Author Topic: Buying Crowdstrike on the dip  (Read 2351 times)

marcus_aurelius

  • 5 O'Clock Shadow
  • *
  • Posts: 83
Buying Crowdstrike on the dip
« on: July 22, 2024, 10:14:55 PM »
In light of the recent trouble (major IT snafu caused by an faulty update), did anyone buy Crowdstrike? I bought $5100 worth when it hit $295, aiming to hold for at least 5 years. (It's at $264 right now.)

P/E is still high at 496 but I think Crowdstrike will bounce back. Their market penetration and growth rate are awesome.

GilesMM

  • Handlebar Stache
  • *****
  • Posts: 2366
  • Location: PNW
Re: Buying Crowdstrike on the dip
« Reply #1 on: July 22, 2024, 10:21:35 PM »
They could bounce back really strong. Or collapse under a flurry of law suits.

Telecaster

  • Magnum Stache
  • ******
  • Posts: 4030
  • Location: Seattle, WA
Re: Buying Crowdstrike on the dip
« Reply #2 on: July 23, 2024, 01:15:25 AM »
I bought CRWD in December of 2020, so I've been following for a bit.  I don't like to talk about individual stocks too much because the back and forth tends to devolve into salesmanship where people stake out their sides rather than dissect arguments.    Anyway, the bull case for CRWD is the same as for a lot of SaaS companies.   They are in an arms race with their competitors to become the biggest, the fastest.   If they succeed, then they have all kinds of pricing power due to network effects and switching costs.   There is an old saying in investing "you only have to be right once."  This could be one of those stocks.  That's the bull case. 

The Bear Case is what we just saw, at least potentially.  The software fails and the customers are all pissed off.  These customers don't give a shit about switching costs.  All they want is to be able to fly their airplanes on time. 100% guaranteed they will blame this on the software vender and switch venders.  Maybe potential customers are spooked and the flywheel of growth is broken.  It growth is broken, the stock price will also be broken, and most likely the whole thing goes "poof."   

In short, I wouldn't buy this stock on the dip, I would only buy it with the extreme conviction that the current problem (and things like the current problem that may happen in the future) won't materially hurt the prospects of the company, and it can maintain something like its current growth trajectory. 


Heckler

  • Handlebar Stache
  • *****
  • Posts: 1832
Re: Buying Crowdstrike on the dip
« Reply #3 on: July 23, 2024, 04:39:46 AM »
  They are in an arms race with their competitors to become the biggest, the fastest.   


In short, I wouldn't buy this stock on the dip…

As an affected downstream customer, I prefer they all slow down and test properly before blue-screening the world.  I had never heard of crowd strike, and to me it stinks of Russian meddling with the West somehow, will never buy stock or software (think of the long con. “Oops, sorry, it was code defect”).  Look up original founders and make your own conspiracy theory.  I have mine.
« Last Edit: July 23, 2024, 05:30:07 AM by Heckler »

ChpBstrd

  • Walrus Stache
  • *******
  • Posts: 7822
  • Location: A poor and backward Southern state known as minimum wage country
Re: Buying Crowdstrike on the dip
« Reply #4 on: July 23, 2024, 12:14:03 PM »
A growth stock with a PE of 495 that suddenly fails their customers and becomes notorious might have a hard time generating the growth implied by their valuation.

A good trick is to imagine yourself changing jobs to become a commission-only salesperson for a company you're considering investing in. Can you imagine the gloom of being a CRWD salesperson right now, trying to book calls with IT directors who have watched or indirectly experienced the chaos, had flights cancelled, lost touch with vendors and customers who got the BSOD, and are feeling deep gratitude that they didn't go with CRWD earlier?


ChpBstrd

  • Walrus Stache
  • *******
  • Posts: 7822
  • Location: A poor and backward Southern state known as minimum wage country
Re: Buying Crowdstrike on the dip
« Reply #6 on: July 25, 2024, 06:31:08 AM »
funny article from The Onion

https://unusualwhales.com/news/crowdstrike-crwd-the-cybersecurity-firm-that-crashed-millions-of-computers-with-a-botched-update-all-over-the-world-last-week-is-offering-its-partners-a-10-uber-eats-gift-card-as-an-apo#google_vignette



.........oh wait that actually happened
So an investment is kinda like a bet on the effectiveness of this gesture, and the people who would make such a gesture, right?

ChpBstrd

  • Walrus Stache
  • *******
  • Posts: 7822
  • Location: A poor and backward Southern state known as minimum wage country
Re: Buying Crowdstrike on the dip
« Reply #7 on: August 05, 2024, 01:19:21 PM »
Now down to $220, 17% below the date of the OP.

GuitarStv

  • Senior Mustachian
  • ********
  • Posts: 24724
  • Age: 43
  • Location: Toronto, Ontario, Canada
Re: Buying Crowdstrike on the dip
« Reply #8 on: August 06, 2024, 07:46:01 AM »
The crowdstrike crash was due to bad data files that the company pushed as part of an update.  The company itself failed on two fronts:

- The initial code that they run as part of Windows Kernel wasn't tested to handle bad data files (causing a null pointer exception, which wouldn't be a big deal except that crowdstrike runs as part of the kernel which means instant blue screens).
- The data they push for worldwide updates isn't tested at all (the bad data files were empty and cause the null pointer exception in the code on 100% of windows machines they were pushed to).

These are really, really big flaws in company process and speak to some very lazy/negligent stuff going on behind the scenes.  This is also coming from a company run by the same man who was responsible for a worldwide shutdown of Windows PC when he was acting as the CTO of McAfee antivirus.  Maybe I'm weird, but if I was a crowdstrike subscriber, I'd be looking to change to something else right now regardless of the costs.

GilesMM

  • Handlebar Stache
  • *****
  • Posts: 2366
  • Location: PNW
Re: Buying Crowdstrike on the dip
« Reply #9 on: August 06, 2024, 10:42:14 AM »

GuitarStv

  • Senior Mustachian
  • ********
  • Posts: 24724
  • Age: 43
  • Location: Toronto, Ontario, Canada
Re: Buying Crowdstrike on the dip
« Reply #10 on: August 06, 2024, 11:57:38 AM »
This problem was 100% caused by gross negligence on the part of Crowdstrike.  They hoped for the best, pushed an untested file globally, and then their software crapped the bed because it didn't have proper data error handling.  If there's any kind of justice in the world, they will lose.

GilesMM

  • Handlebar Stache
  • *****
  • Posts: 2366
  • Location: PNW
Re: Buying Crowdstrike on the dip
« Reply #11 on: August 06, 2024, 01:51:12 PM »
This problem was 100% caused by gross negligence on the part of Crowdstrike.  They hoped for the best, pushed an untested file globally, and then their software crapped the bed because it didn't have proper data error handling.  If there's any kind of justice in the world, they will lose.


One could argue an operation like a major airline should have the forethought to include some redundancy in their systems to prevent a single point of failure such as this from bringing them to their knees so quickly. No doubt the hackers took note.

GuitarStv

  • Senior Mustachian
  • ********
  • Posts: 24724
  • Age: 43
  • Location: Toronto, Ontario, Canada
Re: Buying Crowdstrike on the dip
« Reply #12 on: August 06, 2024, 02:15:54 PM »
This problem was 100% caused by gross negligence on the part of Crowdstrike.  They hoped for the best, pushed an untested file globally, and then their software crapped the bed because it didn't have proper data error handling.  If there's any kind of justice in the world, they will lose.


One could argue an operation like a major airline should have the forethought to include some redundancy in their systems to prevent a single point of failure such as this from bringing them to their knees so quickly. No doubt the hackers took note.

Not sure I follow you.  What sort of redundancy were you picturing that would have prevented problems in this instance?

GilesMM

  • Handlebar Stache
  • *****
  • Posts: 2366
  • Location: PNW
Re: Buying Crowdstrike on the dip
« Reply #13 on: August 06, 2024, 03:03:57 PM »
This problem was 100% caused by gross negligence on the part of Crowdstrike.  They hoped for the best, pushed an untested file globally, and then their software crapped the bed because it didn't have proper data error handling.  If there's any kind of justice in the world, they will lose.


One could argue an operation like a major airline should have the forethought to include some redundancy in their systems to prevent a single point of failure such as this from bringing them to their knees so quickly. No doubt the hackers took note.

Not sure I follow you.  What sort of redundancy were you picturing that would have prevented problems in this instance?


Alternative IT for the systems that failed (operating systems, cybersecurity software, reservations software, etc).  This will be explored in the lawsuits with Crowdstrike claiming the damages to airlines were largely of their own making and the Crowdstrike glitch should have had minimal impact.

GuitarStv

  • Senior Mustachian
  • ********
  • Posts: 24724
  • Age: 43
  • Location: Toronto, Ontario, Canada
Re: Buying Crowdstrike on the dip
« Reply #14 on: August 06, 2024, 03:17:18 PM »
This problem was 100% caused by gross negligence on the part of Crowdstrike.  They hoped for the best, pushed an untested file globally, and then their software crapped the bed because it didn't have proper data error handling.  If there's any kind of justice in the world, they will lose.


One could argue an operation like a major airline should have the forethought to include some redundancy in their systems to prevent a single point of failure such as this from bringing them to their knees so quickly. No doubt the hackers took note.

Not sure I follow you.  What sort of redundancy were you picturing that would have prevented problems in this instance?


Alternative IT for the systems that failed (operating systems, cybersecurity software, reservations software, etc).  This will be explored in the lawsuits with Crowdstrike claiming the damages to airlines were largely of their own making and the Crowdstrike glitch should have had minimal impact.

What you're describing would seem to require a deployment environment where every airline employee has two separate computers connected to two separate networks?  And all the data would have to be instantly transferred through both sets of computers all the time by (most likely) two completely separate sets of software.  You could do one network fully running Windows, and a separate one fully running Linux.  But of course, then you would be hoping that Crowdstrike doesn't push updates to both at the same time - as they also have a history of causing kernel failures on Linux distributions.  So I guess you would have to add in another couple computers at each terminal - to run separate antivirus software on them to be really independent.  So four separate computers for every airline employee, all networked and communicating together.

If the people in the court aren't phenomenally stupid, that's going to be a pretty tough sell . . . when the alternative is:  Test your data updates before pushing them to computers around the world, and wrap kernel level code in a try...catch statement so it doesn't completely fuck systems if things go wrong.  Both of which are basic and pretty standard software practices.

ChpBstrd

  • Walrus Stache
  • *******
  • Posts: 7822
  • Location: A poor and backward Southern state known as minimum wage country
Re: Buying Crowdstrike on the dip
« Reply #15 on: August 07, 2024, 12:19:57 PM »
This problem was 100% caused by gross negligence on the part of Crowdstrike.  They hoped for the best, pushed an untested file globally, and then their software crapped the bed because it didn't have proper data error handling.  If there's any kind of justice in the world, they will lose.


One could argue an operation like a major airline should have the forethought to include some redundancy in their systems to prevent a single point of failure such as this from bringing them to their knees so quickly. No doubt the hackers took note.

Not sure I follow you.  What sort of redundancy were you picturing that would have prevented problems in this instance?


Alternative IT for the systems that failed (operating systems, cybersecurity software, reservations software, etc).  This will be explored in the lawsuits with Crowdstrike claiming the damages to airlines were largely of their own making and the Crowdstrike glitch should have had minimal impact.

What you're describing would seem to require a deployment environment where every airline employee has two separate computers connected to two separate networks?  And all the data would have to be instantly transferred through both sets of computers all the time by (most likely) two completely separate sets of software.  You could do one network fully running Windows, and a separate one fully running Linux.  But of course, then you would be hoping that Crowdstrike doesn't push updates to both at the same time - as they also have a history of causing kernel failures on Linux distributions.  So I guess you would have to add in another couple computers at each terminal - to run separate antivirus software on them to be really independent.  So four separate computers for every airline employee, all networked and communicating together.

If the people in the court aren't phenomenally stupid, that's going to be a pretty tough sell . . . when the alternative is:  Test your data updates before pushing them to computers around the world, and wrap kernel level code in a try...catch statement so it doesn't completely fuck systems if things go wrong.  Both of which are basic and pretty standard software practices.
Yea, fundamentally the difference between an application update which only bricks an app, and a kernel update, which bricks the whole machine with all its apps and leaves users unable to communicate with their IT departments.

Perhaps all Crowdstrike's customers could do is roll out updates in a slower, more gradual way rather than to all machines at once. Then there could be a way to stop the update if it is bricking machines. This is basically pushing testing responsibilities to Crowdstrike's clients, and comes with the added risk of being vulnerable to exploits for a longer period of time. I doubt we will see this argument in court because it was not the service Crowdstrike was selling, and it is probably impractical due to the security risks.