What this company (and every other in that field) is doing: moving small, microservice-able tasks to AWS. Scale and costs are way cheaper. But it's a very long-term move: a couple of years pass and bam!, here's another Oracle multi-million extension contract they must pay to continue to have the lights on. But to the executive's perspective, this is just the cost of doing business in that field, it's the same for the competition. You could argue this is the same "no one ever got fired for hiring Microsoft/IBM" mentality.
Oracle stack is obsolete, but as a business they aren't, simply because it will still take a very long time until most companies reliant of their stack be able to move out without great risk to core business.
But then Oracle moved their Forms client to Java, creating all kinds of versioning headaches and security risks. They should have stuck with the dedicated stand-alone EXE Forms client: a kind of GUI browser. It was a very practical tool for in-house and specialty apps, but Oracle ruined it with Java.
"Extort everyone for obscene license fees because we currently can, and force them to buy cloud credits they'll never use because Amazon."
There's nothing more disheartening then digging into an issue on a new site and realizing someone checked (or didn't uncheck) a box during install years ago and the system is not complying with its licensing.
The licence fees for small features you might not even know your using can be pretty brutal.
Microsoft got into similar hot water tricking people into installing Windows 10 via confusing dialog boxes.
>“Larry Ellison has done a fantastic job with Oracle,’’ Buffett said Monday in a CNBC interview. “I’ve followed from the standpoint of reading about it, but I felt like I didn’t understand the business. Particularly after my experience with IBM, I don’t think I understand exactly where the cloud is going.’’
He doesn't want to fight. He doesn't care about being right in the eyes of others.
I stole this line 10 years ago to decline the invitation "oh, you are a programmer, can you take a look at my computer ?". I don't understand Windows so I can't, because I use Linux. Declaring yourself incompetent is a fantastic social cheat code.
One of the things Buffet (and Munger) do very consciously is their "circle of competence". They know what they're good at. They do that. They know they're not good at other things, so they don't do them. So I think Buffet is saying that he genuinely doesn't know how to evaluate Oracle, and therefore he's out.
I recommend this approach much more.
why not just ask them to pay you for your time, or else say sorry you need to hire a professional instead of giving them more grief
Have to agree with you, seems fairly pointless to kick someone while they’re down. Computer problems are stressful, so it’s not surprising someone would ask the first person they know is remotely involved with computing for help. It’s very easy just to tell someone a few things that they might try googling.
That seems to frequently be perceived as an insult though.
For some reason, everyone thinks that "computer people" can be counted on to solve their Windows PC problems for free, though they wouldn't ask their relative who's a car salesman to fix their car.
I don't have an actual solution here, I'm just pointing out that this seems to be a common problem.
Because if you "charge" them and get paid for the work then everything else that seems wrong from that point forward they blame you for and expect you to fix it [often for free] because in their mind a) you caused it and b) they paid you to fix their computer.
>or else say sorry you need to hire a professional instead of giving them more grief
This is the best response but I think he said they get pushy and insist he do something. The person is going to be pissy at him for telling them off nicely or not.
What this actually translate to is very simple:
"I can't get an edge that I am used to getting information wise from this company so I am pulling my investment".
Now of course that could be because there isn't good info that they are willing or able to give him, or simply they will not play that game with him for some reason.
Honestly he just doesn't like being in the same boat as ordinary people who invest. Or for that matter ordinary analysts or other high level investors. In order to be successful he has to go to the head of the line in some way. You have to have the tea leaves to read the tea leaves.
If the only info you can get is fully public info, the market should have accounted for it already, and in that case, you might as well take your money and but it somewhere more useful. And if you're fine with just riding the market, why risk an individual stock when you can use an index fund?
I will give a very small example to illustrate my point. Years ago I was looking at buying a small business. I was there for about 2 hours. I noticed that the phone did not ring one time in the 2 hours. Let's assume for the sake of the story that the phone ringing was not super indicative of anything (with this business I mean it wasn't a Pizza takeout place) but somewhat meaningful. They didn't have to reveal anything secret at all for me to take that info into consideration.
The problem with an index fund is that since it isn't a gamble you will never win big!
He is also a voracious reader. Most likely after he purchased his shares he continued reading and thinking about the state of things and where they are headed. He read something that made him question his investment thesis and re-think things that he was previously confident in.
Most people, if they continue reading or thinking at all, will just look for things that confirm their previous decision. It's a good lesson to look for things that disprove what you think you know.
 I think he likes the cloud because as opposed to most tech, there are tangible assets involved and significant capital outlays. Combined with the nature of platform lock-in, this keeps out newcomers.
If that is the case then I take my hat off to him. I was also wondering why he would be pulling out of something he'd gone in the previous quarter.
For example, he always cites figures such as return on _tangible_ assets in his interviews.
He is also obsessed with moats. A large reason for the relative lack of moats in tech is that the barrier to entry is low, and has been dropping for decades. You could whip up a respectable competitor (technology-wise) to most of today's unicorns in well under a year or so with only a little bit of money, but given many years and billions of dollars you couldn't re-create Coca-Cola's distribution network.
I am actually curious how you think this is not like his other investments? His investment style has changed for sure, but I would draw two lines, one that he talks about around See's Candy in the 70's. The other I would put about 10-15 years ago, where he started looking at more capital-intensive businesses (like BNSF). Both were gradual shifts, though.
As a Buffett-head, I can't immediately think of another time when he has turned around on a stock so fast. (not counting partnership days, where he would flip stocks pretty relatively quickly, but only after their full value was realized)
His firm got burned by IBM, so he's extra jittery dealing with "big tech" now. IBM created this program to better track and maximize revenue by cranking up revenue incentives for middle managers. It looked good on paper: fine-tune the reporting of and increase incentives. But their customers began to feel squeezed and milked. You can often gain short-term revenue by skimping on customer service, and eventually customers revolted, damaging IBM's cred.
Sales 2014: $38.2 billion; sales 2018: $39.8 billion
No growth beyond inflation. Net income hasn't climbed (roughly stuck at $9-$10 billion for the last five years). Annual debt interest costs have more than doubled to $2 billion as they've made a mess of their balance sheet.
In two years AWS will be as large as their entire business.
In the five years Oracle has seen zero growth, Salesforce has gone from $4b to $10.5b in sales. It's likely that over the next five years Oracle will again see near zero growth and Salesforce will double in size.
Here's what a successful cloud transition with growth actually looks like:
Adobe's business was stagnant for years. Then.... Sales 2014: $4.1b; 2015: $4.8b; 2016: $5.8b; 2017: $7.3b; 2018: $9b.
Their profit skyrocketed from $268m in 2014, to $2.6b for 2018. They earned as much last quarter as they did for all of 2015. The stock has gone from $60 to $250 / share.
You can tell exactly how well Oracle's cloud transition is going by the fact that everybody around them is eating their lunch and seeing growth, from AWS, to Microsoft, to Salesforce, to dozens of other enterprise software companies. Microsoft had watched its top line stagnate for several years as they struggled to find new growth. And then suddenly a big jump: $96b in sales, to $110b in sales, over the prior year. Another example of cloud transition growth in action.
I worked with great people at Oracle but I disliked working there immensely. Compared to a lot of other big-name tech companies, their approach to managing employees and their approach to innovation is extremely outdated and somewhat draconian (in certain arenas). Furthermore, I got the sense that any customers we had were largely the result of good salesmanship and unforgiving lock-in. I'm sure it's probably not like this on every team at Oracle, but that was my experience. It was also incredibly clear from the scarce (once a year? if that?) corporate wide pow-wows and communications from above that the company's top executives not only don't understand their employees but also quite clearly don't care about them, their advancement, or happiness. Hell, Oracle was just starting to bring macs into the workplace in the name of appeasing employees and improving employee-executive relations when I left in early 2018.
I would short Oracle if the overvaluation wasn't so old and stable. Hard to guess if or when it will end.
Umm, ORCL's PE is currently 18.7 and the average PE of the S&P 500 is currently around 21 so your statement about a PE that is pricing in "great growth" is dubious at best:). The average S&P 500 PE ratio is also above 18 if you start at a reasonable date of say 1980..
I think people who are calling Oracle just a database company are really underselling how much value they sell above the database abstraction, they are much closer to Microsoft and AWS (and SAP, Adobe, Autodesk) than people think. They own and sell a ton of cloud "services" that arent "database" but instead specific tasks.
Their "divisions" are Servers, Storage, and Networking; Middleware; Industry Solutions including Communications and Media, Construction and Engineering, Health Sciences, Hospitality and Retail, Utilities; and Applications. They tend to focus on more specific purpose software, compared to Microsoft and Adobe more general purpose software for all industries.
And adding to your point, F/OSS DBs have gotten really good, and they're commoditized.
Oracle is good at extracting the maximum profit from their sales. They're aggressive about sales and policing usage and they don't often cut customers slack. They've learned that clients will try to give them the shaft and they do their best to not get shafted. They're good capitalists as well as the best database developers in the world.
Sounds like a horrible company to work with. Most of the people I know that use Oracle services/products due so because they are stuck with them
My point is that, despite a proliferation of free and open software, companies still often _pay_ for good software that works correctly. And Oracle extracts their pound of flesh w/o taking a single drop of blood. Oracle's is a very strong business model that is no accident and results largely from having no peers.
Today I learned that "capitalist" means "treating the customer as the enemy" and "legitimized exploitation".
Honestly curious - have you absorbed these attitudes from an MBA program, or did you develop them more organically?
*“People of the same trade seldom meet together, even for merriment and diversion, but the conversation ends in a conspiracy against the public, or in some contrivance to raise prices.”
- Adam Smith, "The Wealth of Nations"
This is almost a definitive description of capitalism both by its origin and by its very nature.
Google is a search engine + search engine monetisation engine. They have some third parties participating in the ad market (notably other search engines) and some non-search properties (eg YouTube) that contribute some as revenue but search+AdWords is still the core.
Meanwhile, most of the notable things google do are in unrelated or marginally related to the business model. Android, chrome, self driving cars...
It's as if most of the company has nothing to do with the business. They used to explain this with a VC metaphor, but that seems to be gone.
FB & twitter are even stranger. Why do they have 100X more employees and costs than they had a few years ago, while not apparently outputting much more than they were outputting before. They're still "just" social media sites. With all due respect to ad-tech and content ranking algorithms...
The common denominator is a revenue stream that comes from dominating a niche and very little relationship between what the company invests in and how it makes money. They didn't/couldn't have actually directed themselves towards those things as goals.
Oracle is similar. They're deeply embedded in their customer's businesses. That means a captive market. This generates a revenue stream. There's no predictable way of turning that revenue into a cloud service, no matter how much they're willing to spend.
Ultimately, the problem is that the market as a whole is overfunded. Investors don't want dividends, so they get highly speculative/dubious investments instead.
Chrome, Andriod, Cars, DNS. They all help collect data about a person to better train their core business model which is advertising.
> In June 2011, IBM won the contract to set up the system, using PeopleSoft software; the original contract was for $5.7 million, but IBM was eventually paid $185 million.
...that's 32x the original price.
Everyone in software knows exactly how IBM and Oracle make their money. That part's not a mystery. The bigger mystery to me is why these governments and big companies keep spending money on these professional leeches and expecting a different result.
Do other gov contractors charge 32x the original budget? Do construction, or ad companies, or accountants? Or is it just software?
1. Were there scope changes throughout the process? This is quite common and costs firms and govs a lot of money yearly.
2. Did the Gov know everything they need at beginning?
3. Are there other services or businesses with the amount of experience this one has?
4. Are there continuous services that will cost you monthly/yearly? Did you decide to increase the number of users in that platform?
5. Were there unexpected support or training needed?
6. Would it be political suicide to put the whole price at the beginning?
All these things add up and you can only hope that your gov and big companies take into account before entering this endeavor
And what makes you think that "these governments ... expect a different result"? Isn't is possible that the governments expect (and welcome) exactly this result?
"don't make the mistake of anthropomorphising Larry Ellison" - Bryan Cantrell
skip to 38:20
A bit heavier ranting (with the fallout after his usenix ranting):
Selected quote: "Any explanation of Oracle that doesn't end with a nazi allegory does not fully explain Oracle"
Bryan Cantrill is really fun to listen to, is a tech veteran now, and is still as passionate and energetic as he was the first day (Another interesting video about programming languages he experienced during his (long by tech standards) career: https://www.youtube.com/watch?v=LjFM8vw3pbU)
>[Buffett] Look for simple businesses. If I gave you $10M to invest right now and you only had three weeks to spend it and you could only spend it in Omaha, you’d look for simple, understandable, strong businesses. You look at the Nebraska Furniture Mart (NFM). You wouldn’t look at the third best fast food chain. You might look at McDonald’s, because it is number one and will probably always been number one. They have share of mind. What about Oracle? Too hard. GM? Too hard. You can’t predict the future for these two companies. Too many variables.
That was 2009. Looks like he's decided that was still the case.
Also going by the top comment in https://news.ycombinator.com/item?id=19289583 three days ago:
... "I probably met nearly 100 developers, and when I asked all of them what they use Oracle for EVERY SINGLE ONE OF THEM openly told me that their companies use Oracle right now but they hate it and they were all in the process of transitioning to other providers or open source solutions."...
I'd guess shorting Oracle might be more the thing to do. Typically customer reaction proceeds the financial numbers moving.
Mongo trades at a premium, but its market cap is in the single-digit billions & is nothing Oracle couldn't afford right now. Plus, Oracle's massive legal machine could move the licensing innovation needle even further in the direction of Mongo's recent move to SSPL.
Disclaimer: I work for Salesforce
Proprietary licensing that has a veneer of openness isn't a legal innovation issue, it's a PR innovation issue (how to get the goodwill and community support that comes with open source while using proprietary licensing that doesn't provide the freedom of open source.) Oracle doesn't have any particular special PR competence when it comes to targeting the audience for whom open source is a big draw, and seems quite happy locking anything that they don't want to share with the community behind traditional proprietary licensing, so I wouldn't see the SSPL game and Oracle having a lot to offer each other. SSPL is for people who want to be like Oracle but pretend they aren't, and even if Oracle cared about the pretense, there's not a lot of indication that the SSPL approach works.
Take something like RAC.
Call it "Cloud On-premises."
Viola! Cloud sales are up.
Larry recently bragged how even Amazon is under his cloak using Oracle because there is nothing out there better. Bezos response: We'll build our own database no matter what the cost and then we'll compete against you signing up what were your customers. How do you like that Larry ?
And they are free to use any of the open source ones.
Your narrative makes little sense.
> So this is another claim I'm going to make, Aurora, not built by Amazon, built by us, called MySQL. Amazon just gave it a new name. It's also true of Redshift. Amazon didn't build Redshift, that's not what they do. These are -- they just open-sourced pieces where they built -- that they used to build their cloud. They did a great job. They did a great job of making those pieces available in a coherent -- deliver them in a coherent cloud. I give them a lot of credit. But they don't build databases. Amazon still runs all of their -- all of Amazon on Oracle Database. They are trying, they promised because they don't like me reminding them of that publicly. They said they're trying to get off Oracle by 2020.
> Aurora is our other database. It's our low-end database. It's our database. It's not theirs. Well, it's open source. Anyone can use it for nothing.
(This is bullshit, of course.)
AWS Aurora wasn't built out of spite, however. It was/is designed to be a cloud-native database from the start because there was a perceived desire for the product. As it's AWS's fastest growing service, the team deduced correctly.
Disclaimer: I work for AWS.