At long last, we pick up today with Part 2 of this case study on the Dell leveraged buyout announced in February.
The timing is good because a bunch of factors have changed since the last time we looked at this deal, and it’s unclear who will “win” or if anyone will actually end up acquiring the company.
The more confusing the scenario, the better it is for a case study, right?
Yes, that’s what I thought… so let’s proceed with the revenue and expense side and build in some fun scenarios:
Part 1 Recap
If you have not already read and watched Part 1, go do that now or you’re going to be very confused by everything that follows.
In Part 1 of this case study, I went through the key documents and sources of data we’re going to use, including equity research, industry research, the company’s historical filings, earnings call transcripts, and more.
You also learned how to set up the basic assumptions and some of the “quirks” specific to this deal, like the unconventional Sources & Uses setup and transaction assumptions (due to the offshore cash set to be repatriated and Michael Dell’s rollover).
This time around, we’re going to build on all of that and move into the revenue and expense assumptions.
Sometimes they will just tell you, “Assume XX% for revenue growth, YY% for margins, and the following numbers for working capital.”
But we’re taking a more complex and nuanced approach here by looking at alternate scenarios and using channel checks to come up with numbers that differ from the consensus.
The Tables Have Turned, Trebek!
Since the last time I wrote about this deal, various other bidders have emerged from the forest and set their sights on the company.
Blackstone and Carl Icahn are the most serious ones, and as I write this they have both submitted alternative bids, at (theoretically) higher prices.
I write “theoretically” because both proposals are fundamentally different in that they may leave a portion of Dell’s common shares outstanding and publicly traded, in an attempt to allow large existing shareholders to participate in the potential upside.
In that sense, neither offer is truly comparable to the original Silver Lake offer – because in that one, Silver Lake would acquire all shares (except for Michael Dell’s rollover) and the company would go private.
If you have some extra time and want to learn those bids in detail, go and take a look.
In this case study, we’re going to stay focused on the original offer because that’s the task at hand.
And in real life, of course, you don’t have time to go off on these side quests and then fail to level up for your boss battle.
Revenue and Expense Scenarios: Your Video Tutorial
(I highly recommend full-screening this video in 720p so you can see everything better.)
Table of Contents:
- 0:52 – Recap of Part 1 – Finding Information and Setting Up Assumptions
- 3:04 – How the Deal Has Changed (Blackstone and Icahn)
- 3:51 – How to Tackle Revenue and Expense Scenarios
- 7:07 – Why We’re Using Mostly Market Size and Market Share for Revenue
- 9:10 – How to Find the Numbers and Use “Channel Checks”
- 11:29 – Overall Market Growth Rates
- 13:12 – Dell’s Share and/or Growth in Each Segment
- 20:31 – Triangulating the Numbers
- 25:41 – Expense Scenarios
- 29:29 – Rest of the Model and Working Capital
- 31:41 – Linking the Statements
- 34:06 – Recap and Summary
Here are the documents you’ll need for Part 2, along with all the documents for Part 1 as well:
Here’s the textual version of the video tutorial above:
How to Tackle Revenue
Before attempting to make assumptions for revenue, you need to ask yourself a simple question first: which segments should you use?
In other words, should you use a product-level split (e.g. servers vs. software vs. services vs. desktops vs. laptops), or a “business segment”-level split (e.g. large enterprise vs. public vs. SMB vs. consumer)?
It’s a tough call here because Dell discloses its operating income for the business segments, but does not do so for the product segments (though some research analysts have estimates for that).
But remember: we are more concerned with revenue than with expenses and operating income… and there is much better data for the product segments, so we’re taking that approach.
Also, that split more accurately reflects the nature of Dell’s business: growth in servers/networking, software, and services vs. declines in desktops and laptops… and those trends apply across different customer segments.
Market Share vs. Units Sold
Next, we need to figure out if we’re projecting revenue based on market size and Dell’s market share in each segment, or if we’re going to use a bottoms-up approach and use units sold and pricing as the key factors.
With simpler businesses that have fewer segments, units sold * average selling price is often a better approach that’s more grounded in reality.
But the problem here is that Dell is massive and offers so many different products at so many different prices that it’s hard to come up with even rough estimates for those.
Certainly, it doesn’t disclose that information in its filings.
Some research analysts attempt to estimate this for different segments, but that information isn’t complete or comprehensive enough for a full analysis:
Meanwhile, we do have plenty of data on market share and potential growth in the underlying markets across at least 3 of Dell’s key segments.
So we will tackle the revenue side by focusing on the market size of each segment and Dell’s market share in each one.
Where to Find the Numbers?
This is where the fun begins. You can see that we’ve used a combination of outside industry research, equity research with market data, and historical numbers disclosed by Dell for much of this on a historical basis.
For the projected periods, we can still continue to use some of those sources… but there’s a small problem here, which is that we need to come up with views that are different from consensus estimates.
One of the best ways to do that is to conduct a “channel check,” where you speak with distributors, suppliers, key customers, retailers, and other company partners to figure out what’s going on “at the ground level.”
It’s easy to sit there in a spreadsheet and make up numbers for revenue growth percentages, but you have no real insight into what it will really look like unless you go and speak with real people in the market.
And that’s what I’ve done for you here: please read this document.
This is representative of what you might find if you actually did this, and it’s also very close to what you might see if a fund you’re interviewing with gave you this information in the context of a case study.
Based on these findings, consensus estimates, the company’s internal presentations, and the historical numbers, we can arrive at our own view of the company’s growth prospects going forward.
The Overall Market(s)…
Based on the channel checks and other research, the key takeaways here are:
- Low to flat growth for desktops and laptops – maybe a low percentage in the first few years, and then declining to 0% or negative growth.
- The servers / networking market size may increase at a higher-than-expected rate, due to comments from some of your sources and their plans to increase spending on these products.
Those are the only segments where we’re factoring in market size – software, services, and storage are all based on other metrics or are simple percentage growth rates instead.
Segment by Segment…
You should read the “Channel Checks” document yourself and verify that you understand everything there, but here are the key takeaways by segment:
- Servers / Networking: We’re more optimistic than the consensus because of Dell’s bundled offerings and growth in the past few years, plus because of how businesses are becoming increasingly data-driven… at the same time, though, we don’t quite buy into the 20% market share estimate because this is only a comment from 1 person.
- Desktops and Laptops: If you look at research, many expect Dell’s share to fall by a good amount. The more likely scenario, from these comments, is that Dell’s market share will stay roughly the same, perhaps declining a bit, but the overall market itself may decline due to pricing pressures and the shift to smartphones / tablets.
- Services: Expanding here is one of Dell’s top priorities, and the fact that their strategy with up-selling and cross-selling warranties and value-added services has been successful + solid uptake in foreign markets lead us to believe that growth here will be solid, perhaps even slightly higher than what most are predicting.
- Software/Peripherals: The main issue here is awareness. Yes, Dell has acquired lots of software companies but it seems like overall awareness of its solutions remains fairly low and there isn’t a great differentiating factor. So modest growth seems plausible, but it’s unlikely that these acquisitions and the software from them would result in a dramatic increase in growth unless Dell ramped up marketing and outreach efforts.
- Storage: We ran out of time for channel checks here, so this one will be a simple percentage estimate.
- Tablets: While Dell itself is aiming to get to over $1 billion in tablet revenue according to its investor presentation, we remain skeptical because of the dominance of Apple, Samsung, and Google in this market and the fact that its plans are not even clear at this stage. As a result, we’re not going to factor this in at all.
There are also some comments on expenses in this document, but we’re going to save those for the section on expense scenarios below.
Once you get initial estimates for these numbers in place, you need to “triangulate” a bit by doing the following:
- Base Case – This should line up with what you’re actually thinking based on the research and channel checks above.
- Downside Case – This will take a more pessimistic view that comes close to what analysts with a more negative perspective think, or that corresponds to some of the more negative comments you’ve received.
- Upside Case – This might be more in-line with management’s estimates from investor days and other presentations, but it shouldn’t be an exact match because their numbers are almost always wildly optimistic. So our numbers might be somewhere between the Base Case and management’s estimates. Historically, Dell’s internal estimates have been very far off so it’s not a great idea to rely on them.
- Street Consensus – As the name suggests, this should match up with the consensus revenue numbers, per the instructions. We don’t necessarily need to average every single analyst’s numbers – in this case, we’re paying more attention to the Morgan Stanley report with 5-year revenue projections.
All of this is fairly boring to show you in video form.
It consists of entering numbers, looking at other sources, going back and tweaking those numbers, and repeating that until you get a reasonable result.
It would be as exciting as watching paint dry, so I’ve skipped over a lot of the trial and error in the video and in this description – but that is how you would attempt to make your numbers line up.
There’s very little to no data here, and the channel check comments we’ve gotten are not very insightful.
But the basic point here is simple: if more of Dell’s business shifts to software/services, margins will increase, and if it takes more time to make that shift, margins will stay the same or possibly even go down due to price pressures.
You need to be really careful with these assumptions because even a 0.5% or 1.0% margin is a massive amount of money for a $50B+ revenue company like Dell.
You can’t just go in and say, “Aha! The operating margin will improve by 5% in Year 2!”
Where is that coming from? Are they paying employees less? Reducing the workforce? Paying less for leases?
Unless you have granular data for those variables, it’s better to keep this within a small range and vary margins by relatively small percentages each year.
The Rest of the Model / Working Capital and Other Assumptions
Once you’re done with the revenue and expense scenarios / assumptions, most of the rest of this model is straightforward.
I’ve made some changes over the version in Part 1, because life would just be boring if everything stayed the same:
- SG&A and R&D are combined into a single OpEx line item so that scenarios are easier to support.
- There are some additional items (e.g. financing receivables) in the operating assumptions schedule at the top because I realized they would be necessary once I started looking at the 3-statement projections.
The items here mostly follow what you’d expect (AR is linked to Revenue, AP is linked to COGS, Accrued Expenses to OpEx, etc.), and when all else fails, just make it a percentage of revenue (yes, that sounds silly but it is accurate for many operationally-linked items).
I am mostly using 4-year historical averages for these percentages, except for a few cases where there’s a trend in one direction.
For example, it seems likely that deferred revenue will increase to higher levels in future years as Dell shifts to a software/services model, so I’ve assumed that DR as a % of revenue increases in future years.
Linking the Statements
Most of these links are straightforward, with a few “hacks” here and there because I felt like making this more confusing and/or interesting, depending on your point of view.
For one good example of such a hack, see the “contra-asset” that I’ve mysteriously added to Current Assets (you’ll see why this is necessary if you look at how the BS and CFS are linked).
Be careful of alignment and linking in the right items. The big “Transaction Adjustments” section in the middle can make it easy to make mistakes when copying formulas, but we need it to support the adjustments we’ll make later on.
The debt schedules are mostly blank for now (except for interest, which is set to use beginning balances to avoid circular references).
That is intentional – we’ll get to all of this in Part 4.
I am also assuming numbers for a bunch of items like buying/selling investments, stock repurchases, dividends, etc. that won’t exist in the post-buyout period… those are there mostly just to verify that the model links correctly.
If you leave all of those line items blank, it’s easy to make a mistake where one or more links isn’t working and you never know because everything is set to $0.
The Cardinal Rule About Linking the Statements: Each item on the Balance Sheet must be linked to 1 item on the Cash Flow Statement, and vice versa.
If something is not linked at all, you’ve got a problem, and if something is linked more than once, you’ve also got a problem.
Those two mistakes account for 99% of the problems I’ve seen with Balance Sheets not balancing.
What to Do Next
Download the model and files and practice it yourself, or at least tweak the version I have and see what happens.
Then, think for yourself and do your own channel checks and research: do you agree with these findings and my examples above? Is your view of revenue growth and expenses / margins different in some way?
Have you found something that I’ve missed or haven’t considered?
My goal here is to get you to think for yourself and to give you the tools necessary to do that. So I encourage you to disagree with these numbers and present your own view based on what you’ve found.
Finally, subscribe to our YouTube channel so you can see this and additional future bonus videos / tutorials we’re releasing.
(Yes, there will be more videos soon – right now I am busy finishing a new course, so I haven’t had time to record more videos outside of that recently).
Coming Up Next…
Coming up in Part 3, we’ll delve into how you model the acquisitions and factor in revenue and expenses from them, and then we’ll go into the debt schedules in Part 4 and finally the investment thesis and presentation in Part 5 (NOTE: This order has been changed around – see below).
The Rest of the Series:
- Part 1: Data Gathering, Setup, and Model Assumptions
- Part 2: Creating Revenue & Expense Scenarios
- Part 3: Completing the Debt Schedules and Interest Expense Calculation
- Part 4: Building In Support for Post-LBO Add-On Acquisitions
- Part 5: Answering the Case Study Questions and Presenting Your Findings