PPE accounting refers to long term assets categorized as PPE (plant, property, and equipment), which are assets that are expected to create cash flows over the long term.
These asset categorizations are significant because their depreciation schedules directly influence the bottom line.
However, with the emergence of big data as a result of the invention of the internet, larger and larger portions of company PPE are now being dedicated to computer and networking hardware to store this data among some of the biggest public companies in the stock market.
This is a long term secular trend that investors using old order GAAP accounting have not really had to consider.
Where in the past PPE accounting was only really a focus for capital intensive companies and industries, technology companies are increasingly investing in big data infrastructure, and investors will need to include a greater focus on PPE when examining these tech companies.
This post will cover:
- The Basics of PPE and Depreciation
- Advantages / Disadvantages of PPE accounting
- Where to Find PPE in the 10-k
- Is Big Data Really That Relevant?
- Company Examples of Big Data / Cloud PPE
The Basics of PPE and Depreciation
PPE is also known as a fixed, or tangible, asset. It must have two features:
- A useful life of more than 1 year
- A value that depreciates over time
These are real expenses that the company pays, but it’s tricky because the expenses are accounted for over multiple years rather than just 1 (like a regular expense). That’s where the depreciation comes in.
Some examples of Property, Plant and Equipment (PPE):
- Machinery and Equipment
- Networking Equipment
- Computer software
- Computer hardware
- Office furniture
- Office supplies
The list could go on and on, as long as it satisfies those two features at the start.
PPE will vary depending on industry and even company to company, as there’s no standard for GAAP accounting other than that PPE is a long term asset.
Advantages of PPE – Simplified Income Statement Accounting
One of the great advantages of PPE and depreciation is that its use in accounting helps an analyst determine the long term health of a business.
If a business is in large growth mode, it may strategically spend a lot of money on PPE, which would cause reported earnings to crater if it weren’t for depreciation. With depreciation on PPE, the large investments in the business are spread out over several years so that an analyst can still see the health of the business (through EPS) outside of these big investments.
Along the same vein, if a company needs recurring capex to maintain profitability, using depreciation on PPE also smooths reported earnings over a long time period, balancing the years in which cash flows were spent or not spent to better reflect the long term impact of the capex on profitability.
Disadvantages of PPE – Management’s Discretion
Because of the lack of GAAP standards for PPE, management has wiggle room in how it defines PPE and its categorization and depreciation schedule on the balance sheet.
I don’t find this to be a fault of GAAP accounting, just to be clear. These long term capex investments to grow the business aren’t always black and white, they don’t always create equals amounts of cash flows or ROI, they can sometimes be reasonable categorized in different ways (a computer that is assigned to a network engineer is a computer but also supports the network, so how would you categorize that?). That’s just the reality of business.
But, this discretion allows for potentially large variance in depreciation schedules of these assets, which could affect income statement accounting depending on if this schedule is aggressive or conservative.
This could happen in so many different ways that I couldn’t quantify them all.
Just two examples:
- An aggressive depreciation schedule could crimp short term earnings (due to accelerated depreciation) but set the table for higher longer term profitability—if an asset still contributes to profits long after its depreciation has fully been taken out of earnings
- A conservative deprecation schedule could make short term earnings look great now, while the company has to “pay” the price for a long time down the road
Due to this, and since there’s no black and white answer, it’s important to compare a company to its competitors in the same industry if there’s significant depreciation expenses on its PPE accounting (depreciation is by definition recorded on PPE).
Try to make sure that a company doesn’t have wildly different depreciation schedules for its PPE.
Where to Find Depreciation / PPE in a 10-k
Let’s take two examples of big businesses in the S&P 500 that have very different business structures but the same end market.
This should give us good representation on either end of PPE light or PPE heavy balance sheets, for illustrations of different capex requirements, depreciation, and general PPE accounting.
I want to look at Walmart ($WMT), which is much more capital intensive due to its need to invest in physical store locations– and Johnson & Johnson ($JNJ), which is very capital light due to being a consumer brands company.
First, Walmart. This company has 44.5% of its Total Assets in PPE, as you can see in this screenshot from their last 10-k:
To find the breakdown of this PPE, we need to scroll down to the Notes to the Financial Statements. In this case, they’ve put the breakdown inside Note 1, under Property and Equipment.
Nice of them to include the useful life inside that table as well. We can see that the large majority of Walmart’s PPE is all related to their physical stores (Buildings and improvements, Fixtures and equipment), and possibly much less has to do with distribution (Transportation equipment).
If we go back to the cash flow statment, it seems like depreciation each year is about in-line with new capital expenditures (~$10B per year).
We might imply from this that the company doesn’t seem to be involved with aggressive internal expansion, and there also probably won’t be much in the way of earnings surprises from creative PPE depreciation accounting.
Example: Johnson & Johnson
Now we move on to the 10-k for Johnson & Johnson ($JNJ), with 11.2% of its $157.7B in Total Assets comprised of PPE, net.
Of that PPE, we can see from their Note 1. Summary of Significant Accounting Policies, the estimated depreciation schedules:
And in Note 4: Property, Plant and Equipment, the outline of the company’s PPE assets:
Notice that in contrast to WMT, JNJ has the majority of their PPE in Machinery and equipment, which is probably used in manufacturing and distribution of their various products to retailers like WMT.
I’d also like to note that the PPE table outlines the accumulated depreciation, representing the portion of PPE in the balance sheet that has already been depreciated. The rest that will be depreciated in the future comprises the Net PPE.
Looking at the cash flow statement to compare current depreciation with recent capex and other capital allocation decisions:
Recent depreciation is much higher than recent capex, which means there will likely be less depreciation in the future, as the company appears to have reduced investments in its core business and instead looked for outside investments, such as acquisitions (as highlighted in the screenshot).
Is Big Data Really That Relevant?
Consider the following use cases of the cloud among today’s biggest companies in the S&P 500. So many things that the average consumer touches on a daily basis is directly impacted by big data and the cloud:
- Facebook: serves ads based on user data and activity on its platforms
- Amazon: collects user data and shopping preferences to customize pages viewed on its site
- Netflix: hosts large video files on the cloud, and serves these on its apps
- Apple: most apps in the App Store will store data on the cloud instead of forcing users to download data onto their phones
- Microsoft: has transferred many of its computer applications to the cloud so no download is needed, such as Outlook, Excel, and Skype
Between these companies and many more, the use of data is employed from marketing, to management of costs, to efficiencies in manufacturing and distribution, to computer-assisted features, to content consumption, to everywhere in between.
All of this data needs a place to be stored, and that requires expensive equipment for large data servers and networking– and that requires many types of computers and electronic components that need to be frequently maintained and replaced.
Data is Not Free: FB Example
Take Facebook again, as a quick example. Forget about their scandals and controversies around data leaks, and just observe the physical impact of sorting and storing billions of users’ digital data.
Back in 2011, Facebook built its first in-house data center.
Today, it has announced a new $1B facility in Singapore (1.8 million square foot), and had 6 “cloud” campuses as recently as 2019, and continues to grow its network and its business (and subsidiaries like Instagram, which migrated from AWS to Facebook servers) continue to thrive.
Let’s look at an example of Facebook’s PPE, to show just how different its balance sheet looks from the average company due to its massive data storage needs.
As found in the latest 10-k for Facebook, Note 6. Property and Equipment: Network equipment in 2019 was $17B out of $35.32B in total PPE, a staggering 48%.
On top of that, much of the other PPE such as Land, Buildings, and Construction in progress either directly lead to the construction of more data centers or in-directly or directly support the data centers themselves (you gotta store the data centers inside buildings after all).
These data centers are like other types of “old hat” PPE, they also depreciation in value over time and need to be maintained and replaced.
Facebook here has been vague with their Network Equipment Useful Life, reporting a range as wide as 3-20 years, which could lead to wide variations in depreciation expense.
Other competing companies with large percentages of PPE in networking and data centers reported shorter lifespans in their 10-k’s:
- Microsoft computer software: 3-7 years
- Microsoft computer equipment: 2-3 years
- Amazon servers: 3 years
- Amazon networking equipment: 5 years
- Google IT Assets: 3-5 years
- Google servers: 3 years
- Google Network Equipment: 5 years
Big Data Statistics
I read a fantastic report which put some more relevant statistics to this new digital phenomenon.
- They estimated that every person in 2020 will generate 1.7 MB of data in just 1 second, which would be 86.4 GB in 24 hours, by the way.
- Netflix, with its large volumes of heavy data-intensive video files, estimates it saves $1 billion by retaining customers from big data!
- Another crazy estimate puts 97.2% of all organizations as needing to invest in big data in one shape or form.
While of course all of the data means lots of costs, it also means a physical need for storage of this data.
While obviously most business won’t need to invest in physical server infrastructure like Facebook did, with all of the data generated by the vast majority of businesses now and in the future, all of this data will need a place to live– and if all of that demand can’t possibly be stored by the big cloud hyperscalers (Amazon Web Servies, Microsoft Azure, Google Cloud), well the market will find a way as it always has done.
If more websites like Facebook begin to spring up with massive data storage needs, and the economics dictate that money is saved by building their own data infrastructure, well then much of it will start to show up in financial statements everywhere as PPE.
With much shorter PPE life cycles than other types of equipment, network equipment and computer software will need to become a primary focus relative to other accounting metrics, and its impacts on commonly used balance sheet metrics like P/B or ROE could be significant and permanent.
Data Centers as Large PPE [Examples]
- Amazon (AWS)
- Total Assets = $162,648
- PPE, net = $61,797 (38% of assets)
- PPE, gross = $95,770
- Equipment = $54,591 (57% of PPE)
- Microsoft (Azure)
- Total Assets = $301,311
- PPE, net = $44,151 (14.7% of assets)
- PPE, gross = $87,348
- Computer equipment and software = $41,261 (47.2% of PPE)
- Google (Google Cloud)
- Total Assets = $232,792
- PPE, net = $59,719 (25.7% of assets)
- PPE, gross = $82,507
- IT Assets = $30,119 (36.5% of PPE)
I find it interesting that we are on the cusp of these huge FANG stocks, now thought of as big tech, becoming hugely capital intensive business.
Where these businesses and other technology businesses like them are usually thought of as “capital light”, we might need to think of these big data center empires (AWS, Azure, Cloud) as needing high PPE investments similar to a vertically integrated Exxon Mobil or physical retailer Walmart.
Scale is not infinite, even in the data world.
Data consumption and storage has real world, physical limits, and high capital requirements, as can be seen with these data centers– which are rapidly becoming large PPE components of balance sheets.