Documente Academic
Documente Profesional
Documente Cultură
You should understand by now that given any portfolio variance, leverage
p p y L2((V).
modifies the volatilityy proportionately ) This means that a 5% loss becomes
a 50% loss in this scenario (ask students about the math of this).
Of those covariances go from near zero to one in a market panic, you are dead.
Suppose we take a market basket of stocks, like the index above (but not
restricted to indexes ... this is an example), and suppose we are bullish on the
index, but not every stock in the index. We therefore decide to go long in most
stocks but short some of them (shown in red in the example above, based upon
l t 2007 markets).
late k t) BButt iin what
h t quantities?
titi ? This
Thi is
i where
h statt t arb
b begins,
b i with ith a
portfolio of longs and shorts, like any hedge fund. But stat arb also aspires to be
market neutral.
Stat Arb Market Neutrality
Statistical Arbitrage builds an equity portfolio of longs and shorts (see
previous slide) but stat arb also typically aspires to be "market neutral." This
means that
h ideally
id ll the h portfolio's
f li ' performance
f is
i not improved
i d by
b either
i h a rise
i
or decline in the overall index (or portfolio). In the previous example, this
portfolio, if correctly weighted to be "market neutral," would be insensitive
to a rise or fall in the DJIA
DJIA. [Your teacher once asked a stab arb desk
supervisor, a Mudder, whether he thought the market was going to rise or fall
in emerging months. His answer was, "Who cares?"].
How do they achieve market neutrality? On the risk side, they use some very
complicated algorithms to assess portfolio risk to make the portfolio risk
neutral. We will learn that approach in a few weeks when we cover advanced
volatility.
l tilit
What therefore is the basis of profitability? Generally, you have to be more or
less right about the longs and the shorts. In a rising market, the longs have to
rise more than the shorts. In a falling market, the shorts have to fall more than
the longs.
Stat Arb Problems??
Stat arb has been seen in recent years as the coolest of the quant strategies, but in
the fall of 2007, in the midst of the sub-prime meltdown, that some stat arb
h d ffunds
hedge d hhad
d bbeen llosing
i money, something
hi they
h were not supposedd to do. d
At issue (teacher's comments):
1. Theyy are heavily
y leveraged.
g
2. Their borrowing source is (was) a troubled market.
3. Some bullish bias may have been built into some of these funds.
4. "Quantagion?"
Financial Contagion
g or "Quantagion"
Q g
(a dangerous new phenomenon?)
[Not required, but recommended, for class]: see "What Happened to the
Quants in August 2007?" by Amir E. Khandani and Andrew W. Lo,
September 20, 2007 draft available on
http://www2.hmc.edu/~evans/khandanilo.pdf
l
later versions
i possibly
ibl available
il bl on
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1015987.
Calculate
Reading the Options Chain
IBM's stock info
Out of the
Money
Expiration
p In the Bid/Ask
date St ik Prices
Strike Pi
Money same as
stocks
You can buy the IBM November 19 195 You can buy the IBM November 19 175
Call for $1.82 (OOM), which gives you Put for $8.05 (ITM), which gives you
the right to buy IBM for $195 per share the right to sell IBM for $175 per share
between now and Nov 19. between now and Nov 19.
Note the big Bid/Ask – the less the liquidity the bigger
Source: these spreads.
Potential Call Option Values
(upon expiration)
$12.00
$6.00 This
hi is
i the
h Nov 19 (exp)
( ) IBM OOM
195 Call, purchased at $1.82 (BA)
$4.00 on Sep 29, when IBM was $177.62
(last)
(last).
$2.00
$0.00
Profit/Loss Gross value
The all
all--important p
premium
remium on OTM options
The premium for an in-the-money option
converges to zero as the option
approaches expiration.
The
h premium
i off an out-of-the-money
f h option
i can be b thought
h h off as simply
i l
the price of the option because the option has an intrinsic value of 0 at the
moment.
0.80
0.60
0.40
0.20
0.00
33 32 31 30 29 28 27 26 25 24 23 22 21 20 19 18 17 16 15 14 13 12 11 10 9 8 7 6 5 4 3 2 1
Issue 2: The Impact
p of Volatility
y
Put Option Price Calculator 3.50
Daily Volatility
3 00
3.00
Stock symbol: TLT
Put option: Oct 106 2.50
Date Today: 8/25/2010
Expiration Date: 10/16/2010 2.00
DTM: 52
1.50
Stock Price: 108.14
Strike Price: 106.00
Daily Volatility: 0.0117 1.00
Interest Rate: 0.010
Ti
Time: 52 0 50
0.50
Source: TD Ameritrade The call option that you buy must have adequate
Option Chains, Oct. 6, 2011 open interest (at least a few hundred contracts).
Complex Strategy 2: Strangles
You can buy the IBM November 19 195 You can buy the IBM November 19 175
Call for $1.82 (OOM), which gives you Put for $8.05 (ITM), which gives you
the right to buy IBM for $195 per share the right to sell IBM for $175 per share
between now and Nov 19. between now and Nov 19.
Do you remember
D b ththese from
f th
the last
l t lecture?
l t ? If we
do both it is a strangle!
Source:
Performance and Profitabilityy of the Strangle
g
$20.00 Note: This graph is somewhat misleading. It show the profit and
value of the position on the expiration date only if the option is
$15 00
$15.00
held to expiration. It shows nothing about the possible value of an
option between now and the expiration date.
$10.00
P fi
Profit Profit
$5.00
$0.00
Loss
‐$5.00 Clearly y yyou are
playing volatility
‐$10.00 here. This example is
a little asymmetric.
‐$15.00
150 155 160 165 170 175 180 185 190 195 200 205 210 215 220
We did this in Econ 136 in 2007.
This above is a covered call I wrote specifically for this class last year. I
bought AeroVironment (AVAV) last Spring for an Econ 136 experiment
for around 26. I should have sold it when it popped above 35 but didn’t.
So I wrote this call for us to track until November 22.
Because of high volatility this was expensive for the buyer. I pocket
$1.50 per share and would actually prefer that this be exercised, allowing
me to
t also
l pocketk t a $4 cap gain
i ($1.41
($1 41 had
h d I bought
b ht the
th stock
t k then
th written
itt
the call). The $2.91 gain if executed is more than 10% absolute – not bad
for 50 days. If it doesn’t execute I will just write another call.
0.50
0.00
-0.50
-1.00
SPY: 187.44
W: 2 187 for
2.02
B: 1 184 for
4.21
B: 1 191 for
0 39
0.39.
Net: $1.80
Exam 2 question: When would you use this strategy and what role is played by
each leg??
Butterfly (Call) payoff grid
4.00
SPY Butterfly 3/4/14
SPY stock 187.44
Mar 184 Call 4.21 Buy
3.00
Mar 187 Call 2.02 Write 2X
Mar 191 Call 0.39 Buy
Net -0.56
2.00
1.00
0.00
-1.00
-2.00
180 181 182 183 184 185 186 187 188 189 190 191 192 193
Net Gross
Butterfly (Put) payoff grid
5.00 SPY Put Butterfly 3/4/14
On a butterfly, it doesn’t SPY stock 187.44
much matter whether you Mar 184 Put 1.06 Buy
4.00 Mar 187 Put 2.06 Write 2X
use calls or puts,
puts payoff
Mar 191 Put 4.72 Buy
format is about the same.
Net -1.66
3.00
2.00
1.00
0.00
-1.00
-2.00
180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195
Net Gross
What you must do in HW4/5
STRIKE PRICE PROBABILITY
CALCULATOR Time is frozen to March 4, 2014, 8:52 AM.
Analyst'ss Name:
Analyst Name:
At that moment SPY is tradingg for 187.410.
Date Today: 3/14/14 You have bought a March 22 Call option
Stock
Strike Price 190 at Best Ask for $0.63.
Stock Symbol: SPY
Stock Price: 187.410 There are 8 days to expiration
expiration.
Option
Option Strike Price: 190 You are to use the one-year historical
Option Expiration Date: 3/22/14
Option Price (Best Ask):
p ( ) 0.630 volatility from HW1 to calculate:
Days to expiration: 8
Historical Data (1) The probability that the stock price will
Log Growth Rate Price to Strike
Price:
be higher than the strike price at
Sample Mean Daily Growth Rate:
Sample Mean Daily Growth Rate: 0 000000
0.000000 expiration assuming a zero alpha,
alpha and
Sample Daily Standard Deviation: (2) The probability that the stock price will
Time‐adjusted Standard Deviation:
Calculation
be higher than the strike price at your
Probability of Stock Price Greater estimated alpha (which may produce the
than Strike Price at Expiration: same result).
l)
25.00 1.20 When you ask the question, “What is
The NormBase file ... the prob that price will go from 100
1.00
20 00
20.00 to 105?
105?”, you are also asking “How
How
0.80 many standard deviations is 105
15.00
away from 100, and what is the
0.60
probability of that?”
10 00
10.00
0.40
Ranges Mean to Value Cumulative Probs
+/- 3 0.9973 0.4987 -0.0400 -2 0.0228
5.00
0.20 +/- 2 0.9545 0.4772 -0.0200 -1 0.1587
+/-1.5 0.8664 0.4332 0.0000 0 0.5000
+/-1
+/ 1 0 6827
0.6827 0 3413 0.0200
0.3413 0 0200 1 0 8413
0.8413
0.00 0.00
-0.08 -0.06 -0.04 -0.02 0 0.02 0.04 0.06 0.08
+/-0.5 0.3829 0.1915 0.0600 2 0.9772
3.00 0.40
Probability of X
2.00 105 is 0.164581
0 20
0.20
greater than:
1.00
0.00 0.00
82 85 88 91 94 97 100 103 106 109 112 115 118 121 Helpful to understand ....
Name: Gary R. Evans
Date: March 4, 2014
... a small segment of my Visual Basic
code.
Strangle Implied Daily Volatility Calculator
Calculate