Popular in Professional

Architecture

Connecting Cities and Nature

University of Virginia Timothy Beatley

Human Rights Law

Human Rights and Corruption: Problems and Potential of Individualising a Systemic Problem

Max Planck Institute Anne Peters

Finance

Chinese Exchange Rate Policy: Lessons for Global Investors

University of California San Diego Michael Melvin

Law

A statistical approach to law school citation rankings

University of Virginia Joshua Fischman

Education

Firm Performance Pay as Insurance against Promotion Risk

Stockholm School of Economics Alvin Chen

Constitutional Law

What is constitutional ideology?

University of Oxford Nick Barber

The Early Exercise Risk Premium

View more insights in Professional - Finance

View more figures from University of Manchester

View the page of Kevin Aretz

Read the Study

Next in Finance

Capital Commitment

Elise Gourier, Ludovic Phalippou | ESSEC Business School

The Dodd-Frank Act and Hedge Fund Operational Risk

William N. Goetzmann | Yale University

Kevin Aretz discusses asset pricing implications of being able to optimally early exercise plain-vanilla puts, contrasting expected raw and delta-hedged returns across equivalent American and European puts.

Image courtesy of interviewee. January 29, 2024

Log-in or Sign-up to Faculti
Currently viewing this subject insight as a guest. You have insight(s) remaining for this month. Login to view 8000+ figures on the platform.

Generating Workspace...

Export as PDF

This video video provides an in-depth analysis of the profitability of options investment strategies, specifically focusing on single stock put option strategies. The study distinguishes between American options, which can be exercised at any time before expiration, and European options, which can only be exercised at maturity. The authors introduce the concept of the "early exercise risk premium," which is the expected return difference between equivalent American and European put options due to the flexibility of early exercise.

Through theoretical exploration and Monte Carlo simulations, the authors find that the early exercise risk premium is positive under reasonable market conditions. They also identify direct relationships between this premium and factors such as the financial "moneyness" of options, time to maturity, and underlying asset volatility.

Testing yielded two primary conclusions: the early exercise risk premium is indeed positive, and it correlates positively with options that are likely to be exercised early.

In examining 15 prior strategies that suggest profitable trading methods with put options, the authors replicate these strategies and find they continue to yield significant profits. However, allowing for optimal early exercise changes profitability outcomes for many strategies. Notably, nine strategies become more profitable when considering early exercise, while five lose their profitability altogether.

Overall, the study emphasizes the importance of accounting for the early exercise feature of American options in studies about option profitability. It cautions investors against assuming the profitability levels reported in earlier literature are accurate, as some strategies may be less effective than previously indicated.

Language:

Export as PDF

X-TIMESTAMP-MAP:LOCAL=00:00:00.000,MPEGTS=900000

00:00.185 --> 00:03.965
In our study, we, take a look into the profitability

00:04.105 --> 00:07.325
of options, investment strategies in particular

00:07.875 --> 00:09.045
into the profitability

00:09.425 --> 00:12.085
of single stock put option strategies.

00:12.825 --> 00:15.965
Um, a put option is a financial instrument which gives its

00:15.965 --> 00:20.485
owner the right to sell some underlying asset, in our case,

00:20.645 --> 00:23.365
a single stock at a pre-specified price.

00:23.505 --> 00:26.725
The strike price over a pre-specified period,

00:27.305 --> 00:28.645
the time to maturity.

00:29.405 --> 00:30.725
A good example of one

00:30.725 --> 00:34.325
of the better known options investment strategies comes from

00:34.405 --> 00:36.205
a study by, Goyal

00:36.205 --> 00:39.845
and Reto published in 2009 in the Journal

00:39.865 --> 00:41.165
of Financial Economics.

00:41.555 --> 00:43.925
They advised their readers to buy

00:44.145 --> 00:48.325
to long options written on stocks with a high realized

00:48.385 --> 00:50.165
to imply volatility ratio

00:50.585 --> 00:53.885
and to short to sell those options on stocks

00:53.885 --> 00:55.165
with a low ratio.

00:55.545 --> 00:59.325
So the, realized to implied volatility ratio here

00:59.865 --> 01:03.245
is taken as a measure of the expansiveness of the options.

01:03.545 --> 01:05.765
So they effectively advise their readers

01:05.785 --> 01:09.525
to buy cheap options and to sell expensive ones,

01:09.865 --> 01:11.645
and they show that doing so

01:12.325 --> 01:14.405
generates an economically sizable

01:14.665 --> 01:18.685
and statistically significant positive trading profit.

01:19.385 --> 01:22.605
Now, where we come into this literature is by recognizing

01:22.875 --> 01:25.485
that all the studies in,

01:25.585 --> 01:28.245
the prior literature looking into options,

01:28.245 --> 01:31.085
investment strategies effectively assume

01:31.275 --> 01:33.485
that there underlying options there.

01:33.565 --> 01:36.045
Sample options are European options.

01:36.765 --> 01:39.125
European options can only be exercised.

01:39.125 --> 01:41.685
They can only be used to buy the underlying

01:41.685 --> 01:44.485
to sell the underlying asset at the end

01:44.865 --> 01:46.245
of the time to maturity.

01:46.745 --> 01:49.925
Now this assumption is factually incorrect In the real

01:49.935 --> 01:52.965
world, in real US options exchanges,

01:53.545 --> 01:56.685
all single stock options are American options.

01:57.365 --> 01:59.085
American options are far more flexible

01:59.355 --> 02:00.605
than European options.

02:00.915 --> 02:03.325
They can be exercised, they can be used

02:03.345 --> 02:06.685
to sell the underlying asset at any point over

02:06.705 --> 02:07.725
the time to maturity.

02:08.425 --> 02:12.005
So, my co-author Adnan ga

02:12.105 --> 02:15.485
and I then immediately came up with two questions.

02:15.865 --> 02:20.125
The first question is, does the ability to optimally early

02:20.645 --> 02:24.445
exercise a put option affect the expected return of

02:24.445 --> 02:26.605
that option, both from a theoretical

02:26.865 --> 02:28.325
and an empirical perspective?

02:28.865 --> 02:31.765
And if the answer to this question is yes,

02:32.195 --> 02:35.325
then an immediate follow up question is,

02:35.785 --> 02:38.085
if it affects the expected option return,

02:38.315 --> 02:40.485
does it also affect the profitability

02:41.025 --> 02:43.445
of those options investment strategies

02:43.595 --> 02:45.285
established in the literature?

02:45.465 --> 02:48.725
And that's exactly what we analyze in our paper.

02:49.105 --> 02:51.805
In our theoretical work, we come up with this term,

02:52.185 --> 02:54.005
the early exercise risk premium.

02:54.625 --> 02:56.925
The early exercise risk premium is,

02:57.065 --> 03:01.045
the expected spread return between equivalent American

03:01.105 --> 03:02.685
and European put options.

03:03.025 --> 03:05.845
By equivalent, I mean the options are written on the same

03:05.895 --> 03:06.925
underlying asset.

03:07.515 --> 03:09.285
They have the same strike price

03:09.385 --> 03:11.485
and the same time to maturity.

03:12.185 --> 03:13.525
Um, intuitively speaking,

03:13.625 --> 03:17.205
the early exercise risk premium measures the effect

03:17.545 --> 03:20.525
of being able to optimally early exercise,

03:20.785 --> 03:24.085
put option on the expected put option return.

03:24.705 --> 03:27.165
Now to generate some intuition on the sign

03:27.165 --> 03:28.925
of the early exercise with premium,

03:29.035 --> 03:30.765
whether it's positive zero

03:30.865 --> 03:32.805
or negative, we recall,

03:33.185 --> 03:36.165
par at I 1992 insight

03:36.395 --> 03:39.045
that an American put option can be replicated

03:39.505 --> 03:43.525
by a portfolio containing a long equivalent European put

03:43.525 --> 03:46.325
option and a long exotic derivative.

03:46.665 --> 03:49.525
And the exotic derivative is itself a portfolio,

03:50.005 --> 03:54.285
a portfolio containing an in finite number of binary

03:55.075 --> 03:57.645
cash or nothing but options.

03:58.145 --> 04:00.205
We then recall two further insights.

04:00.305 --> 04:02.925
The first one that the expected portfolio,

04:03.265 --> 04:06.565
return in finance theory is evaluated average

04:06.865 --> 04:09.445
of the expected returns of assets in the portfolio.

04:09.785 --> 04:12.165
And the second that the expected return

04:12.165 --> 04:15.885
of a binary option is usually closer, closer

04:16.025 --> 04:20.005
to zero than the expected return of plain vanilla options.

04:20.345 --> 04:22.285
And the reason for this is that the payoff

04:22.285 --> 04:25.245
of a binary option is just a constant, in our case,

04:25.545 --> 04:28.205
the risk-free rate multiplied by the stripe price.

04:28.505 --> 04:32.205
So the binary option has a systematic risk closer to zero,

04:32.585 --> 04:34.605
and therefore an expected return closer

04:34.665 --> 04:36.925
to zero than a plain vanilla option.

04:37.545 --> 04:42.105
So let's say the European input option has,

04:42.765 --> 04:46.665
for the sake of argument, an expected return of minus 10%,

04:47.365 --> 04:49.745
the exotic derivative, this portfolio

04:49.805 --> 04:54.465
of binary options has a higher expected return minus 5%,

04:54.795 --> 04:57.865
given that the American product option can be replicated

04:57.965 --> 05:00.665
by a portfolio of the long European port option

05:01.165 --> 05:03.745
and the long exotic derivative.

05:04.045 --> 05:07.145
And given that the expected portfolio return is just the

05:07.145 --> 05:09.545
evaluated average of the expected returns

05:09.545 --> 05:12.705
of the assets in the portfolio, it immediately follows

05:12.705 --> 05:16.465
that the American port option should have an expected return

05:16.465 --> 05:20.225
somewhere between minus five and minus 10%.

05:20.485 --> 05:22.545
So the American port option has a higher,

05:23.105 --> 05:25.345
a less negative expected return than

05:25.345 --> 05:26.745
the European port option.

05:27.045 --> 05:29.785
Now unfortunately, this intuition is not

05:29.785 --> 05:31.025
mathematically rigorous.

05:31.205 --> 05:32.465
We cannot prove this.

05:32.845 --> 05:35.985
So what we next do is we conduct a large scale montecarlo

05:35.985 --> 05:37.185
simulation exercise.

05:37.845 --> 05:39.945
So, we simulate the value

05:39.965 --> 05:43.185
of the underlying asset using three stochastic processes,

05:43.475 --> 05:47.145
black and skulls, 1973 geometric in motion.

05:47.645 --> 05:50.705
Two more sophisticated one has since 1993,

05:50.705 --> 05:51.865
stochastic volatility

05:52.365 --> 05:56.305
and Bates 1996 stochastic volatility jump process.

05:57.025 --> 05:58.165
We simulate the value

05:58.165 --> 06:01.605
of the underlying asset using each process under the real

06:01.605 --> 06:04.245
world, the physical and the risk neutral measure.

06:04.585 --> 06:06.925
And we use the outcomes from the simulation

06:06.985 --> 06:10.445
to approximate the expected returns of equivalent

06:11.795 --> 06:13.995
American and European port options.

06:14.215 --> 06:16.515
Now, contrasting these two expected returns,

06:16.855 --> 06:18.835
we reach two important conclusions.

06:19.095 --> 06:20.595
The first one is yes,

06:20.735 --> 06:24.155
indeed the early exercise risk premium is positive

06:24.535 --> 06:27.675
and the reasonable model parameter values, right?

06:27.935 --> 06:30.835
So there is a positive spread in expected returns

06:30.835 --> 06:33.155
between American and European put options.

06:33.615 --> 06:37.115
And second, the early exercise risk premium relates

06:37.635 --> 06:40.715
positively to, the possibility

06:40.935 --> 06:42.875
of the American put being,

06:42.885 --> 06:45.595
early exercised over the time to maturity.

06:46.095 --> 06:48.635
And because this possibility is positively related

06:48.635 --> 06:52.355
to manness and negatively to, time to maturity

06:52.455 --> 06:56.715
and volatility, the early exercise risk premium itself is

06:56.995 --> 06:58.595
positively related to moneys

06:58.855 --> 07:00.755
and negatively to time to maturity

07:01.055 --> 07:02.795
and underlying asset volatility.

07:03.175 --> 07:06.315
And these are exactly the testable implications which we

07:06.315 --> 07:08.075
next take to our data.

07:08.535 --> 07:11.635
Our two main testable implications are first,

07:12.015 --> 07:13.595
the early exercise risk premium.

07:13.735 --> 07:15.955
The expected spread return between American

07:15.955 --> 07:17.915
and European food options is positive.

07:18.295 --> 07:19.355
And the second one,

07:19.575 --> 07:22.355
the early exercise risk premium is positively related

07:22.375 --> 07:25.995
to moneys and negatively to time to maturity

07:26.095 --> 07:27.875
and underlying asset volatility.

07:28.615 --> 07:31.915
We face two challenges in testing these predictions.

07:32.255 --> 07:35.235
The first challenge is that it is not obvious how

07:35.235 --> 07:38.635
to identify optimal early exercises of

07:39.155 --> 07:41.235
American puts in our data.

07:41.735 --> 07:44.715
The literature offers relatively few guidance on this.

07:44.975 --> 07:47.995
We found an outstanding study by Barlow

07:47.995 --> 07:51.075
and Whaley published in 2012 in the Journal of Finance,

07:51.655 --> 07:55.075
and they suggest the, what they call market rule.

07:55.455 --> 07:57.955
Now, the market rule says you should optimally early

07:58.355 --> 07:59.995
exercise an option

08:00.825 --> 08:04.595
when the early exercise payoff exceeds the traded

08:04.685 --> 08:05.835
price of the option.

08:06.895 --> 08:08.195
Now in practice,

08:08.415 --> 08:11.675
the early exercise payoff can never exceed the traded price

08:11.695 --> 08:13.475
of the option because if it did,

08:13.645 --> 08:15.555
there were an arbitrage opportunity

08:16.095 --> 08:18.995
and such blatant arbitrage opportunities do

08:18.995 --> 08:20.075
not exist in our data.

08:20.415 --> 08:22.915
So what we assume is, an

08:23.435 --> 08:26.155
American port should be optimally early exercise.

08:26.615 --> 08:30.955
As soon as the early exercise payoff gets sufficiently close

08:31.015 --> 08:35.515
to the traded option, price comes within a 1%, 2.5%,

08:35.515 --> 08:36.795
5% radius.

08:37.615 --> 08:39.555
The second obstacle, which we face,

08:40.015 --> 08:41.555
and I already alluded to this

08:41.555 --> 08:43.995
before, is that there are no

08:44.715 --> 08:49.115
European single stock options traded in real US

08:49.225 --> 08:50.515
options exchanges.

08:51.135 --> 08:55.115
All single stock options traded in the US in exchanges

08:55.545 --> 08:57.165
are American options.

08:57.785 --> 09:02.285
In order to circumvent this problem, we synthetically

09:02.885 --> 09:05.005
recreate European port options.

09:05.425 --> 09:09.605
How do we do? So we start from Martin's 1973 insight

09:09.755 --> 09:14.205
that an American coil written on an asset not paying out

09:14.205 --> 09:15.245
cash over the time

09:15.265 --> 09:18.885
to maturity should never be optimally early exercised,

09:19.335 --> 09:22.645
which means the early exercise feature is worthless,

09:23.105 --> 09:24.885
you cannot, you should never use it.

09:25.185 --> 09:28.565
And the American call is essentially identical

09:29.105 --> 09:32.125
to a European coil right now, given

09:32.125 --> 09:35.965
that all sample options in our data are written on stocks

09:35.985 --> 09:38.165
not paying our dividends, our

09:38.925 --> 09:42.205
American coils are effectively European coils.

09:42.745 --> 09:44.965
We then use, put coil parity in order

09:44.985 --> 09:49.525
to convert the European coils into European puts.

09:49.865 --> 09:52.165
So put parity as you can,

09:52.475 --> 09:54.765
replicate a European put option

09:54.825 --> 09:58.125
by investments into the equivalent European call option,

09:58.505 --> 10:01.925
the underlying stock and the money market account.

10:02.625 --> 10:05.005
So using our traded American

10:05.105 --> 10:07.965
and our synthetic European put option data,

10:08.585 --> 10:10.285
we then calculate the returns

10:10.385 --> 10:14.365
of optimally early exercised American puts never early

10:14.555 --> 10:17.805
exercised American puts for comparison with the literature

10:18.225 --> 10:20.605
and synthetic European puts.

10:21.435 --> 10:24.375
Now we draw two important lessons from

10:24.445 --> 10:25.815
comparing these returns.

10:26.265 --> 10:28.895
First, the means track return,

10:28.895 --> 10:32.055
between optimally early exercised American inputs

10:32.055 --> 10:34.615
and synthetic European inputs is positive.

10:35.085 --> 10:38.655
It's equal to 2.38% per month.

10:39.005 --> 10:40.735
That is economically huge

10:41.195 --> 10:43.655
and it's also hugely statistically significant.

10:43.655 --> 10:46.455
So it has a key statistic of around five.

10:46.925 --> 10:50.255
Even if we run a sophisticated bootstrap on the T statistic,

10:50.555 --> 10:54.095
we can easily reject the hypothesis that,

10:55.175 --> 10:57.635
the mean spread return is equal to zero.

10:58.735 --> 11:02.755
Um, so the main conclusion is in the data, in the pool data,

11:02.935 --> 11:05.835
the early exercise risk premium is positive,

11:06.415 --> 11:09.555
but then, we also use portfolio sorts

11:09.555 --> 11:11.595
and Pharma Mac best regressions to show

11:11.625 --> 11:12.835
that the spread return

11:12.835 --> 11:15.075
between optimally early exercise American

11:15.415 --> 11:19.035
and synthetic European port options is positively related

11:19.135 --> 11:22.515
to moneyless and negatively to time to maturity

11:23.015 --> 11:26.995
and underlying asset, underlying stock volatility.

11:27.495 --> 11:29.915
So we find support for both

11:29.935 --> 11:32.035
of our main testable implications

11:32.445 --> 11:33.755
after we established

11:34.185 --> 11:37.315
that there is indeed a difference in returns

11:37.315 --> 11:39.075
between optimally early exercise,

11:39.515 --> 11:42.955
American puts never early exercise American puts

11:42.955 --> 11:44.115
and European inputs.

11:44.375 --> 11:46.115
We then returned to those studies

11:46.175 --> 11:49.315
and the literature proposing profitable options

11:49.315 --> 11:50.595
investment strategies.

11:51.015 --> 11:54.445
We identified a total of 12 studies

11:55.095 --> 11:59.965
suggesting 15 of such strategies, involving

12:00.545 --> 12:03.885
raw put options, delta hatch pot options.

12:03.945 --> 12:05.685
So these are put options controlling

12:05.685 --> 12:07.685
for movements in the value of the underlying asset.

12:07.985 --> 12:10.925
And strats strats are,

12:11.055 --> 12:14.765
portfolios containing long positions in the equivalent coils

12:15.065 --> 12:16.925
and puts now each

12:16.925 --> 12:19.165
of these option trading strategies suggests

12:19.165 --> 12:20.365
some sorting variable.

12:20.765 --> 12:23.525
I mentioned the realized to implied volatility ratio.

12:23.785 --> 12:27.885
So in each case, the idea is you go long on those options

12:28.355 --> 12:31.885
with a high sorting variable value or perhaps a low one,

12:32.145 --> 12:35.405
and you go short, you sell those options which have the

12:35.645 --> 12:36.805
opposite value, right?

12:37.385 --> 12:40.165
So to begin with, we then simply,

12:40.535 --> 12:44.405
replicated the 15 investment strategies conducted

12:44.465 --> 12:45.485
in the 12 studies.

12:46.145 --> 12:47.845
The news is good for the studies.

12:47.985 --> 12:51.645
We find that even over our updated sample period ranging

12:51.645 --> 12:56.245
to December, 2022, we still find that all of the strategies,

12:56.425 --> 13:00.965
all 15 continue to be, economically, continue

13:00.985 --> 13:04.445
to create economically large trading profits which are

13:04.445 --> 13:07.245
statistically different from zero.

13:08.105 --> 13:11.325
We then however, introduce one small change

13:11.625 --> 13:15.005
to the investment strategies we allow for the possibility

13:15.145 --> 13:17.285
of an optimal early exercise

13:17.825 --> 13:21.885
in the American puts contained in the trading strategies

13:21.975 --> 13:25.275
where we again, allow for

13:25.875 --> 13:28.475
optimal early exercises using Barlow

13:28.475 --> 13:31.275
and Wale's 2012 trading strategies.

13:31.695 --> 13:34.435
And the re results from this exercise are striking.

13:34.935 --> 13:39.515
We find that for 14 out of 15 trading strategies allowing

13:39.615 --> 13:42.515
for optimal early exercises in the American puts

13:42.905 --> 13:47.235
significantly changes the profitability of the strategies.

13:48.215 --> 13:51.195
Now, whether the strategy becomes more

13:51.255 --> 13:54.915
or less profitable depends on whether there are more highly

13:54.915 --> 13:57.835
likely to be early exercised options in the long,

13:57.935 --> 14:01.235
in the buy lack or in the short the select.

14:01.815 --> 14:03.715
In nine out of 15 cases,

14:03.965 --> 14:06.475
there are more such options in the long lack

14:06.705 --> 14:09.235
that increases the return of the long lack

14:09.375 --> 14:11.475
and the strategy becomes more profitable.

14:11.895 --> 14:13.195
But in five cases,

14:13.415 --> 14:15.395
and that's, one third

14:15.615 --> 14:17.835
of all the investment strategies we look into,

14:18.045 --> 14:20.035
there are more highly likely

14:20.135 --> 14:23.875
to be early exercise options in the short lack in the salt

14:23.985 --> 14:26.915
lack that increases the return of the salt lack.

14:27.095 --> 14:30.555
And so in this case, the strategies become less profitable.

14:30.855 --> 14:33.235
In fact, what we show in our empirical work is

14:33.235 --> 14:36.355
that if once you allow for optimal early exercises,

14:36.645 --> 14:41.115
these five strategies no longer produce economically

14:41.215 --> 14:43.835
or statistically significant trading profits.

14:44.095 --> 14:46.875
Our recommendation is if you are a real world investor,

14:47.085 --> 14:48.995
definitely don't invest into these

14:48.995 --> 14:50.555
strategies. There are two,

14:50.695 --> 14:53.445
Um, main lessons to be taken away from our study.

14:53.985 --> 14:57.005
The first main lesson is that, the ability

14:57.145 --> 15:00.245
to optimally early exercise a put option,

15:00.585 --> 15:04.045
has a large effect on the expected return of the put option,

15:04.515 --> 15:08.325
both from a theoretical and an empirical perspective.

15:08.865 --> 15:11.285
So, studies looking

15:11.915 --> 15:15.525
into put options should definitely take the early

15:16.085 --> 15:19.645
exercise feature of American options into account because

15:19.795 --> 15:23.125
otherwise they produce distorted bias returns.

15:23.585 --> 15:27.125
And then because ignoring the early exercise feature

15:27.145 --> 15:29.045
of American options leads

15:29.065 --> 15:32.525
to distortions biases in their returns,

15:33.735 --> 15:37.195
one should not take at face value the profitability

15:37.535 --> 15:39.075
of the trading strategies

15:39.215 --> 15:42.275
so far adjusted in the existing literature, right?

15:42.575 --> 15:46.675
So some of these strategies nine are way more,

15:46.675 --> 15:49.635
profitable than suggested in the original studies.

15:49.855 --> 15:52.195
But then on the other hand there are five strategies,

15:52.335 --> 15:55.275
one third, which are actually not profitable at all.

15:55.615 --> 15:59.795
And so readers, investors need to be careful,

16:00.045 --> 16:02.515
about the profitability of these strategies.

Evaluating Profitability of Put Option Strategies

00:00:00 - In our study, we, take a look into the profitability of options, investment strategies in particular into the profitability of single stock put option strategies. a put option is a financial instrument which gives its owner the right to sell some underlying asset, in our case, a single stock at a pre-specified price. The strike price over a pre-specified period, the time to maturity. A good example of one
00:00:34 - a study by, Goyal and Reto published in 2009 in the Journal of Financial Economics. They advised their readers to buy to long options written on stocks with a high realized to imply volatility ratio and to short to sell those options on stocks with a low ratio. So the, realized to implied volatility ratio here is taken as a measure of the expansiveness of the options.
00:01:05 - to buy cheap options and to sell expensive ones, and they show that doing so generates an economically sizable and statistically significant positive trading profit. Now, where we come into this literature is by recognizing that all the studies in, the prior literature looking into options, investment strategies effectively assume that there underlying options there.
00:01:36 - European options can only be exercised. They can only be used to buy the underlying to sell the underlying asset at the end of the time to maturity. Now this assumption is factually incorrect In the real world, in real US options exchanges, all single stock options are American options. American options are far more flexible than European options. They can be exercised, they can be used to sell the underlying asset at any point over
00:02:08 - So, my co-author Adnan ga and I then immediately came up with two questions. The first question is, does the ability to optimally early exercise a put option affect the expected return of that option, both from a theoretical and an empirical perspective? And if the answer to this question is yes, then an immediate follow up question is, if it affects the expected option return,
00:02:41 - of those options investment strategies established in the literature? And that's exactly what we analyze in our paper. In our theoretical work, we come up with this term, the early exercise risk premium. The early exercise risk premium is, the expected spread return between equivalent American and European put options. By equivalent, I mean the options are written on the same underlying asset. They have the same strike price

Impact of Early Exercise on Put Option Returns

00:03:12 - intuitively speaking, the early exercise risk premium measures the effect of being able to optimally early exercise, put option on the expected put option return. Now to generate some intuition on the sign of the early exercise with premium, whether it's positive zero or negative, we recall, par at I 1992 insight that an American put option can be replicated
00:03:43 - option and a long exotic derivative. And the exotic derivative is itself a portfolio, a portfolio containing an in finite number of binary cash or nothing but options. We then recall two further insights. The first one that the expected portfolio, return in finance theory is evaluated average of the expected returns of assets in the portfolio.
00:04:12 - of a binary option is usually closer, closer to zero than the expected return of plain vanilla options. And the reason for this is that the payoff of a binary option is just a constant, in our case, the risk-free rate multiplied by the stripe price. So the binary option has a systematic risk closer to zero, and therefore an expected return closer to zero than a plain vanilla option. So let's say the European input option has,
00:04:47 - the exotic derivative, this portfolio of binary options has a higher expected return minus 5%, given that the American product option can be replicated by a portfolio of the long European port option and the long exotic derivative. And given that the expected portfolio return is just the evaluated average of the expected returns of the assets in the portfolio, it immediately follows that the American port option should have an expected return
00:05:20 - So the American port option has a higher, a less negative expected return than the European port option. Now unfortunately, this intuition is not mathematically rigorous. We cannot prove this. So what we next do is we conduct a large scale montecarlo simulation exercise. So, we simulate the value of the underlying asset using three stochastic processes, black and skulls, 1973 geometric in motion.
00:05:50 - stochastic volatility and Bates 1996 stochastic volatility jump process. We simulate the value of the underlying asset using each process under the real world, the physical and the risk neutral measure. And we use the outcomes from the simulation to approximate the expected returns of equivalent American and European port options. Now, contrasting these two expected returns, we reach two important conclusions.

Analyzing Early Exercise Risk Premium in Options

00:06:20 - indeed the early exercise risk premium is positive and the reasonable model parameter values, right? So there is a positive spread in expected returns between American and European put options. And second, the early exercise risk premium relates positively to, the possibility of the American put being, early exercised over the time to maturity. And because this possibility is positively related to manness and negatively to, time to maturity
00:06:56 - positively related to moneys and negatively to time to maturity and underlying asset volatility. And these are exactly the testable implications which we next take to our data. Our two main testable implications are first, the early exercise risk premium. The expected spread return between American and European food options is positive. And the second one, the early exercise risk premium is positively related to moneys and negatively to time to maturity
00:07:28 - We face two challenges in testing these predictions. The first challenge is that it is not obvious how to identify optimal early exercises of American puts in our data. The literature offers relatively few guidance on this. We found an outstanding study by Barlow and Whaley published in 2012 in the Journal of Finance, and they suggest the, what they call market rule. Now, the market rule says you should optimally early
00:08:00 - when the early exercise payoff exceeds the traded price of the option. Now in practice, the early exercise payoff can never exceed the traded price of the option because if it did, there were an arbitrage opportunity and such blatant arbitrage opportunities do not exist in our data. So what we assume is, an American port should be optimally early exercise. As soon as the early exercise payoff gets sufficiently close
00:08:35 - 5% radius. The second obstacle, which we face, and I already alluded to this before, is that there are no European single stock options traded in real US options exchanges. All single stock options traded in the US in exchanges are American options. In order to circumvent this problem, we synthetically
00:09:05 - How do we do? So we start from Martin's 1973 insight that an American coil written on an asset not paying out cash over the time to maturity should never be optimally early exercised, which means the early exercise feature is worthless, you cannot, you should never use it. And the American call is essentially identical to a European coil right now, given that all sample options in our data are written on stocks

Evaluating American and European Put Options

00:09:38 - American coils are effectively European coils. We then use, put coil parity in order to convert the European coils into European puts. So put parity as you can, replicate a European put option by investments into the equivalent European call option, the underlying stock and the money market account. So using our traded American and our synthetic European put option data,
00:10:10 - of optimally early exercised American puts never early exercised American puts for comparison with the literature and synthetic European puts. Now we draw two important lessons from comparing these returns. First, the means track return, between optimally early exercised American inputs and synthetic European inputs is positive. It's equal to 2.38% per month.
00:10:41 - and it's also hugely statistically significant. So it has a key statistic of around five. Even if we run a sophisticated bootstrap on the T statistic, we can easily reject the hypothesis that, the mean spread return is equal to zero. so the main conclusion is in the data, in the pool data, the early exercise risk premium is positive, but then, we also use portfolio sorts
00:11:11 - that the spread return between optimally early exercise American and synthetic European port options is positively related to moneyless and negatively to time to maturity and underlying asset, underlying stock volatility. So we find support for both of our main testable implications after we established that there is indeed a difference in returns between optimally early exercise, American puts never early exercise American puts
00:11:44 - We then returned to those studies and the literature proposing profitable options investment strategies. We identified a total of 12 studies suggesting 15 of such strategies, involving raw put options, delta hatch pot options. So these are put options controlling for movements in the value of the underlying asset. And strats strats are, portfolios containing long positions in the equivalent coils
00:12:16 - of these option trading strategies suggests some sorting variable. I mentioned the realized to implied volatility ratio. So in each case, the idea is you go long on those options with a high sorting variable value or perhaps a low one, and you go short, you sell those options which have the opposite value, right? So to begin with, we then simply, replicated the 15 investment strategies conducted in the 12 studies.

Reevaluating Options Strategies: Early Exercise Impact

00:12:47 - We find that even over our updated sample period ranging to December, 2022, we still find that all of the strategies, all 15 continue to be, economically, continue to create economically large trading profits which are statistically different from zero. We then however, introduce one small change to the investment strategies we allow for the possibility of an optimal early exercise
00:13:21 - where we again, allow for optimal early exercises using Barlow and Wale's 2012 trading strategies. And the re results from this exercise are striking. We find that for 14 out of 15 trading strategies allowing for optimal early exercises in the American puts significantly changes the profitability of the strategies.
00:13:51 - or less profitable depends on whether there are more highly likely to be early exercised options in the long, in the buy lack or in the short the select. In nine out of 15 cases, there are more such options in the long lack that increases the return of the long lack and the strategy becomes more profitable. But in five cases, and that's, one third of all the investment strategies we look into, there are more highly likely
00:14:23 - lack that increases the return of the salt lack. And so in this case, the strategies become less profitable. In fact, what we show in our empirical work is that if once you allow for optimal early exercises, these five strategies no longer produce economically or statistically significant trading profits. Our recommendation is if you are a real world investor, definitely don't invest into these strategies. There are two,
00:14:53 - The first main lesson is that, the ability to optimally early exercise a put option, has a large effect on the expected return of the put option, both from a theoretical and an empirical perspective. So, studies looking into put options should definitely take the early exercise feature of American options into account because otherwise they produce distorted bias returns.
00:15:27 - of American options leads to distortions biases in their returns, one should not take at face value the profitability of the trading strategies so far adjusted in the existing literature, right? So some of these strategies nine are way more, profitable than suggested in the original studies. But then on the other hand there are five strategies, one third, which are actually not profitable at all.