Option Market Activity and Behavioral Finance

Option Market Activity and Behavioral Finance

Josef LakonishokInmoo Lee

Allen M. Poteshman

2

How Do Investors Use Options?

Over last three decades much work on pricing and hedging stock optionsLittle is known about how investors actually employ stock options What are the levels of long and short open

interest and trading volume by different types of investors?

What are the cross-sectional determinants of option activity?

How did option market activity change during the bubble of the late 1990s and early 2000?

3

This paper

Establish some key facts about option market activity In general During the bubble

Offer conjectures about the facts based upon Discussions with option market

participants Behavioral finance literature

4

Unique Dataset to Establish Facts

Detailed daily open interest and volume information for all CBOE listed options from 1990 through 2001All data broken down by different types of investorsOpen interest: long and shortVolume: buyer/seller initiated and open/close option positionsLimitation: Cannot see combined trades

5

Our Approach

For different types of investors and categories of stocks get daily average Long and short, put and call open interest Call and put buys and sells to open new

positions

Regress four volume categories on stock returns over various past horizonsDo analyses over entire 1990-2001 period and for subperiods

6

Results 1: Option Market Activity Levels

Fact: Non-market makers have about four times more long call than long put open interest

Comment: Surprising because harder to short stocks Conjecture: Analysts make pos. recs. Easier to manage calls

Fact: Non-market makers have more short than long call open interest

Conjecture: Much of the short call open interest is part of covered call positions. Compared to holding stock long, covered calls lose in fewer states of the world and the losses are less severe. Consistent with loss aversion and narrow framing

Fact: Non-market makers have more short than long put open interest

Conjecture: Puts are sold on stocks viewed as undervalued. If stock increases just keep the premium, if it decreases buy the stock at an even lower price

Corroboration: More short put open interest is observed on value than growth stocks

7

Results 2: Cross-Sectional Determinants of Option Activity

Fact: All types of non-market makers buy more calls to open new positions after high past returns (extending up to two years in the past)

Conjecture: Option market participants are trend-chasers, and sentiment about stocks is established over long horizons

Fact: More calls are also sold after high past returns Conjecture: This also represents trend-chasing because

the short calls are part of covered call positions which are viewed as conservative long positions

Fact: More puts are sold to open new positions when past returns on underlying stock are low over past quarter

Conjecture: Contrarian investors like to sell puts on beaten down stocks

8

Results 3: Option Market Activity During the Bubble

Fact: Unsophisticated investors significantly increased buying of calls on growth but not value stocks during the bubbleFact: Sophisticated investors did not increase their call buying during the bubbleFact: None of the investor types increased their put buying during the bubble No evidence that investors used the option

market during the bubble to overcome short-sales constraints (compare to Ofek and Richardson (2003))

9

Outline

DataOption market activity levelsCross-sectional determinants of option market activityOption market activity during the bubbleConclusion

10

DataDaily data on all CBOE listed options from Jan 1990 through Dec 2001Three investor classes: firm proprietary, discount, full-serviceOpen interest: long and short for each investor classVolume: buyer/seller initiated and opening/closing option position (also for each investor class)CRSP/COMPUSTAT: stock returns, shares outstanding, BM ratios

11

Levels of Investor Sophistication

Most sophisticated: Firm proprietary traders (Poteshman and Serbin (2003))Intermediate sophistication: Full-service customers Includes hedge funds Mahani and Poteshman (2003) Pan and Poteshman (2003)

Least sophisticated: Discount customers Odean (1998), Odean and Barber (2000) Mahani and Poteshman (2003) Pan and Poteshman (2003)

12

Measuring Option Market Activity

Measure option open interest or volume on a trade date (t), underlying stock (s), investor class (i), and type of open interest (k):

Similarly for OptionVolPercentageShares

,,, , , ,

1,,

,

100

100

Callss tN

k i Calls j t s j t

jk is t Shares

s t

OpenInterest

OpenInterestPercentageSharesN

13

Table 1: Avg. Daily Open Interest as % of Shares Outstanding

Average aggregate open interest is 0.56% of shares outstanding Turnover 6%/year

Investor Type Long Call Long Put Short Call Short PutNon Market Makers 0.200% 0.047% 0.245% 0.063%

Firm Proprietary 0.042% 0.014% 0.031% 0.010%Discount Customers 0.031% 0.004% 0.023% 0.008%

Full-Service Customers 0.126% 0.029% 0.191% 0.046%

Large Underlying StocksType of Open Interest

14

Table 1: Avg. Daily Open Interest as % of Shares Outstanding

Underlying Stocks Long Call Long Put Short Call Short Put

Large Growth 0.044% 0.015% 0.032% 0.011%Large Value 0.039% 0.019% 0.041% 0.011%



Type of Open Interest

Panel A: Firm Proprietary Traders

Panel B: Discount Customers

Panel C: Full-Service Customers

15

Profit to Covered Call

Long Stock

Stock Price at Expir.

Short Call

Covered Call

K

Profit

SCurrent

(OTM)

16

Levels of Open InterestAcross the investor classes open interest for long calls is about four times greater than for long puts Somewhat surprising because it is more

costly and difficult to go short than long in the stock market but not in the option market

On the other hand: financial advisers push long stock positions, long calls are easier to manage

17

Levels of Open Interest (Cont.)

More short than long call open interest On surface contradicts idea that investors are relatively

averse to short positions But many short calls are part of covered call positions

Heavily promoted as a conservative long position Consistent with loss aversion and mental accounting

For discount and full-service more short put than long put open interestOver entire 1990-2001 time period there are no major differences across different types of underlying stocks

Largest difference: Full-service customers have more short put open interest on value than growth

18

Table 2: Avg. Daily Open Volume as % of Shares Outstanding

Investor Type Buy Call Buy Put Sell Call Sell PutNon Market Makers 0.00527% 0.00178% 0.00432% 0.00139%

Firm Proprietary 0.00077% 0.00032% 0.00064% 0.00026%Discount Customers 0.00067% 0.00016% 0.00029% 0.00011%

Full-Service Customers 0.00383% 0.00130% 0.00339% 0.00102%

Type of Open VolumeLarge Underlying Stocks

19

Table 2: Avg. Daily Open Volume as % of Shares Outstanding

Underlying Stocks Buy Call Buy Put Sell Call Sell Put




Panel B: Discount Customers Average Daily Open Volume

Panel C: Full-Service Customers Average Daily Open Volume

Panel A: Firm Proprietary Trader Average Daily Open Volume

20

Table 2: Avg. Turnover Time (in Trade Dates)

Investor Type Buy Call Buy Put Sell Call Sell PutFirm Proprietary 55 44 48 38

Discount Customers 46 23 81 71Full-Service Customers 33 23 56 45

Large Underlying StocksType of Open Volume

21

Levels of Option Volume

For both calls and puts there is more open buys than open sells for all investor classes.Since there is more short than long call open interest for full-service customers, they must hold short call positions longer than long call positions Short call positions are held for an average of 56 days Long call positions are held for an average of 33 days Suggests full-service (and discount) customers use

short calls in longer-term investment strategies

22

Cross-Sectional Determinants

Dependent variables: Open buy call volume, open sell call volume, open buy put volume, open sell put volume.Explanatory variables: Returns on the underlying stock over the past week, month, quarter, year, and two yearsControls: BM ratio and volatility of underlying stockReport time-series averages of the intercept and slopes from daily cross-sectional regressions

23

Table III: Large Stock Regressions

Dependent Variable Intercept Rsameday Rweek Rmonth Rquarter Ryear R2years ln(B/M) Volatility

Open Buy Call -0.0002 0.0074 0.0048 0.0023 0.0011 0.0005 0.0000 0.0002 0.0048Volume (-1.69) (5.03) (4.85) (3.96) (2.71) (3.20) (-0.09) (4.08) (5.69)

Open Sell Call -0.0001 0.0064 0.0025 0.0016 0.0010 0.0007 0.0002 0.0002 0.0038Volume (-1.26) (4.78) (2.55) (2.81) (3.03) (5.39) (2.43) (5.18) (6.02)

Open Buy Put -0.0001 -0.0057 -0.0017 0.0000 0.0001 0.0001 0.0002 0.0001 0.0019Volume (-1.39) (-8.04) (-6.23) (-0.01) (0.80) (1.63) (3.82) (5.54) (7.21)

Open Sell Put -0.0001 -0.0074 -0.0015 -0.0003 0.0000 0.0001 0.0001 0.0000 0.0014Volume (-1.76) (-11.36) (-6.47) (-1.63) (0.49) (3.83) (2.51) (4.09) (9.38)


Independent Variables

24



Open Buy Call -0.0013 0.0023 0.0007 0.0006 0.0006 0.0008 0.0005 0.0000 0.0062Volume (-16.40) (4.11) (1.83) (2.59) (3.94) (9.49) (8.49) (0.65) (24.17)

Open Sell Call -0.0007 0.0066 0.0009 -0.0002 -0.0002 0.0001 0.0002 0.0000 0.0037Volume (-26.70) (22.79) (6.25) (-1.89) (-2.98) (3.12) (7.03) (2.18) (33.10)

Open Buy Put -0.0003 -0.0005 0.0004 0.0002 0.0001 0.0002 0.0001 0.0000 0.0014Volume (-11.98) (-3.08) (4.12) (2.69) (1.75) (7.83) (5.88) (-1.96) (20.13)

Open Sell Put -0.0003 -0.0023 -0.0008 -0.0004 -0.0001 0.0001 0.0001 0.0000 0.0015Volume (-20.65) (-17.60) (-11.24) (-8.54) (-5.04) (4.44) (7.36) (0.33) (26.96)



25



Open Buy Call -0.0048 0.0411 0.0089 0.0050 0.0030 0.0021 0.0010 0.0005 0.0360Volume (-13.68) (8.23) (2.64) (2.95) (2.53) (4.98) (4.11) (4.80) (18.28)

Open Sell Call -0.0035 0.0496 0.0145 0.0051 0.0018 0.0015 0.0011 0.0003 0.0285Volume (-13.43) (14.20) (5.61) (3.70) (2.15) (4.45) (4.55) (4.02) (19.02)

Open Buy Put -0.0015 -0.0278 -0.0031 -0.0004 0.0003 0.0009 0.0006 0.0002 0.0117Volume (-14.28) (-17.94) (-3.69) (-0.53) (0.95) (7.32) (5.32) (4.89) (21.25)

Open Sell Put -0.0016 -0.0188 -0.0061 -0.0019 -0.0005 0.0002 0.0007 0.0002 0.0118Volume (-10.01) (-15.35) (-5.61) (-2.74) (-1.65) (1.75) (5.25) (5.93) (18.26)



26

Table IV: Impact of One Standard Deviation Shock

VolumeVariable Rweek-R2years Rweek-Rquarter Ryear-R2years

Open Buy Call Volume 93.99 74.34 19.65

Open Sell Call Volume 102.38 60.55 41.82

Open Buy Put Volume 7.42 -22.34 29.76

Open Sell Put Volume -8.63 -35.36 26.73


Dependent Vars. Receiving Positive One Std. Dev. Shock

27





Open Buy Put Volume 77.86 25.58 52.28




28





Open Buy Put Volume 24.36 -10.53 34.90




29

Open Buy Call VolumeDiscount and full-service customers have significantly positive coef. on the return variables for all past horizons Trend-chasing: Daily data, option contracts can be

created and destroyed at will Sentiment developed over extended periods impacts

investment decisions

The impact of past returns is economically large One standard deviation shock to past return variables

increases daily open buy call volume for discount and full-service by 78% and 57%, respectively.

30

Open Sell Call Volume

Full-service customers have positive and significant coefficients on past returns variables. One standard deviation shock increases vol. by 63% Most of volume is from writing covered calls Conservative, income enhancing strategy Prospect theory maintains that investors are more

likely to write calls on their stocks with gains than losses

Less wealthy discount customers are less likely to own underlying stock and the impact of past returns is less clear

31

Open Buy Put VolumeDiscount customers: Buy more (fewer) new puts on underlying stocks that have increased (decreased) in price Expected from prospect theory: Investors are

locking in gains Contrarian investing

Full-service customers: Similar to discount, but clear effect is limited to returns more than three months in pastSince buying puts is not a popular activity, the average option market investor is a trend-chaser (from open buy call regressions)

32

Open Sell Put Volume

There is more selling than buying of putsFor discount and full-service investors the coefficients on past returns are negative over the last quarter and positive over longer horizons Suggests investors believe weakness in stock

is temporary Investors sell puts thinking that if stock price

increases they will just keep premium while if it decreases they are happy to buy the stock at the strike price

33

Option Market Activity During the Bubble

Compare option market activity of the various investor classes before during and after the bubbleDefine Pre-bubble 1990 - 1994 Beginning of bubble 1995 - 1997 Height of bubble 1998 - March 2000 Post-bubble April 2000 - 2001

34

Fig. 2: Price/Book of Russell 1000 Growth Divided by Price/Book of Value

0

0.5

1

1.5

2

2.5

3

3.5

4

4.5

5

Dec-9

0

Dec-9

1

Dec-9

2

Dec-9

3

Dec-9

4

Dec-9

5

Dec-9

6

Dec-9

7

Dec-9

8

Dec-9

9

Dec-0

0

Dec-0

1

(Gro

wth

P/B

)/(V

alue

P/B

)

35

Table V: Avg. Daily Open Volume as % of Shares Outstanding (Large Stocks)

Time Period Firm Discount Full-Service1990-1994 0.00091% 0.00047% 0.00432%1995-1997 0.00077% 0.00082% 0.00433%

1998 - March 2000 0.00065% 0.00119% 0.00380%April 2000 - 2001 0.00050% 0.00034% 0.00165%

Investor typeOpen Buy Call Volume

36

Table V: Avg. Daily Open Volume as % of Shares Outstanding (Large Stocks)

Time Period Firm Discount Full-Service1990-1994 0.00034% 0.00014% 0.00141%1995-1997 0.00024% 0.00018% 0.00137%

1998 - March 2000 0.00028% 0.00023% 0.00125%April 2000 - 2001 0.00042% 0.00009% 0.00095%

Open Buy Put VolumeInvestor type

37

Open Buy Call VolumeDiscount customers

Doubles from pre-bubble to beginning of bubble Increases another 50% at height of bubble Declines by a factor of three post-bubble Increased other open volume as well, but during

height of bubble 56% was buying calls

Full-service customers Stable from pre-bubble to beginning of bubble Decreases a bit during height of bubble Cut in half in post-bubble period

Firm proprietary traders The bubble is a non-event

38

Open Buy Put Volume

Well-known that it can be difficult to establish short positions in the stock marketOfek and Richardson (2003) suggest that short-sales constraints were an important contributor to the bubbleBut, we see no major increase in open buy put volume during the bubbleApparently, during a bubble it is not easy to be contrarian

39

Table VI: Large Stock Open Buy Call Volume Regr. for Discount Customers

Time Period Intercept Rsameday Rweek Rmonth Rquarter Ryear R2years ln(B/M) Volatility

1990-1994 -0.00064 0.00215 0.00033 0.00063 0.00064 0.00021 0.00004 -0.00003 0.00457(-13.63) (2.99) (0.65) (2.15) (3.17) (3.30) (0.76) (-2.24) (20.95)

1995-1997 -0.00164 0.00212 -0.00040 -0.00060 -0.00030 0.00073 0.00060 0.00011 0.00940(-18.13) (1.51) (-0.42) (-1.20) (-0.76) (5.06) (6.09) (3.14) (20.84)

1998 - March 2000 -0.00283 0.00615 0.00401 0.00257 0.00206 0.00248 0.00132 -0.00007 0.00892(-10.32) (4.07) (4.21) (4.43) (8.26) (9.65) (7.51) (-1.95) (12.10)

April 2000 - 2001 -0.00034 -0.00196 -0.00068 -0.00025 -0.00002 0.00021 0.00030 0.00004 0.00153(-8.60) (-4.66) (-3.55) (-2.25) (-0.40) (8.01) (7.85) (3.38) (12.36)



40

Table VI: Large Stock Open Buy Call Volume Regr. for Full-Service Customers

Time Period Intercept Rsameday Rweek Rmonth Rquarter Ryear R2years ln(B/M) Volatility

1990-1994 -0.00519 0.05502 0.01482 0.00937 0.00630 0.00309 0.00092 0.00092 0.04506(-6.78) (5.15) (2.01) (2.53) (2.40) (3.38) (1.71) (4.24) (10.92)

1995-1997 -0.00616 0.04728 0.00586 0.00142 -0.00084 -0.00030 0.00013 0.00011 0.04455(-20.70) (7.06) (1.31) (0.59) (-0.48) (-0.68) (0.42) (0.52) (27.42)

1998 - March 2000 -0.00510 0.02975 0.00795 0.00450 0.00345 0.00407 0.00269 0.00070 0.02740(-12.36) (5.21) (2.43) (3.09) (4.57) (10.00) (8.17) (4.39) (13.90)

April 2000 - 2001 -0.00084 0.00457 -0.00201 -0.00072 -0.00024 0.00064 0.00076 -0.00006 0.00575(-5.99) (2.10) (-2.72) (-1.64) (-1.07) (6.58) (4.96) (-0.67) (16.78)



41

Table VII: Impact on Open Buy Call Volume of One Std. Dev. Shock

Subperiod Rweek-R2years Rweek-Rquarter Ryear-R2years

1990 - 1994 48.18 32.81 15.37

1995 - 1997 36.26 -13.60 49.86

1998 - March 2000 153.91 58.00 95.91

April 2000 - 2001 31.01 -16.86 47.88

1990 - 1994 82.06 54.67 27.39

1995 - 1997 5.80 6.77 -0.97

1998 - March 2000 86.43 32.44 53.99

April 2000 - 2001 14.68 -11.57 26.25




42

Open Buy Call Volume Across the Subperiods

Over the whole period there is trend-chasing, but the results are not consistent across the subperiodsThe height of the bubble period definitely stands out Coefficients on past returns are all positive and highly

significant For discount customers a one standard deviation shock to

return variables increases open buy call volume by 154% Longer term past returns are especially important Suggests least-sophisticated investors contributed to run-up in

prices of stocks that did well in the past

43

Open Buy Call Volume Across the Subperiods: Full-service

Table V (looking at open buy call volume unconditionally) makes it look like they did not increase their positive speculation during the bubbleTables VI and VII show that they did not escape the frenzyDuring the height of the bubble they chased better performing stocksPost-bubble there is a dramatic change in behavior for both full-service and discount

Contrarian with respect to short-horizon past returns Impact of longer horizon past returns is much more muted

44

Table IX: Growth/Value Stock Average Daily Open Buy Call Vol.

Time Period Firm Discount Full-Service

1990-1994 0.00105% 0.00070% 0.00633%1995-1997 0.00080% 0.00049% 0.00382%

1998 - March 2000 0.00066% 0.00054% 0.00312%April 2000 - 2001 0.00056% 0.00032% 0.00175%

1990-1994 0.00073% 0.00046% 0.00382%1995-1997 0.00075% 0.00092% 0.00440%

1998 - March 2000 0.00067% 0.00188% 0.00482%April 2000 - 2001 0.00055% 0.00050% 0.00216%

Open Buy Call VolumeInvestor type

Panel A: Large Value Stocks

Panel A: Large Growth Stocks

45

Table IX: Growth/Value Stock Average Daily Open Buy Put Vol.

Time Period Firm Discount Full-Service

1990-1994 0.00045% 0.00021% 0.00185%1995-1997 0.00029% 0.00008% 0.00101%

1998 - March 2000 0.00031% 0.00009% 0.00096%April 2000 - 2001 0.00043% 0.00008% 0.00086%

1990-1994 0.00029% 0.00012% 0.00133%1995-1997 0.00021% 0.00018% 0.00137%

1998 - March 2000 0.00027% 0.00033% 0.00150%April 2000 - 2001 0.00068% 0.00014% 0.00105%

Open Buy Put VolumeInvestor type

Panel A: Large Value Stocks

Panel A: Large Growth Stocks

46

Value Versus Growth: Discount Customers

Buying calls is the most important activity for all subperiodsGrowth Stock Open Buy Call Volume

Doubles from pre-bubble to beginning of bubble Doubles again at height of bubble Declines by a factor of four post-bubble

Value stocks: Did not increase volume that opens new callsThus, option market evidence is that during height of bubble discount customers dramatically increased their speculation that growth stock prices would rise but did not increase speculation that value stock prices would rise

47

Value Versus Growth: Full-Service/Firm Proprietary

Mild increase in open buy call activity on underlying growth stocks during the bubbleA reduction in open buy call activity on underlying value stocks during the bubbleFor firm proprietary traders the bubble period is a non-event in terms of their option activity

48

Table X: Open Buy Call Volume Regressions, 1998-3/2000

Underlying Stocks Intercept Rsameday Rweek Rmonth Rquarter Ryear R2years ln(B/M) Volatility

Large -0.0028 0.0062 0.0040 0.0026 0.0021 0.0025 0.0013 -0.0001 0.0089(-10.32) (4.07) (4.21) (4.43) (8.26) (9.65) (7.51) (-1.95) (12.10)

Large, Growth -0.0085 0.0124 0.0074 0.0047 0.0041 0.0046 0.0014 -0.0017 0.0098(-7.61) (3.94) (4.07) (3.77) (7.25) (7.42) (5.23) (-5.95) (6.69)

Large, Value -0.0005 0.0034 0.0026 0.0010 0.0002 -0.0003 0.0004 0.0002 0.0036(-7.79) (3.53) (4.10) (2.90) (0.66) (-2.56) (3.64) (2.93) (10.91)

Large -0.0051 0.0297 0.0080 0.0045 0.0035 0.0041 0.0027 0.0007 0.0274(-12.36) (5.21) (2.43) (3.09) (4.57) (10.00) (8.17) (4.39) (13.90)

Large, Growth -0.0117 0.0336 0.0065 0.0066 0.0063 0.0068 0.0012 -0.0021 0.0218(-7.80) (4.53) (1.83) (3.06) (5.90) (7.10) (2.43) (-4.77) (9.60)

Large, Value -0.0008 0.0338 0.0120 0.0031 0.0020 -0.0022 0.0037 0.0036 0.0196(-1.79) (4.18) (3.07) (1.56) (1.25) (-2.96) (3.88) (5.09) (7.75)




49

Table XI: Percent Impact of One Standard Deviation Shock

UnderlyingStocks Rweek-R2years Rweek-Rquarter Ryear-R2years

Large Growth 210.95 94.02 116.92Large Value 50.40 46.73 3.67

Large Growth 109.12 48.08 61.04Large Value 51.28 38.41 12.87




50

Cross-Sectional Determinants of Open Buy Call Volume During the Bubble

Discount: One standard deviation shock to past return variables results in a 211% increase in open buy call volume on growth stocks and 50% increase on value stocksFull-Service: 109% for growth stocks and 51% for value stocksThere was substantial trend-chasing across the board during the bubble, but it was most intense for Growth stocks Customers of discount brokers

51

Conclusion

Long put positions are relatively unimportant: Only ¼ of long call positions and did not increase during the bubble periodShort call positions are more prevalent than long call positions among non-market makers Covered-calls result in losses in fewer states of

the world than stock positions and only give a lower payoff than stock alone when the stock has increased in price

Consistent with loss aversion and narrow framing

52

Conclusion -- ContinuedShort put positions are more prevalent than long put positions – especially so on value stocksAll three investor classes display trend-chasing behavior in their purchases of calls to open new positions. Returns as far back as 2 years in past matterLeast sophisticated investors dramatically increased purchases of calls to open new positions during the bubble, and this activity was focused in growth stocksNone of the investor groups increased purchases of puts during the bubble. Even when securities are available, investors do not seem to have the courage to bet against a bubble

Option Market Activity and Behavioral Finance

Documents

Transcript of Option Market Activity and Behavioral Finance