Thread by @steve_larrison

AST SpaceMobile Q4/21 Earnings Report - a thread

#SpaceMob
$ASTS

1/n

First, the subject that was on everyone's mind going into the call - would AST be ready for the launch window of Summer '22, or would there be a delay?

I am more than happy to say that everything is on track for BW3 to be ready in time to meet the launch window.

If you will
2/n

recall when the March/April '22 launch was delayed, the company had to notify SpaceX 90 days before the launch if they weren't going to be able to make the launch.

For purposes of preparing for the launch, the customer has to assume that SpaceX is going to launch on the
3/n

first day of the launch.

You can see a detailed sequence of events that have to be met at various points in time relative to the first day of an agreed upon launch window.

You can find the list at the bottom of this SEC filing for the Launch Services Agreement that was signed

with SpaceX on March 3rd.

sec.gov/Archives/edgar…

The final Mission Integration Analysis Inputs had to be provided to SpaceX by March 23rd. We also know from documents filed with the SEC related to the multi launch agreement with SpaceX that part of the $23 million
4/n

spent was related to technical changes that caused a need for adjustments on SpaceX's side.

So given the fact that Abel told us that we are on track, and we recently passed a milestone that is critical for a go/nogo decision, anything that causes that launch to be delayed
5/n

would be a surprise.

So now that we know the BW3 launch is highly unlikely to be delayed again, the next question is "Where is BW3?"

SpaceMob is a curious mob, and being a curious mob, I've seen speculation about launch dates ranging from April to early June.

I have always
6/n

been of the belief that "Summer '22" doesn't start before June 21st.

There are two reasons for that:
1) Summer starts June 21st
2) Engineers can be overly literal

I am an engineer, and there are a lot of engineers at AST and SpaceX. I'm relatively confident that the actual
7/n

start date of Summer is not a closely held secret.

But never-the-less, there was still debate about a Summer '22 launch window starting at the beginning of summer and a June 1st date that was traced to a Wikipedia article.

Hmmm, here is an angle that I didn't think of.
8/n

We can make summer appear to start earlier simply by editing the wiki on Summer Solstice.

But on a more serious note, the reason why there was still debate after the earnings call is because Abel's accent makes him difficult to understand at times - at least it does for me.
9/n

That's why I found it necessary to wait until I had a transcript to write this thread - I had too many questions about details in what was discussed to provide the level of analysis that would make me feel good about this thread.

Here is what Abel said about where BW3 is

10/n

First, you can tell that Abel was talking freely and not off of a script.

Next, this is definitely an automatically generated transcript as has mistakes.

For example RS should be RF.

The first marked section covers RF testing and packing the satellite.

11/n

Relevant slides are the left-most and right-most slides below:

RF testing is clear enough, but "LMDS 2" probably isn't clear to most.

The acronym stands for Local Multipoint Distribution Service. The link below is to a book on the subject when applied to a different
12/n

satellite system.

amazon.com/Performance-An…

In laymans terms, it's a simulator that tests quality of service with different numbers of users and modulation types.

I *THINK* this might be synonymous with the aforementioned RF testing.
13/n

The second marked area talks about an environmental test - this is not the same as an RF test.

The 3rd marked area mentions that the environmental test is expected to take 4 weeks.

Then it adds a final inspection prior to shipping to Florida.

14/n

Put it all together, and my read is that they have to

1) Pack the satellite in the cube for LDMS 2 simulation
2) Perform the LDMS 2 simulation
3) Perform the final environmental test
4) Perform a final inspection
5) Ship it!

Of course it is the "Happy Path" - it is always
15/n

possible for a problem to be found along the way.

In addition to the state of readiness of BW3, we have a clear view of what the total cost to build and launch BW3 is.

That number is approximately $80 million.

A couple of points of clarification are needed here.
16/n

The capex section of the chart below shows spending on BW3 up until the end of Q4. We won't see final numbers in the chart until the next earnings report (hopefully in a few weeks)

17/n

Pay attention to the Orange part of the bar - this is where cost related to R&D, Engineering Services, Manufacturing labor for building and testing the satellite, and the cost of materials are reflected.

The gray bar reflects property and equipment. It can be assumed that
18/n

a lot of those expenses are for things that are reusable for the BlueBird builds.

The other thing to keep in mind when you read about an $80 million price tag to build BW3 is that as a test satellite, it has a lot of features that are not needed in the production satellite.
19/n

There are a variety of sensors on BW3 that exist to collect data that will be used to inform design decisions on future BlueBird satellites. Those same features will not become part of the BlueBird satellites.

Also, because BW3 is using FPGAs instead of ASICs, there are
20/n

differences between the solar panels (BW3 is more expensive because of increased power usage by FPGAs) and other aspects of the electrical and cooling system that add to the cost of BW3.

These power related differences caused by FPGAs being used instead of ASICs is the
21/n

reason behind most of the cost differences of the early BlueBirds ($13-15 million per satellite expected) and the later satellites ($11 million per satellite expected).

And, of course, being the first of it's kind, a lot of lessons were learned in the build of BW3 that
22/n

impacted the final build of BW3. These are the kind of lessons that only have to be paid for once. What was learned will not have to be paid for again when building BlueBirds.

Outside of BW3 readiness for summer, the next biggest question that SpaceMob had was probably
23/n

about what is going on with the FCC.

For those of you who may not be following the FCC closely, what's wrong with you?????

Sorry, I couldn't help myself.

So if you aren't following the FCC closely, there are 3 licenses that are pending approval.
24/n

1) Approval of an experimental license to test BW3
2) Approval of use of a terrestrial frequency - V band - for satellite communications
3) Approval to operate SpaceMobile in the United States

While all of these issues are important, the experimental license for BW3 is the
25/n

most time sensitive.

Though issues 2 and 3 are more important in the long run, neither license is needed before phase 2. So there is still plenty of time to get approval.

But there have been questions about what happens if the experimental license for BW3 is not granted
26/n

before the summer launch.

Some people have speculated that SpaceX won't laucnh BW3 from within the United States without FCC approval.

Others have speculated that AST can launch, but they just can't test from within the United States.

AST stated in the earnings call that
27/n

they expect to have the experimental license before summer, but if they don't have it in time, they are still going to launch, they just won't test from within the United States.

On the business side, they formally announced the MOU with Orange.

Orange is the 11th largest
28/n

Mobile Network Operator (MNO) in the world and has some 259 million subscribers.

But the MOU only covers 1 African country. It is expected that if Orange likes what it sees, they will make SpaceMobile available in other countries

But it is important to recognize that just
29/n

because AST has an MOU with MNOs with N number of subscribers does not mean that AST will be selling data to all of those customers through the MNO.

Not all potential customers will become actual SpaceMobile users

Not all customers of an MNO are even eligible to use the
30/n

service because they live in countries not covered by the MNO.

So while it was great to hear about 1.8 billion customers of the MNOs that AST has MOUs with, don't expect SpaceMobile to have anywhere near that number of customers when SpaceMobile is first turned on.

31/n

in addition to the Orange MOU that was announced, my friend CatSE believe that he has found an unannounced MOU.

https://twitter.com/CatSE___ApeX___/status/1509655526531125252

32/n

It was also announced that the African country covered in the MOU with Orange will be participating in the testing process of BW3.

Another issue that was discussed, but didn't contain enough detail to satisfy me personally was the supply chain.

It was mentioned that the
33/n

approach to handle supply chain problems was to order long-lead items well in advance of when they are needed.

This was also mentioned in their Q2 and Q3 reports.

In addition to ordering long-lead items in advance, it was also mentioned that they have been working on
34/n

expanding the supply chain to meet ramped up production for the BlueBirds.

Frankly, I would have liked to get more supply chain information.

In my opinion, the issue that is most likely to cause delay in future BlueBird builds is going to be access to components and
35/n

sub-assemblies needed to build BlueBirds.

It's important enough of an issue that I asked Investor Relations a few questions on the subject before the earnings call.

Sadly, none of those questions were chosen for the Q&A session.

On the financial side, they actually have a
36/n

better cash position at the end of the quarter than I expected.

Here are their cash positions stated from the 3 earnings report they have had as a public company.

Q2/21 - 402.6 million
Q3/21 - 360.4 million
Q4/21 - 321.8 million

37/n

So the cash burn between the first and second quarter was 42.2 million, and the cash burn between the third quarter and forth quarter fell to 38.6 million.

Given the acquisition of the 100,000 square foot and build out of the facility in Midland, and equipment needed to
38/n

build out the 16,000 square foot leased building used as a SNOC (Satellite Network Operations Center - AST refers to them as SOC and NOC. But SNOC is a term we used on that NASA program that I worked on, so I prefer that term) in Lanham, Maryland, I expected the cash burn
39/n

to be higher.

In addition to the building expenses (purchase of the Midland facility, and equipping both buildings), there was an increase in Opex related to growth of the company.

40/n

As you can see, there is steady growth since Q2 for General and Administrative expenses, but that growth is expected to level off for the remainder of 2022.

G&A expenses are costs related to people working in manufacturing who aren't engineers, and office people in areas
41/n

office areas like IT, HR, Accounting, Purchasing and Production Control.

Engineering Services are the engineers working on keeping the manufacturing floor running smoothly.

R&D bounces around a lot. That shouldn't be a surprise. At AST, the R&D function is made up of
42/n

consultants who design various systems and subsystems of the satellites that AST builds.

The reason for the large jump in R&D between Q3 and Q4 is because R&D work for BlueBird satellites is coming online while there is still R&D that was involved in completing BW3.
43/n

You can expect expenses in both Engineering Services and R&D to grow in coming quarters.

The total number of employees in Assembly, Integration and Test (AIT) was 241 at the end of Q3 and 287 at the end of Q4.

It was also stated that head count at the end of 2022 is
44/n

expected to grow to 350 by the end of 2022.

It's worth noting that the expected head count of AIT was originally expected to be 300 people.

I am not sure if this 16.67% growth is related to :
1) A change in expectations of the number of people needed to build the planned
45/n

planned satellites
2) An increase in the number of satellites planned to be built in a given amount of time
3) Some combination of 1 & 2

But understanding the trend is helpful in projecting future cash burn rates

And that is both an important and a timely thing to mention
46/n

It was mentioned in the call that AST will likely be raising capital before phase 1 is completed.

Now that you are back at your seat after running around and screaming "dilution!", keep in mind that there are multiple ways to raise capital.

1) Warrants can be called - that
47/n

alone will bring in $202.4 million
2) Money can be borrowed - even though they are a pre-revenue company (if you don't count what NanoAvionics contributes), they did borrow $5 million for the purchase of the 2nd facility in Midland.
3) They could do a stock offering

48/n

Given that Abel owns 88% of the voting rights in the company and 43% of the company, whatever he wants is the way the company will go.

And given that he stands to be hurt more than anyone else by dilution, I highly doubt that dilution will be his choice when there are
49/n

alternatives.

On the subject of the new facility in Midland, a few details came out during the earnings call.

1) $20 million was earmarked but facilitating automated build and test processes
2) The clean room in the facility is the first activity and completion is expected
50/n

in Q2
3) The new building will be focused on micron and solar array assembly. It is not clear whether that is the only activity that will be done there. But if you have to move things between buildings, it is easier to move a stack of microns than it is to move an assembled
51/n

BlueBird.
52/52

@threadreaderapp unroll

This will be the last time I use ThreadReaderApp. It appears to have deleted 2 messages and broke this thread.

You can resume reading this thread on the body of tweet 26/n

There are a total of 52 tweets.