Elon Musk *Critical* Tesla Stock Flip | FSD & Robotaxi

hey everyone me Kevin here so Elon Musk

just dropped the price of full

self-driving to $99 per month he left

the full price option where you just pay

once and then you don't have a

subscription fee at $122,000 which

really is probably a sign that they're

not really converting many of those

$112,000 individuals and so they're

purposefully leaving that price at 12K

because now you're not insulting that

many people who paid 12K recently

although there are going to be some uh

and what you're doing is you're trying

to drive this value for the $99 version

by showing people look it's going to

take you like 10 years to break even if

you spend 12K so why not take this

reasonably priced offer of $99 so now

the question is what is this actually

going to do for the fundamentals of

Tesla does this actually change the

thesis for the stock so what I did is I

put together a spreadsheet and we'll put

it up on screen and it is basically

three

different like impacts that this does

for the stock uh now I purposefully went

out to 2026 for this rather than going

out to you know 2030 or 2040 or

something crazy because I I want to see

what the near-term impact for the stock

could be and really the only way to do

that is going out um honestly 2 and A2

years is already quite a bit the end of

2026 I think if you go out uh any more

than that you're you're not going to see

that immediacy in the stock movement and

I think that's what most people care

about so uh let's talk about this as we

uh look at that Florida beach this why I

love Florida I grew up in Florida uh

okay so the first thing we're going to

do is we're going to look at what I

consider my base case scenario

and there's a big problem that a lot of

people have a lot of people are not

realizing that there are costs

associated with full self- driving a lot

of people think Kevin full self-driving

is going to be 100% 100% to the bottom

line it's the best margin driver ever

there's no capex and it's just going to

be pure

profit so there's realistic and then

there's unrealistic and unrealistic is

no cost there's a lot of cost you have

to keep in mind they just spent probably

somewhere around $3 to $500 million

buying 10,000 h100s and guess what those

are now the old school version of chicks

Blackwell which will be available

towards the end of the year you're

probably able to train FSD four times as

fast some of the newest

chips so you're consistently probably to

have the latest and greatest chips going

to have to spend just on chips alone

probably and and I should just say

graphic chips which that's literally

what FSD is trained off of right

Graphics video graphics like you might

be able to move some of these chat-based

neural Nets off of some of these gpus in

the longer term but you know maybe to

like A6 or something more more specific

to just Tech space but video video

neural Nets it's always going to need

the latest and greatest stuff to really

get better because it's not just a

matter of training at once like getting

to the version you know 1234 that we

have now that's great but really to iron

out the rest of these edge cases we're

just going to need a lot more training a

lot more compute by using a lot more

data like quite frankly the fact that

1233 and

1234 is making left and right turns from

center lanes totally dodging the left

and right turn Lanes is weird there

should be plenty of data to tell the

systems this is not what it's supposed

to do now I'm not saying those things

aren't solvable in fact we could bandage

them tomorrow just by writing a line of

code but that defeats the purpose we're

trying to get away from writing lines of

code so we need this data we need the

training frankly we need better chips to

really advance I think one of the

reasons we got to 12 versus 12 frankly

the anti- neuronet is because the h100s

came out and so obviously that's a mode

for NVIDIA uh but that's just going to

get as the chips get better and better

FSD will get better and better and these

edge cases will go away but that takes

cap X and that's what a lot of people

forget is you're probably throwing 500

to a billion dollars at chips alone once

you consider that we don't just need

gpus but we need CPUs we need memory we

need server racks uh and and then we

need infrastructure uh and then we got

to pay for that infrastructure the

electricity uh the the power the the

people who are maintaining the servers

the people who are upgrading the servers

whatever and then of course you need the

R&D folks who actually uh implement this

into the systems on Tesla vehicles the

software updates to push this to people

uh there's there's a lot that goes into

it so I think initially we're probably

looking at spending quite a bit uh of

fsds Revenue

on R&D so your AI training expense uh

your chips whatever the things I just

mentioned so I think if we go with my

base case scenario we're probably

looking in 2024 if we had a 12 1 12%

take rate keep in mind that's already

adjusting for the fact that some cars

are never going to have F because if you

have a fleet right now of about 82

million vehicles uh worldwide by the end

of the year so that's assuming you know

another 2 million of production we're

probably going to and that might be

generous there's a percentage that are

in other countries where we don't do FSD

there is a percentage that uh are

scrapped vehicles that just got

destroyed I mean obviously and there are

a percentage of people who are just

never actually going to want to buy a

FSC uh because some people don't drive a

lot you know if I'm uh if if I'm

somebody who uh you know if all I do is

I drop my kids off at school 4 minutes

away and I pick them up and that's all I

do twice a day 5 days a week do I re

really need FSD for 10 minutes of

driving a day no of course not so

there's always going to be a percentage

of people who won't get it and so hence

when we Factor all of these things in

the other countries scrap Vehicles

people who just refuse to get it uh and

people who have already paid for FSD so

you're not getting that recurring

monthly Revenue you're going to be

looking at probably a pretty low take

rate initially hopefully it's in excess

of 12 1/2% but let's assume that it's 12

1/2% right now uh and then what we're

going to do is we're going to multiply

that obviously by uh uh the9 $99 a month

time 12 right okay so let's go ahead and

take a look at the base case scenario

and explain how I'm doing this so if you

look at the chart you'll see 8.5 million

yeah that's always loud 8 half million

Fleet size multipli by a 122% take rate

you get the value how much money that

could generate 1.2 billion take out how

much you're probably spending on AI

expense I think $800 uh million is quite

frankly low uh I think they'll probably

spend more than that but let's just say

it's 70% uh divide that by shares

outstanding you're only adding about 12

cents to EPS which adds about 4% to EPS

it basically does nothing uh for EPS at

some point in the longer term honestly I

think you'll probably want to just

include FSD uh in the actual price of

the vehicle but then margin will go up

so whether you calculate it here on take

rate or you calculate it as part of

vehicle margin you have to put the value

of FSD somewhere so for 2024 on my base

case I don't really see a big move in

this $99 feature it's it's not as

valuable as as I think we hope now if we

go out to 20126 I think we'll be able to

move that take rate to nearly a quarter

of vehicles that'll be about 22 1.2% of

vehicles and I think our AI expense will

still be around a billion 1.3 billion is

what I wrote down about 40% and uh that

add

63 cents to our EPS which if we add a

40x multiple on that because hopefully

we're growing at you know we're growing

EPS at 25% a year a 1.6 Peg right we

multiply that out we get a 40 multiple

that adds about $25 of Share value to

the stock in 2026 which I think is fair

to look at now in fact I think frankly

if we didn't have version 12 FSD Tesla

stock would probably be around

$140 maybe $145

so if you look at the last 30 days the

last 30 days have been pretty wild for

Tesla think about it for a moment you

went from version 11 in the last 30 days

to version 12 121 122 123 12 31 32 33

34 you got half uh off on monthly FSD

Revenue you got an August 8th Robo taxi

announcement and the potential

cancellation of the $225,000 car which I

know a lot of ultra Tesa you're like why

is everyone so concentrated on a $25,000

car well I'll explain that in just a

moment but the last 30 days have been

crazy and so one of the reasons Tesla

stock is the fact that it's flat over

the last 30 days with the horrible q1

deliveries number is actually a sign

that the market is adding back in some

of the value of FSD or Robo taxi or or

whatever otherwise the stock would be

down substantially more of the delivery

numbers that we had so what does this

mean well this means we're we're

probably already factoring in the base

case into Tesla stocks value now and I

think what we do is we can look at Bull

cases and Ultra bull cases in my bull

Case by 2026 I think we have 35%

adoption and only 20% in expenses uh and

that's going to drive about $52 in Share

value I do not think that's being priced

in right now uh and then in the ultra

bull case which I personally think is

unrealistic in the ultra B case we're

only adding about $80 a share in value

by 2026 which we would start pricing in

relatively soon now that is pretty good

especially since I think the vehicles

for the company you're probably adding

about $140 to the share price maybe $120

to the share price so 80 bucks should

put us in the low 200s in the ultra bll

case uh the problem with all of this is

we can't

really understand what the take rates

are for Tesla because they don't give

that data to us so without having that

Clarity we're kind of at a loss if we

don't have the clarity on what take

rates are now then how can we know if

the things we're changing like pricing

or otherwise are actually being

respected by the markets we don't and we

don't really see a lot of FSD Revenue

recognition in each quarter so it makes

it very difficult as an investor to have

Clarity this lack of clarity has always

made me excited for the icing on the

cake at Tesla but I've always

underwritten Tesla as a car company

that's growing rapidly and so for me

when we're able to grow EPS by 30% a

year well I can give Tesla a 50p ratio

and when we can get to 5 million

vehicles at some point in the future you

know this is a $400 $500 stock for me

just underwriting on cars but that

thesis has falling apart now because

Elon is shifting from something that is

very clear to underwrite vehicle

deliveries and a cheaper car to we're

going all in on Robo taxi but Robo taxi

comes with this massive level of

investment

that first of all we're not near Robo

taxi now we're probably 10 years away

from robotaxi and not only are we

probably 10 years away from Robo taxi

when Robo taxi first starts you're still

competing with Uber and I know people

think to us how the hell are we

competing with Uber we're going to

bankrupt Uber they have Labor right but

you're trying to get Uber's Revenue so

if I'm going to spend 20 bucks for an

Uber for you know a 10 or 15minute drive

or 20-minute

drive one of the reasons I'm willing to

pay for that is because I value my time

and I'm not going to sit at a corner you

know waiting for an Uber for 30

minutes or a robo taxi for 30 minutes if

I could get an Uber in 2 minutes you

need a lot of vehicles a lot of

vehicles to actually get people using

the robo taxi Network you need FSD

perfected Vision based FSD perfected

we're not close to that then you need

people saying wow the robo taxi is

better I don't have to talk to a driver

it's cheaper which that lowers your

margin anyway uh and I don't have to tip

a driver this is great but they still

have to manufacture the car you know

even if you create a $30,000 Robo taxi

80% of that car so $24,000 of that car

is still going to be cost for Tesla

we're not actually convinced they can

manufacture a car for $24,000 yet so

it's not like these cars are free and

you know Uber is doing millions of rides

a day so in order for us to pull that

off

I mean we're going to need hundreds of

thousands of vehicles so we're going to

need one to two quars of of robotaxi

vehicle production just to be able to

start competing with Uber in in certain

markets first where you get your

delivery times down low enough at a

reasonable price where people are

willing to take a Tesla Robo taxi versus

an Uber or a lift a lot of competition

there so yes even though there is a

human there and even though you know

Uber spends somewhere around oh boy they

uh they're they've got about $37 billion

of Revenue they spent about $22 billion

just on the labor so more than half is

just going to the drivers right that's

all future potential margin for Tesla

which is hopium because the price is

just going to go down the price of

providing the service is going to go

down you know if you take $22 billion of

cost away somebody else is able to

operate Uber basically for uh in that

case that'd be about $15 billion so they

can drive the prices down by 50% so the

$20 ride is now a $10 ride and no tip

which is great but again Des has to

provide the actual cars somebody has to

charge the cars the cars have to be

maintained who cleans the cars between

each ride the person who throw up in the

back right uh and that's all assuming

we're at level five autonomy so I I it

it's great but like it's great hope but

as an investor trying to trying to price

it in I can't price Robo taxi in yet I I

can only price in the hope of it but

what it can price in is some reasonable

level of speculation on FSD which that's

why yes I think there is an ultra bull

case where the expenses are virtually

nothing for FSD and the take rates are

like you know upwards of 45 to 65% but I

think it's heavily unrealistic I think

right now we're maybe adding 25 bucks a

share now the difference is if we can

get EPS growth from this and we start

proving wow we're really growing from

this then guess what happens if instead

of valuing Tesla at a 40 multiple you

start pricing it in at an 80 multiple on

future growth because now you're getting

EPS growth from software which

eventually will be higher margin they

could prove and this is the real hopium

method they could prove to be the most

profitable software AI play that exists

I actually think that there's very

little profitable software right now uh

that's AI may maybe maybe maybe paler

let's get windy uh P tiers an option uh

I think most of the companies like uh

you know the sales forces and the msfs

you know they're they're trying and I

think they can they can probably boost

their sales a little bit on Office 365

oh get it with AI it's $5 more a month

whatever problem is that's going to

become so so Baseline that I'm not

actually convinced you're going to see

such Revenue explosion uh at at at some

of the core software companies now that

have run as much as they have on AI so I

think those are a little bubbly I

actually like Nvidia more than I like

some of those plays but realistically

Tesla's one of the most profitable AI

plays or will be one of the most

profitable AI plays the question is and

that's from a revenue side the question

is how much are they going to spend to

get there on uh on cap

backs and when are we actually going to

get transparency on how much EPS growth

we can actually expect from these things

if Tesla can lay out a road map for

here's how much we think we can do in

earnings per share

and here's how much revenue we think

we're going to make based on a certain

take rate that's when this stock can

actually finally unleash and take off

because you're going to get away from

just focusing on vehicle deliveries

which we know are in the toilet right

now and you can actually start

underwriting okay rather than just a hm

point of view we can actually underwrite

how much can we expect to show up in

earnings for the stock okay we're going

to I mean elon's done this before he

just tells us look we're going to Target

uh 60% take rate on new vehicles and a

fleet take rate of 30% which we want to

grow to 40% over time uh and then what

we're going to do is uh you know once we

get this going we're going to f we're

still going to focus on a compact car

you know it's not going to be called a

model 2 or whatever we're going to call

it who knows whatever but you're going

to be able to get it as roboox incapable

this we just get a

vision then we can actually price the

Stu so that's my take hopefully that's

useful and if it is let me know in the

comments down below I made some money on

uh trading some uh Tesla calls yesterday

it was really good Tesla was just like

straight up and down uh yesterday it was

down 2% on the day but I bought it was

down 2% so I was arving the difference

between that 2% and like

1755 and uh we were killing it on some

calls yesterday so if you want all those

Buy sell alerts and trade alerts make

sure to check the link down below stocks

and psychology