations left, it should be easier to arrive at REE. Calculate Period-RAD conditional on the number of yet-to-enter generations until maturity. – split again between H and L markets. Black horizontal line represents RAD for period with dividend-collecting investors present. Speculating investors have difficulties in forming RE even if only one future generation is left.
56
Hypotheses Hypothesis 1: Deviations of prices from the fundamental value are the same during periods when only speculating investors are present compared to periods when dividend-collecting investors are present in the market. Hypothesis 2: For a security of a given maturity, the deviation of prices from the fundamental value is not affected by the length of investors’ holding periods. Hypothesis 3: Prices will be the same irrespective of the total amount of cash in the market.
57
Results: Hypothesis 2 Treatment (Average holding periods) T1 (16.0 periods) T2 (10.7 periods) T4 (6.4 periods) T8 (3.6 periods) High-liquidity session (H) 0.421 (95) 0.586 (94) 0.739 (96) 1.187 Low-liquidity session (L) 0.116 0.355 0.429 One may argue that our experimental results are consistent with the theoretical predictions of previous literature, showing that investors’ short-term speculation gives rise to price bubbles (e.g., Allen, et al. 2006, Blanchard and Watson 1982, DeLong, et al. 1990a, 1990b, Dow and Gorton 1994, Froot et al. 1992, Tirole 1985). We should note, however, that our findings on the price indeterminacy associated with short-term speculation are obtained even in markets where the security has finite maturity and the dividend value has no uncertainty and is common knowledge, which excludes important factors postulated to cause bubbles in the prior literature. In our markets price deviations and volatility stem from the difficulties of investors in forming rational expectations, whereas the above literature assumes that investors can form rational expectations of future prices. Average holding period and price deviation from fundamentals are negatively related – reject H2.
58
High-liquidity Session (H) Low-liquidity Session (L)
Panel B: Differences between Average Period-RAD across Treatments Table 5: Comparison of Average Period-RAD between Treatments with High and Low Liquidity. Panel B: Differences between Average Period-RAD across Treatments High-liquidity Session (H) T2 T4 T8 T1 0.165* 0.318*** 0.766*** 0.153 0.601*** 0.448*** Low-liquidity Session (L) 0.239*** 0.313*** 0.075 0.074 0.000 Notes: Two-sided t-test significance levels * (10%), ** (5%) and *** (1%).
59
Hypotheses Hypothesis 1: Deviations of prices from the fundamental value are the same during periods when only speculating investors are present compared to periods when dividend-collecting investors are present in the market. Hypothesis 2: For a security of a given maturity, the deviation of prices from the fundamental value is not affected by the length of investors’ holding periods. Hypothesis 3: Prices will be the same irrespective of the total amount of cash in the economy.
60
Results Hypothesis 3 Prices tend to be The average of Period-RD
above fundamental value in H-sessions below fundamental value in the L-sessions. The average of Period-RD = in H-sessions = in L-sessions Difference between H and L is statistically significant at the 1%-level – reject H3. Given the observation, we consider a measure taking into account the direction of mispricing. Difference between H and L is statistically significant (1%)
61
Periods with dividend- collecting investors
Results: Hypothesis 3 Periods with dividend- collecting investors (2) Periods with only speculating investors Difference (2)-(1) High-liquidity session (H) 0.295 (177) 0.741 (204) 0.446*** Low-liquidity session (L) -0.087 (178) -0.340 (203) -0.253*** Liquidity effect is larger when only speculating investors are present. We conjecture that lower prices in low-liquidity treatments could be caused by speculating investors’ fear of future market illiquidity. Suppose that speculating investors have difficulties in rationally expecting future sales prices and observe weak buy-order and low transaction prices in some period of our experiment. Then, they may sell the security now even at prices below 50, fearing that they may not be able to sell all their securities before their exit, or may be forced to dump them in fire sales. This behavior of speculating investors would tend to drive prices below fundamentals. This conjecture is supported by theoretical analyses of financial liquidity crises by Bernardo and Welch (2004) and Morris and Shin (2004). They point to speculating investors selling securities expecting future market declines, and causing price drops. It is also consistent with an empirical study by Cella et al. (2013) who find that during episodes of market turmoil, short-term investors sell more than long-term investors, and stocks held mostly by short-term investors experience larger price drops than stocks held mostly by long-term investors. In addition, Morris and Shin’s (2004) model predicts a V-shaped pattern in prices around the liquidity crisis; after the crisis, prices go back to fundamentals through the long-term investors’ arbitrage transactions. Cella et al. (2013) also report that stocks held mostly by short-term investors experienced large price reversals after the turmoil. These V-shaped price paths from theoretical and empirical studies are also observed in our low-liquidity sessions. As Figure 3 shows, in T2L, T4L, and T8L markets, prices tend to decline when there exist only speculating investors, but they generally recover and converge to fundamentals once dividend-collecting investors (the last generation) enter the market. In the high-liquidity treatments this is less likely as each individual investor has enough money to “buy the whole market”, i.e., buy all the assets in the market at their fundamental value. In H, spec. investors magnify the overpricing (bubbles). In L, spec. investors magnify the undervaluation (liquidity crisis).
62
Results: Hypothesis 3 Speculating investors have difficulties in forming RE even if only one future generation is left.
63
Figure 4: Average Period-RAD for each period number: Comparison between the markets with dividend-collecting investors and those with only speculating investors. Figure 4: Average Period-RAD for each period number: Comparison between the markets with dividend-collecting investors and those with only speculating investors. Notes: In periods 15 and 16 dividend-collecting investors are present in all treatments (see, Table 3). Therefore only black bars are shown for these two periods.
64
Price Predictions/Expectations
Hirota and Sunder (2007): results show that when subjects cannot do backward induction, they resort to forward induction, and simply project past data in forming their expectations about the future In long-horizon sessions, future price expectations are formed by fundamentals. Speculation stabilizes prices. In short-term sessions, future price expectations are formed by their own or actual prices. Speculation may destabilize prices.
65
Price Expectations Speculating investors have difficulty in forming RE of future prices. Then, how do they expect future prices? The fundamental model The trend model The combined model We utilize the price expectation data submitted by the predictors as proxies for investors’ price expectations and investigate how price expectations are formed in our experimental markets. We examine the price expectation formation process using two models; one is the fundamental model and the other is the trend model. The fundamental model assumes that investor form expectations about future prices based on the price deviation from the fundamental value of the security. With this model, investors expect future price appreciation (depreciation) if the fundamental value, Ft, is higher (lower) than the current price, Pt. We should also argue that =1 in this model corresponds to the rational expectation formation supposed by the standard asset pricing models. The trend model assumes that investors expect the future price based on the recent observed price changes. In this model, if beta>0, recent price increases (decreases) cause investors to expect further price increases (decreases) in the future; if beta<0, recent price increases (decreases) cause investors to expect future price decreases (increases). With this model, investors’ expectation of future prices are solely affected by recent price movement itself, irrespective of the fundamental value of the security. The combined model allows for the possibility that investors use some combinations of the fundamental model and the trend model.
66
Price expectations model estimates
High-liquidity Session Periods with div. coll. investors Periods with speculating investors FUND TREND COMBINED Const. 1.672** -0.709 1.733** 4.159** -2.611* 0.515 (0.622) (1.595) (0.620) (1.895) (1.310) (1.449) (Ft - Pt) 0.197*** 0.211*** 0.109* 0.078 (0.043) (0.053) (0.061) (0.057) (Pt - Pt-1) 0.020 0.067 -0.301*** -0.270*** (0.031) (0.044) (0.049) N 173 167 186 168 F 20.96 0.42 8.09 3.19 37.71 25.16 p 0.000 0.522 0.002 0.092 adj. R2 0.38 0.00 0.39 0.14 0.30 0.36 Let us first look at the results for high liquidity sessions. For periods with long-horizon investors, the data are more consistent with the fundamental model than the trend model. These results suggest that in the periods with long-horizon investors, the fundamental value of the security not only determines transaction prices but also critically affect the future price expectations. (but, note that 1, RE formation is not supported). In contrast, for the periods with only short-horizon investors, the trend model is more supported over the fundamental model. These results indicate that in a market with only short-horizon investors, investors form future price expectations not based on the fundamental value, but based on current or past transaction prices. Also, the negative coefficient of (Pt -Pt-1) shows that market participants expect price reversals. This observed reversal expectations are in a sharp contrast with the momentum (extrapolative) expectations of investors reported by previous studies in the field (Greenwood and Shleifer 2014, Vissing-Jorgensen 2003) and laboratory (Hirota and Sunder 2007).
67
Price expectations model estimates
Low-liquidity Session Periods with div.coll. investors Periods with speculating investors FUND TREND COMBINED Const. -2.275*** 1.054 -2.543** -0.804 -0.248 -1.636** (0.684) (0.737) (0.742) (0.524) (0.399) (0.671) (Ft - Pt) 0.401*** 0.419*** 0.070*** 0.065** (0.092) (0.096) (0.017) (0.024) (Pt - Pt-1) -0.088 -0.016 -0.162* -0.180** (0.079) (0.031) (0.081) (0.074) N 171 162 186 168 F 19.36 1.26 10.10 16.25 3.95 5.82 p 0.000 0.274 0.001 0.063 0.012 adj. R2 0.43 0.01 0.08 0.13
68
Figure D4: Average period price predictions and average period price realizations
69
Consequences for Financial Reporting
Importance of shared observables in formation of market prices Role of Verified financial reports Role of dividends: relevance of dividend policy What happens (happened) in a world of mark-to-market accounting: loss of verified anchors
70
Wrap Up Prices are close to the fundamental values when Investors have long-term horizons. Prices deviate from the fundamental values and become indeterminate when there are only short-term investors in the market. Investors fail to backward induct to bring prices to the fundamental values. The shorter the investment horizon (the larger number of generations), the more difficult the backward induction. Prices in high liquidity treatments are higher and deviate more from fundamentals than those in low liquidity treatments.
71
Implications Bubbles are known to occur more often in markets for assets with (i) longer maturity and duration (ii) higher uncertainty These field observations line up with the lab data Inputs to expectation formation matter: Dividend policy matters! Ex post, market inefficiency, anomalies, and behavioral phenomena more likely to be observed in markets dominated by short-horizon investors (difficulty of backward induction)
72
Table 3: Average Period-RAD for each period
Panel A: High-liquidity session Period 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 T1 1.423 0.582 0.354 0.293 0.329 0.301 0.321 0.390 0.374 0.382 0.396 0.303 0.323 0.286 0.387 0.259 T2 1.825 1.016 0.310 0.406 0.467 0.536 0.541 0.477 0.676 0.865 0.705 0.313 0.232 0.468 0.179 T4 1.552 1.471 1.342 1.038 1.182 0.960 0.798 0.499 0.697 0.509 0.470 0.559 0.325 0.210 0.167 0.040 T8 1.879 1.249 1.373 1.392 1.409 1.498 1.177 0.991 1.108 1.082 1.607 1.733 1.019 0.647 0.550 0.273 Panel B: Low-liquidity session Period 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 T1 0.226 0.139 0.106 0.098 0.077 0.101 0.103 0.138 0.152 0.158 0.147 0.070 0.084 0.083 0.085 T2 0.596 0.425 0.299 0.278 0.503 0.685 0.743 0.760 0.342 0.352 0.222 0.146 0.071 0.053 0.115 T4 0.385 0.489 0.495 0.543 0.517 0.527 0.556 0.653 0.535 0.530 0.511 0.459 0.341 0.163 0.110 0.052 T8 0.214 0.249 0.398 0.315 0.313 0.355 0.499 0.446 0.584 0.628 0.741 0.663 0.679 0.230 0.066
73
Table 5: Treatment averages for market efficiency measures
Relative Absolute Deviation Relative Deviation Bid-Ask Spread Std. Dev. of Log Returns Share Turnover T1L 11.43% -5.24% 8.79% 4.70% 1.60 T2L 35.47% -18.95% 19.92% 17.56% 1.69 T4L 42.92% -34.07% 22.52% 25.16% 1.56 T8L 43.03% -30.48% 21.69% 26.24% 1.05 T1H 41.99% 37.39% 29.61% 14.08% 2.01 T2H 77.02% 38.81% 55.04% 22.70% 1.59 T4H 73.86% 52.11% 23.37% 17.72% 1.57 T8H 118.67% 103.48% 65.95% 18.17% 1.07
74
Table 6: Differences between averages across treatments, same Liquidity (RAD, RD, SPREAD, VOLA, and ST two-sided Mann-Whitney U) RAD T2L T4L T8L T2H T4H T8H T1L 24.03%** 31.48%*** 31.60%*** T1H 35.03% 31.87% 76.67%* 7.45% 7.56% -3.16% 41.64% 0.11% RD -13.70%** -28.83%*** -25.23%** 1.42% 14.71% 66.09% -15.13% -11.53% 13.29% 64.67% 3.60% 51.38% SPREAD 11.13%* 13.73%** 12.90%** 25.43% -6.24% 36.34% 2.59% 1.76% -31.67% 10.91% -0.83% 42.58%** VOLA 12.86%** 20.46%*** 21.54%*** 8.61% 3.64% 4.09% 7.60% 8.68%* -4.98% -4.53% 1.08% 0.45% ST 0.09 -0.03 -0.55** -0.43 -0.44 -0.94** -0.12 -0.64*** -0.02 -0.52** -0.51 -0.50**
75
Table 7: Differences between averages across treatments, different Liquidity (RAD, RD, SPREAD, VOLA, and ST two-sided Mann-Whitney U) H minus L RAD RD SPREAD VOLA ST T1 30.56%** 42.64%*** 20.82%*** 9.39%* 0.42 T2 41.56% 57.76%** 35.12%* 5.14% -0.10 T4 30.95% 86.18%*** 0.86% -7.44% 0.01 T8 75.64%* 133.96%*** 44.26%** -8.07% 0.02
76
Figure D1: Exiting generations’ share of initial security holdings by period
77
Table D1: Forfeiture rates (fractions of all securities not sold by exiting generation to entering generation) across treatments
78
Figure D2: Security concentration (SC), calculation example.
79
Figure D3: Security concentration (SC) over time in the four treatments
80
High-liquidity session (H) Low-liquidity session (L)
Table D2: The number of security transfers and average period-SC by treatment Treatment T1*** T2*** T4 T8*** High-liquidity session (H) 0.697 (96) 0.666 0.610 0.690 Low-liquidity session (L) 0.603 0.580 0.613 0.617
81
Table D3: Comparison of Average Period-SC between Periods with Dividend-collecting Investors and Periods with only Speculating Investors Periods with dividend- collecting investors Periods with only speculating investors Difference (2)-(1) High liquidity Session (Treatment H) 0.725 (180) 0.613 (204) -0.112*** Low liquidity Session (Treatment L) 0.633 0.577 -0.055*** Notes: Sample size is in parentheses. *** indicates that the difference is statistically significant at 1% level by two-sided t-test.
82
Table D4: Comparison of average period-SPEC between periods with dividend-collecting investors and periods with only speculating investors (1) Periods with dividend- collecting investors (2) Periods with only speculating investors Difference (2)-(1) High liquidity sessions (Treatment H) 0.289 (178) 0.301 (204) 0.012 Low liquidity sessions (Treatment L) 0.232 0.322 (203) 0.091*** Note: Sample size is in parentheses. *** indicates that the difference is statistically significant at 1% level by two-sided t-test.
83
Table D5: Prediction Accuracy (regression analysis)
Panel A: High Liquidity Sessions dependent variables Abs(EP-P)/P Abs(EP-P)/50 Intercept 0.250 *** 0.197 ** 0.230 0.130 (0.048) (0.011) (0.037) (0.051) Number of Entering generations Left 0.016 0.046 * (0.018) (0.026) Number of Periods Left 0.010 0.021 (0.007) (0.005) R2 0.004 0.008 0.038 0.050 N 381 Panel B: Low Liquidity Sessions 0.224 0.184 0.106 0.060 (0.045) (0.015) (0.014) 0.041 0.020 (0.017) (0.006) 0.013 (0.002) 0.028 0.017 0.084 Notes: Standard errors clustered by session in parenthesis. Significance levels: * (10%), ** (5%) and *** (1%).
84
Thank You! Shyam.sunder@yale.edu
Similar presentations
© 2025 SlidePlayer.com. Inc.
All rights reserved.