Stochastic Modeling

Categories: Metrics

Stochastic Modeling is a process in which we try to track, or possibly predict, the behavior of a whole series of inter-connected variables. Typically, a simulation will be created that models the behavior of a variable we’re interested in tracking, like the future price or possible returns on an investment.

This variable will almost depend on a whole variety of other variables, like confidence in the company, the financial position of the company, the overall market performance, etc. We’ll run the simulation with a certain set of initial conditions and determine the resulting effect on our stock price. We’ll then slightly, or even drastically, change one or more of those initial conditions and re-run the simulation. We’ll keep doing this over and over to get a whole profile of predicted values for our variable. Analyzing that profile can give us insight into how the stock price might actual shake out in the end.

Example:

Dirt, Inc.’s stock has been rising pretty steadily over the last six trading days, since they announced a breakthrough in their dirt production process...which has led to tons of investors jumping on board as buyers driving up the price. You want to hitch your wagon to the gravy train.

But what if the price gains are slowing? In other words, the price is still increasing, but increasing by less and less each day, indicating that the stock is possibly reaching a maximum value before recorrecting downward? If we could also measure the rate at which the prices are changing, we might be able to predict when or if the price is going to reach a maximum value.

Now say you’re the getaway guy for a bank robbery, but the pressure gets to be too much for you, and you take off. Now, what if we not only track the speed of your car, but also the rate at which the speed is changing? Your car might still be speeding up from 50 mph to 60 mph to 65 mph to 67 mph, but it’s speeding up at a slower and slower rate. The decreasing rate at which the speed is changing indicates that you, the driver, might be reaching a top speed soon. We stress the “might” because it’s not a guarantee here...nor is it a guarantee in forecasting future movements.

So, getting back to stocks. It’s absolutely true that changes in the rate of change of a stock price can often precede the stock price reaching a highest or lowest point...or even topping out and then dropping or bottoming out, and then rising. Stochastic analysis is a great way to forecast possible price changes. We’re getting a look under the hood of the stock... not just at the actual performance of the stock, but at the rate at which that performance is changing. Effectively, we’re getting not only a look at the way the stock looks in the mirror, but also a look at its inner workings. Like an x-ray. What we see in the mirror is just a chart of closing prices plotted over time. What we see in the stochastic analysis x-ray machine is a measure of the rate at which those prices are changing.

Related or Semi-related Video

Finance: What is Stochastic Analysis?0 Views

00:00

Finance Allah Shmoop What is Stochastic analysis Stochastic analysis is

00:11

a method of determining the rate at which the value

00:14

of a random variable like the price of a stock

00:17

is changing But obviously we contract the price of a

00:20

stock like every trading platform ever devised Contract the closing

00:24

prices of the stock and plot um for us But

00:27

there's more to it than just tracking the stock price

00:29

and reaching some arbitrary price point And then you know

00:32

selling or buying based on what some crystal ball tells

00:35

you to do What about the speed at which the

00:38

price is changing Dirt inks Stock has been rising pretty

00:42

steadily over the last six trading days since they announced

00:44

a breakthrough in their dirt production process which has led

00:47

to tons of investors Jumping on board Is buyers driving

00:51

up the price right That's what drives up prices more

00:54

demand than supply You want to hit your wagon to

00:56

the gravy train But what if the price gains are

00:58

slowing In other words the price is still increasing but

01:02

increasing by less and less each day indicating that the

01:05

stock is possibly reaching a maximum value before Oh no

01:09

then maybe a correcting downward If we could also measure

01:13

the rate at which the prices air changing Well we

01:15

might be able to predict when or if the price

01:18

is going to reach a maximum value Like example Let's

01:22

say you're the getaway guy for a bank robbery but

01:24

the pressure gets to be too much for you and

01:27

you take off the stranding your former friends Now what

01:30

if we not only track the speed of your car

01:33

but also the rate at which the speed is changing

01:36

Your car still might be speeding up from fifty miles

01:39

an hour to sixty sixty five sixty seven but it's

01:43

speeding up at a slower and slower rate But the

01:46

decreasing rate at which the speed is changing indicates that

01:49

you the driver might be reaching a top speed Soon

01:53

We stress the might because it's not a guarantee here

01:56

Nor is it a guarantee in forecasting future movements So

01:59

getting back to sea stocks It's absolutely true that changes

02:02

in the rate of change of stock price can often

02:04

precede the stock price reaching a high point or a

02:06

low point or even topping out and then dropping or

02:09

bottoming out and then rising again right It's kind of

02:12

random Well stochastic analysis is a great way to forecast

02:15

possible price changes We're getting a look under the hood

02:18

of the stock Not just that the actual performance of

02:21

the stock but at the rate at which that performance

02:24

is changing kind of helps investors attach odds of a

02:27

stock going up and hitting some point or going down

02:30

hitting some other bad point And they can actually then

02:32

make bets that kind of reflect the odds or what

02:35

they think will happen effectively We're not only going to

02:37

look at the way the stock looks in the mirror

02:39

but also look at it inner workings like an X

02:42

ray we see in the mirror It's well just chart

02:45

of closing price is plotted over time What we see

02:47

in the stochastic analysis X ray machine is a measure

02:50

of the rate at which those prices are changing So

02:52

let's start by acknowledging the elephant in the room How

02:55

does measuring the mo mentum of the stock help us

02:58

We'll take a look at the closing prices of a

03:00

big fat growth mojo company Facebook For a time the

03:03

closing prices of Facebook were closing higher and higher values

03:07

each day from left to right across this timeline we

03:09

see that the prices in general continually closing at higher

03:12

and higher values But if we examine the growth we

03:15

see it doesn't always happen at the same rate For

03:17

the first half of this plot well we can see

03:19

the price increases from one seventy knew about one seventy

03:21

eight a difference of eight bucks Check out the middle

03:24

of the graph about three quarters the way across there

03:26

the price started one seventy eight The middle went upto

03:28

one eighty three a difference of five bucks While the

03:30

price is still increasing but not as fast it went

03:33

up fewer dollars in less time than the first time

03:36

right from three quarters of the way to the end

03:39

It's still increasing but now at an even slower rate

03:42

that about three quarters the way along the plot we

03:44

were at price of one eighty three And crossing that

03:47

final quarter on Lee showed growth of another in to

03:50

wish bucks right Well the rate at which the prices

03:52

are climbing is slowing clearly In other words the moment

03:56

um of the stock price is decreasing from what it

03:58

was at the far left Over here we can mash

04:01

the accelerator the floor and our speed goes up very

04:04

quickly Or we can use a light touch on the

04:06

gas pedal and increase our speed slowly The first case

04:09

our speed has large momentum In the second case our

04:11

speed has little mo mentum There are four obvious patterns

04:15

related to a stocks mojo which are calculated using stochastic

04:18

analysis that we should be aware of There are other

04:20

more subtle patterns here is well but well we're just

04:23

going to skip him for now But stocks price could

04:25

be climbing from day to day but the momento of

04:27

the stock could be decreasing In other words the rate

04:30

of growth is slowing This could but doesn't have to

04:33

indicate that the price is going to reach a peak

04:35

and then begin to drop The stock's price could be

04:37

decreasing and the mo mentum of the stock also decreasing

04:41

In other words the rate of the decline is slowing

04:43

down This might be a clue that the price will

04:45

bottom out soon rebound and start to increase again again

04:48

with the word might here being the key words No

04:51

guarantee ever particularly in stock picking We could have a

04:53

stock with a continually increasing price profile This stock could

04:56

also have an increasing MO mentum profile In other words

04:59

the price is going up Mohr each time Interval This

05:03

might be the ultimate siren Call that a stock is

05:05

headed to unheard of heights and we want to own

05:08

it But lastly we might have a stock whose price

05:10

is dropping daily The mo mentum might also be increasing

05:13

In other words the price is dropping at a rate

05:16

more and more and more each time interval which might

05:19

mean we're headed for a berth on the Titanic Go

05:21

get your violent So how do we actually calculate the

05:25

mo mentum or speed at which the price is changing

05:28

Well almost every investing platform whether professional or for the

05:31

home investor like you and me does it for you

05:34

already But it can also be done by hand The

05:37

first process involves calculating the stochastic oscillator That's the percent

05:41

K thingy there which is a direct measure of stocks

05:44

Mo mentum the lowest low closing price and highest high

05:48

closing price are just what they sound like the lowest

05:50

lowest lowest closing price stock took on over our time

05:53

period typically a fourteen days We look at things here

05:56

the highest highs the highest closing price the same time

05:58

period Once we have a percent K for a bunch

06:01

of successive time periods we also need a simple well

06:03

three day moving average of percent Cave denoted by percent

06:06

D This moving average will take the first three values

06:09

of percent k and find their simple average And we'll

06:11

drop the first of the three percent K values and

06:13

add the first unused value of percent k next in

06:16

line and find a new average and so on Until

06:18

we've used all the percent K values what the two

06:21

sets of values percent Kane percent D are plotted on

06:24

the same graph checkout line graph of the bottom of

06:26

this chart and then we start interpreting Seriously though no

06:30

one does it by hand but we will well because

06:32

of the nature of how it's calculated The value of

06:34

K will always be between zero and one hundred So

06:36

certain way points between zero hundred become indicators of whether

06:39

a stock might either be over bottom or over Seoul

06:42

a stock that's over bots one that analysts believe that

06:45

for a variety of reasons is trading above its actual

06:48

value They believe it may be due for a downward

06:50

price correction at some point in the future Example a

06:53

company experiences a bump in stock price after a press

06:55

release saying they have a new hotshot CEO coming on

06:57

board to right the ship While the stock price may

07:00

rises people jump on board but may also auto correct

07:03

downward Once the company gets back to business as usual

07:06

a stock that is oversold is considered to be priced

07:09

below its actual value The price is probably likely then

07:13

to rebound upward in the opposite direction right But let's

07:15

say a report untruthfully link the company to human rights

07:19

violations in one of their overseas factories The stock price

07:22

might drop is an immediate overreaction Until the real story

07:25

comes out exonerating the company well At this point the

07:28

price would probably rebound right well Stocks that have a

07:31

percent K above eighty are considered overbought while stocks of

07:34

the percent K below twenty or considered oversold The actual

07:38

value percent K however isn't the only way we can

07:41

use to casting analysis help us track of stocks momentum

07:44

and also compare the values of percent k the stochastic

07:47

oscillator there in the present di the simple three day

07:49

moving average of percent K to gain other insights into

07:52

a stocks mojo When the two lines present Cade present

07:55

here plotted together they'll often be very close to each

07:57

other But time to percent K line will either dropped

08:00

down through the present the line or cross upwards through

08:03

the present deal A line When we have sent Kay

08:05

dropped down through the venti This is signaled that the

08:07

price may be on its way up When the percent

08:09

K line rises up through the present the line well

08:11

possibly signals a drop in price Or rather the odds

08:14

get higher based on these patterns Now because some of

08:17

us don't have the memories of elephants let's recap a

08:19

bit stochastic analysis and specifically the stochastic oscillator measure the

08:24

speed at which a value like the stock price is

08:26

changing Often called mo mentum the stochastic oscillator percent K

08:30

and a three day moving average plenty or typically calculated

08:33

for us and plotted over a period of days weeks

08:35

or even months when percent K gets above eighty the

08:38

stocks considered over bought and once below twenty it's considered

08:41

oversold Also the behavior of the percent came percent D

08:44

lines taking together can produce signals to either buy or

08:47

sell Remember it's not just about the trend in the

08:50

stock price itself It also matters what's going on inside

08:53

the stock in terms of the momentum of the price

08:55

And in a showdown of these trading patterns versus fundamental

08:58

activities in the company like how well or poorly business

09:01

is doing the latter always wins The latter always directs

09:05

the stock price over just these simple patterns that statistician's

09:08

like the point to to make up little games in

09:10

their heads So that's stochastic analysis so complicated A cave 00:09:14.162 --> [endTime] man can't do it you know Sorry there Aug

Find other enlightening terms in Shmoop Finance Genius Bar(f)