this is the trend Saul’s board members are investing in.
Software doesn’t work like this. Microsoft might spend a lot of money to develop the first unit of a new program, but every unit after that is virtually free to produce.
As somebody who manages software development I have to say this is misleading. While the production itself is free, your costs rise with the rate of adoption and with the age of your software. The more users adopt your software the more extreme cases you have to cover. It has to support more types of hardware. The fringe features get more use, unlikely bugs surface every day when your software is used by millions of people. You need a much larger support and development team the more your software gets used and as it ages you will have dozens of different versions in the field each with its own challenges.
In a lot of cases you can not get rid of these cumulative additional costs. Since they quoted Microsoft, they should have looked at their costs for Windows XP and Windows 7. Even though both have been end of life products for years, they still incur costs for MS. People still use them and as new security vulnerabilities get discovered MS has to fix them or face a massive PR disaster.
Selling a physical commodity on the other hand is fire and forget for the most part, easily predictable.
Since they quoted Microsoft, they should have looked at their costs for Windows XP and Windows 7.
Would Bill Gates not be familiar with these costs? Sure Gates’ explanation might be a bit over-simplified, but I think it is basically right too.
Matt
Long MSFT
for many types of non-cloud/subscription sw, you do pay yearly maintenance fees
This reminds me of how many see and misunderstand the stocks many of us are invested in. From Bill Gates.
These businesses can’t be properly “valued” because we can’t fathom what their revenue and products will be in the future.
“It took time for the investment world to embrace companies built on intangible assets. In the early days of Microsoft, I felt like I was explaining something completely foreign to people. Our business plan involved a different way of looking at assets than investors were used to. They couldn’t imagine what returns we would generate over the long term.”
The declining need for capital in the post-industrial service economy has another big effect: declining returns on capital.
Notably, low interest rates. Or no interest at all.
Aside from the implications for investors, together with the decline in the working-age population, this creates a situation where the neutral interest rate (the interest rate at which full employment is achieved with modest inflation) is negative.
Since the interest rate can’t really be negative, the result is “secular stagnation”.
He is not wrong, because the production costs for new copies are almost zero. However, it’s also misleading because it’s the maintenance costs that grow with the amount of copies distributed.
As to why he would say something misleading like that I can’t really judge, but He’s been out of MSFT for almost twenty years.
Look a it this way, if he’s right, then why is MSFT’s gross margin on a long term downtrend?
https://www.macrotrends.net/stocks/charts/MSFT/microsoft/gro…
Remember, R&D (developing new features/versions) does not influence the gross margin, this downtrend only happens if the cost of goods produced is rising.
I meant cost of goods sold, the proper term for this accounting item - the copies of Windows already sold do incur costs on MSFT.
“Capitalism without Capital” is a terrible way to put it because it’s misleading. Capital is needed just not in the same old proportion of brain to brawn. That’s exacty what Being Digital means, copyright 1995
Being Digital 1st Edition by Nicholas Negroponte (Author)
https://www.amazon.com/Being-Digital-Nicholas-Negroponte/dp/…
As investors what interests us is the difference between the old and the new business models. The old was mostly based on the economic law of decreasing returns, each additional unit of output tended to cost more than the previous one save for economies of scale, which are not possible everywhere. When you have exhaused the labor force the extra help costs more. When you have used up all the prime land, more land is less productive. Typically decreasing returns applies to commodity type of goods and servces. The digital world has accelerated and amplified the economy based on the law of increasing returns which is best described as “the more you sell, the more you sell.” It’s not constrained by decreasing availability. You can download a billion copies of software with hardly any cost increase. You can make several orders of magnitude more transistors than before for less cost, Moore’s Law! You won’t see “Peak Sand” in the news! LOL
Because digital business is not held back by decreasing returns the growth rates are astounding, overwhelming Ben Graham’s Security Analysis. To complicate matters GAAP just hides the reality of increasing returns. R&D is a good example. In the old economy you needed physical assets to be productive and loans were usually made against the security of physical assets, land and other properties. Since R&D only produces knowledge it couldn’t be used as security for loans so it made sense to expense it. In the digital world the most valuable asset is knowledge but by expensing R&D GAAP says it’s not an asset but an expense. According to GAAP Windows is worthless, Office is worthless, MacOS is worthless, Mongo Database is worthless, CUDA is worthless. All the stuff that is making Saulites rich is worthless! So says GAAP!
Of course, were Intel to buy out Nvidia, it would have to put on the books what it paid for CUDA as “Goodwill” laying bare GAAP’s lie. Just because CUDA has new owners it now shows up on the Asset side of the Balance Sheet, how ridiculous is that?
The capital is still there, it’s just that GAAP is clueless.
Denny Schlesinger
Look a it this way, if he’s right, then why is MSFT’s gross margin on a long term downtrend?
Gross margin has two inputs, what it costs and what you can sell it for. If price drops faster than cost, margin will shrink.
Denny Schlesinger
Windows 7 original price was $119.99 for Home Premium full license
https://answers.microsoft.com/en-us/windows/forum/windows_7-…
Windows 8 original price - $119.99 for Home Premium equivalent (called “core”)
https://www.itprotoday.com/windows-server/windows-81-pricing…
Windows 10 original price - $119 for Home edition
https://www.cnet.com/news/microsoft-prices-single-windows-10…
The inflation rate change between Win 7 and Win 10 releases was 10%. Yet the margin dropped by more than 20%. This is despite large scale expense saving measures implemented by Satya Nadella.
I think we are getting a bit off-topic for this board, but trust me, supporting a large user base is expensive, even for client side software like operating systems. (in some sense client side is worse as you have less control of what users can do and rolling out updates is more expensive).
Note that I don’t think these are bad. On the contrary - Nadella was able to keep these rising expenses manageable and they do represent a good moat. Don’t think anybody can support a 1 Billion user base of an universal operating system as efficiently as MSFT. (iOS is not universal, Android is supported by smartphone vendors and Linux does not have that kind of support and nowhere near the user base).
I’m just saying Gates is wrong when he says you don’t need additional capital once youve “finished” your software product.
I think this conversation would be better suited to METAR so I’m not going to respond here any more.
This comment is by TheStreet about Nvidia:
Just look at its price-to-earnings ratio of 44, compared to the tech-heavy Nasdaq’s 23.
Nvidia has 5000 engineers who, according to GAAP, are just an expense. If Nvidia would recognize their R&D as an asset instead of as an expense, would profits rise sufficiently to bring the P/E down to 23?
Denny Schlesinger
https://www.thestreet.com/markets/nvidia-risks-to-watch-for-…
Just look at its price-to-earnings ratio of 44, compared to the tech-heavy Nasdaq’s 23.
…
If Nvidia would recognize their R&D as an asset instead of as an expense, would profits rise
sufficiently to bring the P/E down to 23?
No.
With a five year life: ~37.
With a ten year life: ~35.
Ears
No.
With a five year life: ~37.
Thanks for the numbers. My point is that TheStreet is making an apples vs. aardvarks comparison.
BTW, GAAP’s treatment of R&D is a subsidy for high tech, less income tax! What’s not to like?
Denny Schlesinger
I am completely in sync with this new trend of capitalism without capital: I practice a form of investing known as “investing without capital.”
Rich
who promises he will not do this again for at least 12 months
You won’t see “Peak Sand” in the news! LOL
The World is Running Out of Sand
The little-known exploitation of this seemingly infinite resource could wreak political and environmental havoc
Sand and gravel are now the most-extracted materials in the world, exceeding fossil fuels and biomass (measured by weight). Sand is a key ingredient for concrete, roads, glass and electronics. Massive amounts of sand are mined for land reclamation projects, shale gas extraction and beach renourishment programs. Recent floods in Houston, India, Nepal and Bangladesh will add to growing global demand for sand.
https://www.smithsonianmag.com/science-nature/world-facing-g…
You’re welcome.
While the production itself is free, your costs rise with the rate of adoption and with the age of your software. The more users adopt your software the more extreme cases you have to cover.
Agreed, there is incremental cost. But it is incremental. The cost for producing 1000 cars to 2000 cars almost linear, but not so with software.
Selling a physical commodity on the other hand is fire and forget for the most part, easily predictable.
Not true. Servicing the product is an on-going commitment. Ask either GE or Rolls Royce on their engine part failure. Relative to them to get fixes for software bugs users need to have support. They not only pay one time fee to procure the product but pay yearly fees for support.
Agreed, there is incremental cost. But it is incremental. The cost for producing 1000 cars to 2000 cars almost linear, but not so with software.
This is what Bill Gates tried to tell Steve Jobs, but Jobs wasn’t having it. He needed control of the entire box, from conception to user in the chair. Gates understood that producing the first box of software cost a lot, but every box after that was just $1, and still sold for the same price as the first box. He didn’t care if other people made the macnhine, that was capital intensive and had low leverage for additional production. Just making another box of software for $1 and selling it for $100 was high leverage, and put Microsoft on top of the world for a time.
Not taking sides which approach was right, but we have become a ‘winner take all’ mentality society, and the VC’s love it when you can give them a business model that works this way now.
This is what Bill Gates tried to tell Steve Jobs, but Jobs wasn’t having it.
That’s because Gates was out to make a lot of money and Jobs was out to change the world.
And what’s great is that both men succeeded.
Oh, and Microsoft ended up making hardware after all…